PDA

View Full Version : clouds?



infinice
05-31-2018, 07:53 AM
Howdy folks-
I just parachuted in, to ask a question. Tho info is hard to come by (people tend to get preoccupied with explosions, fire, or fluids if they get interested in particles at all), i suspect that LightWave is the best general 3D app for skies/clouds.
Am i right? Or should i be looking elsewhere (like Terragen, or???)?
I have limited time so my priorities are very good clouds and great ease of use/production, not fabulous clouds but very difficult to make (i.e. Houdini).
Also, could somebody summarize user reaction to Volumetric Primitives vs. Hypervoxels? I've not seen any A:B screenshots comparing them. I know one thing tho: NewTek shot itself in the foot by discarding the catchy name, HyperVoxels, for the generic "volumetric primitives". (But NT seems to shoot itself in the foot a lot!) Even HyperVoxels2 would've been a better label. Anyway.
Thanking you in advance...
Canadian guy looking to make fly-thru's for HMD's.

What did the sadist do to the masochist? Nothing

SBowie
05-31-2018, 08:03 AM
I have nothing helpful to offer re: rendering clouds, howevah ... :)


I know one thing tho: NewTek shot itself in the foot by discarding the catchy name, HyperVoxels, for the generic "volumetric primitives"."Voluprims"?


Canadian guy looking to make fly-thru's for HMD's.A Vancouverite will certainly know something about what realistic clouds look like! (jk, my eldest brother lives there, and has since the 60's - lovely, lovely place.)

Axis3d
05-31-2018, 08:17 AM
LW 2018 does have the new volumetric primitives, which for clouds, does render a heck of a lot faster than doing them with Hypervoxels. I do still use the sprite mode of Hypervoxels for some things because sometimes you just need a 2d cheat that renders faster. I do wish that the new volumetrics had a sprite mode also.

With the new volumetrics, you can even do a fly-through the clouds, which with the old hypervoxels, would send your render times through the roof.

https://youtu.be/sICnY0c12kg

Marander
05-31-2018, 08:18 AM
Prometheus to the rescue ; -) He's the LWMOC (LW Master Of Clouds)

The answer very much depends on your requirements like rendering time, control, realism, if you do compositing etc.

In my opinion good looking clouds can be done with LW2018. Other applications offer similar but they are mostly particle based. What you have to consider is the high rendering time using LW2018 to achieve a grain-free render.

Specialized apps would be Vue (easiest for simple clouds but complex with nodes for more advanced stuff) or Terragen (best cloud quality in my opinion). They both offer high realism but depending on lighting also take time to render. I use both and with Vue xStream you have the best integration (but not compatible with LW2018).

Marander
05-31-2018, 08:26 AM
I think one issue that was mentioned is the limitation on scale, the new LW2018 volumetrics only work in small (and unrealistic) scales. Another problem is the limited usage because you can only use Nulls and the few Geometry shapes. However they seem to blend together nicely. But if you want to art direct your clouds to a specific shape, not possible n LW as far as I know.

Signal to Noise
05-31-2018, 09:37 AM
I found Ozone/CloudFactory (E-on Software), which is a pared down version of Vue focusing on clouds/atmospheres, quite good with LW albeit a tad slow. But since the e-on hack last year and the company's lack of transparency on the issue to their existing customers I have decided not to renew any of my maintenance subscriptions (Vue Infinite, Ozone, and Plant Factory) with them. They are pretty much offline now so any communication including license renewals are handled via email.

In light of the E-on fiasco I've auditioned Terragen Pro and it seems really good and I might go that route since I really like good clouds. Admittedly have not had time (it was a 30 day trial, now expired) to use Terragen in conjunction w/ LW. I've heard Planetside offer a "side-grade" discount, if coming from another package, to buy Terragen which might be a good incentive.

However, I also have been using, erm...learning, Houdini Indie which is supposedly very good at particles/dynamics/fluids/fx, but have not tried clouds per se yet. Still coming to grips with the learning curve really.

Marander
05-31-2018, 10:40 AM
I found Ozone/CloudFactory (E-on Software), which is a pared down version of Vue focusing on clouds/atmospheres, quite good with LW albeit a tad slow. But since the e-on hack last year and the company's lack of transparency on the issue to their existing customers I have decided not to renew any of my maintenance subscriptions (Vue Infinite, Ozone, and Plant Factory) with them. They are pretty much offline now so any communication including license renewals are handled via email.

In light of the E-on fiasco I've auditioned Terragen Pro and it seems really good and I might go that route since I really like good clouds. Admittedly have not had time (it was a 30 day trial, now expired) to use Terragen in conjunction w/ LW. I've heard Planetside offer a "side-grade" discount, if coming from another package, to buy Terragen which might be a good incentive.

However, I also have been using, erm...learning, Houdini Indie which is supposedly very good at particles/dynamics/fluids/fx, but have not tried clouds per se yet. Still coming to grips with the learning curve really.

Yep exactly the same here.

Actually e-on got hacked about 5-6 years ago and they only found out last December. I even tried to renew my Vue xStream maintenance via Mail but they got confused and sent me my license details instead so I didn't bother again.

Got Terragen Pro with a decent sidegrade discount too.

Both tools have their strengths and weaknesses anyway so I use both. Unfortunately both don't support the new LWO3 format or native 2018 integration (and it would actually be up to NewTek but I doubt they will do anything).

Houdini is also on the radar but not this year because I want to get more advanced with all the other tools I have first and Houdini needs alot of time.

prometheus
05-31-2018, 11:14 AM
I think one issue that was mentioned is the limitation on scale, the new LW2018 volumetrics only work in small (and unrealistic) scales. Another problem is the limited usage because you can only use Nulls and the few Geometry shapes. However they seem to blend together nicely. But if you want to art direct your clouds to a specific shape, not possible n LW as far as I know.

Not quite right, you can use the volumetrics in 2018 on particles, but not point clouds as far as I now..unless some node can extract the info.

I may have done some cloud stuff before...getting old, hard to remember :D

Check this thread specificly dealing with the old volumetrics/hypervoxels and the new volumetrics.
http://forums.newtek.com/showthread.php?155488-Hypervoxels-VS-New-Volumetrics&highlight=hypervoxel

And this thread may give you some hints..
http://forums.newtek.com/showthread.php?155945-Goody-Rays-and-Clouds-Again

Some cloud experiments here...
https://plus.google.com/photos/100944643113557837045/albums/5978023753291712209



To sum it up a bit short...best clouds are also a matter of integrating it fully with sky atmospherics, and Lightwave of today do not have that close to realistic physic volumetrics ..like an old plugin that was ogo taiki, if only that developer would have been in contakt with newtek..I think they tried ...and if they would have improved speed and settings, Lightwave would have a system that would be almost close to Terragen, Probably better han vue since the cloud fractals are not that great..and Lightwaveīs procedurals offer quite good stuff.

Ogo taiki pluign is only 32 bit, it wouldnīt work with 2018 since 2018 now is only 64 bit.
Sure it may be a dead horse, but there is nothing In Lightwave that comes close to a complete sky realism , as well as planetery render as Ogo Taiki did, the problem was render speed VS quality and also setting it up, too many quality settings as well.


The very best clouds per Hero item would be fluid simulated clouds, like from the Plugin for lightwave called turbulenceFD or other software which you then also may import to lightwave as VDB in 2018, it does however require either smoothing operations or very high detailed simulated cache files, so it doesnīt get blocky.

The very best clouds per Global sky scenery, Hands down Terragen...and itīs quite fast compared to many other stuff..though it is still slow in itīs nature..mostly all volumetrics is.

The new Volumetrics..

On the plus side..
+ Now the new volumetrics can look awesome, and with the new ligths and global scattering volumetrics, godrays can be created nicely..a bit slow in conjunction, but at least it is better than trying it with old hypervoxels.
+ Shading is more realistic, but it can be tricky to set it right if you are a newbie, but light and scattering offers much more cloud realism in lighting in clouds, though it may be more trickier to bring forth cloud noise detail within th cloud volume.
+ The edge softness of the new volumetrics is now much much better than hypervoxels.
+ Blending is much better than hypervoxels, but I think this is just when you deal with cloning nulls, when you use a particle emitter, the blending is not the same and not as good, the old dynamite blending was better for particles and point clusters.





On the minus side..
- The new volumetrics doesnīt work on point clouds...major let down.
- The new volumetrics Requires you to enter nodes just to set a hypertexture, the workflow has taken a step back, it takes longer time to just get a simple texture in there, unlike old hypervoxels.
- The new volumetrics doesnīt have a direct random particle scale, unlike old hypervoxels, the new system requires you to enter nodes and set up a network of various nodes to do such simple task, or you do have to switch to the other item, particle emitter and use itīs -+ settings, very awkward.
- The new volumetrics require you to select itīs item and no other item while tweaking it, since all the UI properties is build in to that Items property window, so you can not move around other stuff and then Use a seperate window to tweak volumetrics, unlike old hypervoxels which had a more direct Overview of deactivating or activating many voxel items in the same window, and that while moving other objects like aeroplane etc, but still having the controls open for you.
- The new volumetrics are not acceptable at close to real cloud scales when it comes to renderspeed and quality ratio, the scenes provided are at 1 m scale, and that is probably because of this drawback.
The old Hypervoxels IN 2018 however, does render Much faster than the new volumetrics in 2018 when it comes to larger scale, the old system has fixed levels of quality, and it isnīt dependend on the scale, while the new volumetrics is dependend on scale, where you set your step quality lower for large volumetrics...otherwise they will becom sooo slow to render, but at the same time that means too bad quality.

And correct, You can not use any geometry polygon shape to shape your volumetric item cloud, unlike some other software, only the use of cubic, spherical or cone shape.
If you use TurbulenceFD plugin, you can however "fill" a polygon shape on single framesm and add displacement noise, so that is a workaround and you would need to save that to a cachefile and reloadas VDB if you want it to work
with the new global volumetric scattering system for godrays, since turbulenceFD do not work with this new volumetric system.

https://lh3.googleusercontent.com/Z1cL-8wBS4IFe0TZk4_VSEP36cYb8K5pC6yqHr3aAav9_4w5RxgBAQa 6yny9MEI9pO3f5yL7WdF4pg=w460-h220
https://lh3.googleusercontent.com/Yl2Vkkk9uoCQKKvYPe35yBuw-GxyhxV_4bZtmaBFlkDLZBw771TLcehivuqrRvajaaZSfy6oV05 Pqt5qcA=w229-h220



But if used at lower scale without dependencies to other scene elements, and taking some time, Hero Clouds can look nice...godrays with the new global scattering can be done, but it is tricky, you need to have the light quite close behind the clouds(thatīs not physicly correct or near phyicly good settings), and further on add additional lights so the clouds themself do not get to dark and uggly.

http://forums.newtek.com/attachment.php?attachmentid=139902&d=1517614108


So Improved render speed at high step quality levels, and make it work on point clusters...Wonder what the dev team thought of here? did they simply not have time to implement the new volumetrics full out?
That is what I suspect..we may see it soon, but they seem to simply had the need to release it as it was for a deadline.
and put that D...texture choics directly in a drop down list ..next to where it actually should be one, then it will become really really nice.

prometheus
05-31-2018, 11:32 AM
I have nothing helpful to offer re: rendering clouds, howevah ... :)

"Voluprims"?

A Vancouverite will certainly know something about what realistic clouds look like! (jk, my eldest brother lives there, and has since the 60's - lovely, lovely place.)

I can imagine the cloud scenery over there, and in the us, the vast landscape and valleys, mountains, plains in all that granduer..is a receipe for dramatic weather changes and phenomena not ever seen here in sweden.
Oh and by the way...havenīt hardly seen any clouds for more than 15 days now..itīs a record may, not since the earliest recorded temperatures in 1739 have we had such warm month in may (Yay..itīs all global warming)
so itīs been clear sunny sky all these days, no drop of rain...and temperatures between 24-28 + celsius degrees ...Very Very unusal that itīs been like that almost the whole month, we do occasionally get a few days of very high temps ..but not this long time, so early that is high summer, when we usually have a bit chilly spring still.

rustythe1
05-31-2018, 12:02 PM
- The new volumetrics doesnīt work on point clouds...major let down.



so if you can use particles, just set an emitter on the point cloud to emit the exact number of points on the first frame and set the to emit from verts as the points are counted as verts in pointcloud,
like this,
141901

RPSchmidt
05-31-2018, 12:23 PM
I have been working with LW2018 volumetrics recently for the specific purpose of trying to build realistic clouds.

Here is my opinion;

1. Ease of use - Extremely easy to use. The best way to approach it (that I have found worked for me personally) is to download the Lightwave content for LW2018. You will find Lightwave scenes of just clouds in it, and those scenes provide an excellent starting point to build your own custom clouds.

2. Yes, as mentioned earlier, you do have to use nodes for texturing and that may be a downside if you aren't a fan of nodes. I have used them off and on (although I don't consider myself an expert by any stretch of the imagination) so I felt pretty comfortable using them.

3. At large scales, they do render slowly. I am currently working with a scene where the main cloudbank is around 4km square and on a dual xeon system it takes about 16 minutes or so to render a single frame.

Now, I do want to caveat that with the fact that some of my render settings are perhaps too much; first, I'm rendering brute force (no interpolation) so I'm taking a performance hit right there. Second, I'm using 125 rays, which is probably overkill.

However, I can't argue with the results. They are easily the most realistic clouds I have ever created using Lightwave, and setup was fairly straight forward.

prometheus
05-31-2018, 12:33 PM
so if you can use particles, just set an emitter on the point cloud to emit the exact number of points on the first frame and set the to emit from verts as the points are counted as verts in pointcloud,
like this,
141901

Two things...
I know that is a workaround, itīs too messy though.
I wait for them to fix this, or I simply do not upgrade.

prometheus
05-31-2018, 12:39 PM
I have been working with LW2018 volumetrics recently for the specific purpose of trying to build realistic clouds.

Here is my opinion;

1. Ease of use - Extremely easy to use. The best way to approach it (that I have found worked for me personally) is to download the Lightwave content for LW2018. You will find Lightwave scenes of just clouds in it, and those scenes provide an excellent starting point to build your own custom clouds.

2. Yes, as mentioned earlier, you do have to use nodes for texturing and that may be a downside if you aren't a fan of nodes. I have used them off and on (although I don't consider myself an expert by any stretch of the imagination) so I felt pretty comfortable using them.

3. At large scales, they do render slowly. I am currently working with a scene where the main cloudbank is around 4km square and on a dual xeon system it takes about 16 minutes or so to render a single frame.

Now, I do want to caveat that with the fact that some of my render settings are perhaps too much; first, I'm rendering brute force (no interpolation) so I'm taking a performance hit right there. Second, I'm using 125 rays, which is probably overkill.

However, I can't argue with the results. They are easily the most realistic clouds I have ever created using Lightwave, and setup was fairly straight forward.

1. Hypervoxels is easier to use...Even though I may be biased, I know how to use the new volumetrics anyway..I still think hypervoxels is easier to use, and by no Point Extremely easy to use for a newbie.

2. Itīs a downside to Everyone..not just for those who is not a fan of nodes, itīs a different matter that nodes may expand on what you can do..but it is by no way faster to acess any kind of hypertexture..Regardless if you know nodes as king of the nodes, I dare to challenge Anyone to set up a volumetric item with voxels HYPERTEXTURES faster than what I can do with old hypervoxels, that means however..no use of any kind of preset.

3. yes...at large scale they do render slowly..unless raising step size, but that defeats the purpose of having clouds with good quality, AGAIN...think twice or thrice on Why the scene content doesnīt provide scenes at more real cloud scale, and instead use 1 m.

Final act...Agreed, clouds can be..Is more realistic than before....But, the Caveat is scale denpendency, so you either render only the clouds in small scale only, or render small scale together with downsized items, or wait extremely long time for more real cloud scale.

rustythe1
05-31-2018, 12:55 PM
Two things...
I know that is a workaround, itīs too messy though.
I wait for them to fix this, or I simply do not upgrade.

Why is it a work around? You only need to set 3 settings in the first window, and this method opens up so many possibilities, such as being able to then dynamically effect the point cloud with forces/dynamics etc, create instances and clones very quickly, you could just use the l script commander to make the whole thing 1 click

prometheus
05-31-2018, 01:01 PM
Why is it a work around? You only need to set 3 settings in the first window, and this method opens up so many possibilities, such as being able to then dynamically effect the point cloud with forces/dynamics etc, create instances and clones very quickly, you could just use the l script commander to make the whole thing 1 click

Why isnīt it a workaround when you obviously have to do many more steps to acheive something that previously took lesser steps?
And no..you do not only need to set 3 settings in the first window, you need to add particles etc..

Quite simply, itīs faster to use only point clouds ...rather than having to add emitter, set nozzle, rate etc...it takes too long time compared to plain point clouds, and additionally calculating.
Now itīs a completely different matter if you want to expand on it with effects as you say...and if you know how and like it, fine but..itīs still not reducing the fact that it is a workaround.

And you do not dynamicly directly effect the point cloud, you effect the particle cloud inherited from the vertex emission, but now I am picky :) I get your point of what you can do with that approach, and it is interesting from VFX point of view if you want expandin on that.

Destroyed functions needs fixing, not standard techniques of old to compensate in more tedious ways ..thereīs nothing new in 2018 in that regards that you couldnīt do in 2015 with particle vertex emission.
Might add ...Ivé been a fan of using particles for hypervoxels, but that is not the point.
And ..there is the issue of random particle scale that is Way more tedious now than with Hypervoxels, either directly use partice -+ or go to nodes and mess with that instead of a direct value slider in hv tab.



https://www.youtube.com/watch?v=-6PDjQdPyus

For fly throughs I would recommend ..doing fluid sims and cache files and reload various pre-simmed frames, either in Lightwave or in other software directly or use other software and import to Lightwave and render with OPEN VDB.

There is something really nice coming up for some other software though that looks awesome for fluid sims of clouds and re-loading..which I wonīt be mentioning here.

rustythe1
05-31-2018, 01:21 PM
all you need is an emitter, change to per frame, set to stupid value on life and birth, change emitter to verts, done, it took me 6 seconds to set up, how is that such a chore to do?
so should your reply be, "yes, but you have to click the mouse a few more times so its not worth the time until some one adds 1 click I got a cloud functionality" and not "no you cant use point clouds"? and how come its more steps than adding a null, adding particles, and effecting particles into the shape you want when you already stated that volumetrics work with particles so at some point you must have done it?
because hypothetically you would have to have the same amount of steps if you were adding volumetrics to a solid object rather than just a null, as you would still need a way to define a particle/blending size at each point etc i.e do you want gaps between spaced out points or just one solid object, so even if they introduced that feature, it would still take as many steps as any work around to be able to have flexibility.

prometheus
05-31-2018, 02:39 PM
all you need is an emitter, change to per frame, set to stupid value on life and birth, change emitter to verts, done, it took me 6 seconds to set up, how is that such a chore to do?
so should your reply be, "yes, but you have to click the mouse a few more times so its not worth the time until some one adds 1 click I got a cloud functionality" and not "no you cant use point clouds"? and how come its more steps than adding a null, adding particles, and effecting particles into the shape you want when you already stated that volumetrics work with particles so at some point you must have done it?
because hypothetically you would have to have the same amount of steps if you were adding volumetrics to a solid object rather than just a null, as you would still need a way to define a particle/blending size at each point etc i.e do you want gaps between spaced out points or just one solid object, so even if they introduced that feature, it would still take as many steps as any work around to be able to have flexibility.

Itīs not more steps than adding a null, you are way out confusing things..itīs about point clouds, as having a pre-made point cloud ready in layout, adding a hv on to that will not even take you a second, donīt mix in adding nulls and particles..whatīs that got to do with point clouds?

And yes, why would you use particles when you could use the points, when point cloud hvs is the only thing you may need for that case, you are just propagating for a a slower process albeit may it be 5 seconds more..and given the relation/ratio between 1 second and 5 seconds, that is a LONG time for things that is a one click workflow and should be...it makes no sense to propagate for a workflow to compensate for the lack of properly implemented function.

All this debate wether itīs a workaround is most likely meaningless, especially since you do not have any decent assertments to verify that you can do it faster in anyway..I am pretty sure they will fix this as working on points in the way ahead, the question is how long it takes.

And no...I have not done it the way you think as you said I have done it..why do you mix in your thoughts of my appoach..which isnīt at all what I use to do, though I have used vertex emission on points for effects, as for vid sample, itīs only particles.

I used to do these sprite hvīs on displaced geometry, that is impossible with the new volumetrics, you can not use the new volumetrics on a mesh either, apart from that it lacks sprites, it would be insane trying to calculate particles on so many vertices.


https://www.youtube.com/watch?v=lBYUqMucoBw

prometheus
05-31-2018, 02:51 PM
because hypothetically you would have to have the same amount of steps if you were adding volumetrics to a solid object rather than just a null, as you would still need a way to define a particle/blending size at each point etc i.e do you want gaps between spaced out points or just one solid object, so even if they introduced that feature, it would still take as many steps as any work around to be able to have flexibility.

no you are Wrong...there is no need to define a particle blending size at each point, it depends on inital particle scale or if you use random particle scale.
and that is Irrelevant, that is secondary steps you would only do when needed, and not the initial steps to set volumetrics up on a point cloud.

It all depends on scene case scenarios as well, what you saw on my cloud sample session is pure particle emitters and it works well, random scaling...that is not the point though.
and random scaling works the same wether or not it is points or particles with the old hv system

random scaling is not the same with the new system, and it frankly looks awful compared to both the ease and results are when using old hypervoxels.
Same goes with trying to use that kind of vast particle emitters for clouds as I did, at a distance of12 000 meter maybe....try that and the needed volumetric scale on those particles and see how that looks with the new volumetrics, not good I can tell you.

prometheus
05-31-2018, 03:05 PM
Please rustythe1.
if you can ...record the process of bringing in a point cloud, make it an fx emitter and make it render with the new volumetrics and see how that goes and how long time it takes.
I think you would get a surprise.

and also...to get it working and show volumetrics you need to deal with changing source item for it to work with particles based on vertices point clouds..additionally more steps, I dare to say that you are not serious if you say it took you 5 seconds, got to take longer time than that.

prometheus
05-31-2018, 03:37 PM
You also need to constantly switch between these two element, the volume item..then the actual point cloud vertices..when arranging your clouds, like moving and placing it, then having to switch to the volume item and change scale, not satisfied, go back to the point mesh..change position, then go back and select the volum item, and change scale etc...a lot of unnecesary selecting steps when tweaking, with hvīs you just add the hvs to points, move the point cloud tweak the hv since that module window can be open even though you move other items, this is not possible with the new volumetrics since all itīs UI is in itīs selected form for properties.

I certainly do not like the workflow of setting up the new volumetrics, and how to move around, copy and paste, turn it on or off..so much better before.
I Certainly do like the way the new volumetrics look, up to a certain point.

We also lack the ability of using sprites, clips, using vertex map painted hv density on a mesh, but thatīs another story not related to this very case.
Either they need to keep old Hypervoxels working for a long long time, or they need to consider re-write a lot of the new volumetrics.

Currently it seems to be in early stages, it may of course evolve and I hope it does, it Certainly Isnīt any replacement for hypervoxels all kind of smörgåsbord flexibility.
I see it as a not fully working addition, promising but just that.

I must add, adding the old hypervoxels legacy is of course a painstaking workflow process in 2018, compared to it being directly there in 2015.3, I do not count that as a valid factor as to measuring speed on setting up old hypervoxels vs the new volumetrics on point clouds, that is a mishap and the guilty part is 2018, not how it used to be.

I must add, old hypervoxels in 2018 is now Much faster than in 2015 when stacking several nulls of hvīs and cloning them, previously it would halt the system when overlapping nulls, that is now gone in 2018...so there is some small good things, though this particular case is still pertaining the old hvīs.

Thomas Leitner
06-01-2018, 05:34 AM
I think one issue that was mentioned is the limitation on scale, the new LW2018 volumetrics only work in small (and unrealistic) scales....
What do you mean with "...only work in small scales"? We did cloud renderings in real scale.


...at large scale they do render slowly..unless raising step size, but that defeats the purpose of having clouds with good quality, AGAIN...think twice or thrice on Why the scene content doesnīt provide scenes at more real cloud scale, and instead use 1 m.

Final act...Agreed, clouds can be..Is more realistic than before....But, the Caveat is scale denpendency, so you either render only the clouds in small scale only, or render small scale together with downsized items, or wait extremely long time for more real cloud scale.
Narrowly: higher resolution = longer render times. Itīs ridiculous to expect something else. If you increase the scale by the same factor as you lower the resolution you keep the same render times. What did I miss?

ciao
Thomas

Marander
06-01-2018, 05:42 AM
What do you mean with "...only work in small scales"? We did cloud renderings in real scale.


That's why I said

"I think one issue that was mentioned..." - I haven't found time to test this myself in real scale (and I don't know if I ever would because I use other software for environment renders).

Good to know that it seems to work the same with a similar Size / Resolution ratio according to your experience.

prometheus
06-01-2018, 07:06 AM
What do you mean with "...only work in small scales"? We did cloud renderings in real scale.



Narrowly: higher resolution = longer render times. Itīs ridiculous to expect something else. If you increase the scale by the same factor as you lower the resolution you keep the same render times. What did I miss?

ciao
Thomas
Why do you narrow down and specify that I in some way don't understand that higher resolutions yields longer render times.
Not in any way has I given that impression.

You have to realise differences between high render times. ..an too high render times.

The current volumetrics doesn't hold up in speed comparing to terrace.
. .Neither does it seem to be as fast as the old hypervoxels..within 2018
And don't confuse that with hvs in 2015.

When I say it isn't at an acceptable render speed...and if you try to reach such speed..The quality would suffer too much.
Old hv in 2018 will work better in quality vs render speed. At real cloud scaling.

So don't wrangle my statements of too slow renders to such wrong perceptions.

Thomas Leitner
06-01-2018, 07:30 AM
Why do you narrow down and specify that I in some way don't understand that higher resolutions yields longer render times.
Not in any way has I given that impression.

You have to realise differences between high render times. ..an too high render times.

The current volumetrics doesn't hold up in speed comparing to terrace.
. .Neither does it seem to be as fast as the old hypervoxels..within 2018
And don't confuse that with hvs in 2015.

When I say it isn't at an acceptable render speed...and if you try to reach such speed..The quality would suffer too much.
Old hv in 2018 will work better in quality vs render speed. At real cloud scaling.

So don't wrangle my statements of too slow renders to such wrong perceptions.

Hi,
I donīt wrangle your statments: Your satements indicate that clouds with larger scales render slower than small ones. I donīt see this!
....If you increase the scale by the same factor as you lower the resolution you keep the same render times. What did I miss?...
So you get the same render times for the same quality.

ciao
Thomas

Tobian
06-01-2018, 08:04 AM
volumetrics (both the shader version and the volumetric kind) have a step size

https://docs.lightwave3d.com/display/LW2018/Primitive+Type+-+Volumetric

this confers to a sample resolution, which increases render time, as it goes up. For a small object (less than 1m) you might make the step size a few mm, but for real clouds you'd make it 1m or more, depending on their physical scale. As you can see it's envelopable, should you want to get closer to the object.

RPSchmidt
06-01-2018, 08:07 AM
1. Hypervoxels is easier to use...Even though I may be biased, I know how to use the new volumetrics anyway..I still think hypervoxels is easier to use, and by no Point Extremely easy to use for a newbie.

My experience with hypervoxels has been mixed, with the balance of that being "bleh". I rarely got the results I was looking for, regardless of what I was using them for (clouds, smoke, etc). Part of that is that I simply did not have the time to invest in mastering hypervoxels... I had production deadlines I had to meet. Previously, in almost any situation where I could, I would use composited stock video instead of hypervoxels.

In contrast, I opened up one of the cloud scenes from the LW content, saved it to a different file name, animated the texture frequency and amplitude over frames, changed some emission parameters, animated the whole thing moving over frames, and made a duplicate and changed the animation and VOILA... beautiful, gorgeous clouds rolling across the sky in about an hour.

I never could have achieved that with hypervoxels.


2. Itīs a downside to Everyone..not just for those who is not a fan of nodes, itīs a different matter that nodes may expand on what you can do..but it is by no way faster to acess any kind of hypertexture..Regardless if you know nodes as king of the nodes, I dare to challenge Anyone to set up a volumetric item with voxels HYPERTEXTURES faster than what I can do with old hypervoxels, that means however..no use of any kind of preset.

I'm not sure what you are referring to as hypertextures in connection to the new volumetric primitives. I just used a 3d texture node (turbulence) and that seemed to work quite well.

I also tried it with several other 3d texture nodes to see what types of effects I could achieve. They were all interesting and pretty fast.... add a 3d texture node, connect to the texture input on the material, new volumetric effect to play with.

So since I am part of "everyone" and I personally feel like screwing around with the old hypervoxel setup was a major pain in my backside compared to the far simpler process of using nodes, I guess I disagree on the "everyone" statement.

I am certainly not a master of the old hypervoxels (like you) or even the new volumetrics... but to me, the question seemed aimed at getting good clouds quickly. Volumetrics certainly does that, especially for those of us who may be hypervoxel-challenged.


3. yes...at large scale they do render slowly..unless raising step size, but that defeats the purpose of having clouds with good quality, AGAIN...think twice or thrice on Why the scene content doesnīt provide scenes at more real cloud scale, and instead use 1 m.

I have no problem working at a smaller scale because in almost every case I am compositing the end result with a 1080p or 720p composition that includes other effects. So really, the actual scale is irrelevant to me, as long as the rendered scene composites well with all of the other elements.

I remember you mentioning this in another thread and I think I asked you why it would be necessary to render an animation at real-life scale. I mean, my first test was at a very large scale, but I honestly can't see myself using the same scale for any future renders.

Anyhoo... I think the OP should definitely load a scene from the LW content and give it a spin around the block. It certainly can't hurt, and they will definitely get clouds.

Marander
06-01-2018, 08:16 AM
volumetrics (both the shader version and the volumetric kind) have a step size

https://docs.lightwave3d.com/display/LW2018/Primitive+Type+-+Volumetric

this confers to a sample resolution, which increases render time, as it goes up. For a small object (less than 1m) you might make the step size a few mm, but for real clouds you'd make it 1m or more, depending on their physical scale. As you can see it's envelopable, should you want to get closer to the object.

This makes sense and is actually the same with all Volumetric plugins and software I use (e.g. TFD). The difference is that they can optionally display the Step Size in form of a grid (back / side / top / bottom / all / off), which makes it much more clear and easy to set the correct Stepping. Again, poor UI design by NewTek.

prometheus
06-01-2018, 08:18 AM
Hi,
I donīt wrangle your statments: Your satements indicate that clouds with larger scales render slower than small ones. I donīt see this!
So you get the same render times for the same quality.

ciao
Thomas

Of course you are wrangling my statements, Look at what you are saying (indicates) then you throw in completely irrelevent assumption on what you think I think.

You are having blindfolds on, you can not compare the small scale VS the large scale and needed equal change in step size...the result isnīt the same.
Please go ahead and show me a good quality render of a large scale render and One eqaully changed in step size for a small scale and compare render time.

Do you really think all the scene content provided is made at 1 meter scale by mistake? if the same quality can be made by equal ratio change in step size for the larger clouds, why would they not provide that more real scale clouds? it doesnīt make sense.

No..I dare to say you do Not get the same render times for the same quality...when it comes to small VS large scale clouds, also look at the following post by RPSchmidt
.who donīt mind working on smaller scale, and yes ..you can approach it that way of course.

Thomas Leitner
06-01-2018, 08:22 AM
...Your satements indicate that clouds with larger scales render slower than small ones. I donīt see this!
So you get the same render times for the same quality.


....You are having blindfolds on, you can not compare the small scale VS the large scale and needed equal change in step size...the result isnīt the same.
Please go ahead and show me a good quality render of a large scale render and One eqaully changed in step size for a small scale and compare render time....

Here is a quick test. It's "Cloud_04.lws" from the LW2018 Content (without fog). First the original scale form the scene with a approx. 11m cloud (I didn't find any 1m cloud in the whole cloud content...). Second a approx. 1100m cloud with a similarly framing. This is done without much tweaking any setting, so not all aspects are equal. The resolution of the big one is 100 times lower than the resolution of the small cloud (=step size is 100 times bigger). The big one renders slightly faster (only seconds). I don't see any huge difference in quality.

SBowie
06-01-2018, 08:40 AM
The difference is that they can optionally display the Step Size in form of a grid (back / side / top / bottom / all / off), which makes it much more clear and easy to set the correct Stepping. Again, poor UI design by NewTek.Yes, yes, we all know that they are idiots ... we've all heard this from several repeatedly so it must be true. I was just wondering, have you submitted this great idea as a feature request? Or do you prefer to just flog the devs instead?

prometheus
06-01-2018, 08:53 AM
My experience with hypervoxels has been mixed, with the balance of that being "bleh". I rarely got the results I was looking for, regardless of what I was using them for (clouds, smoke, etc). Part of that is that I simply did not have the time to invest in mastering hypervoxels... I had production deadlines I had to meet. Previously, in almost any situation where I could, I would use composited stock video instead of hypervoxels.

In contrast, I opened up one of the cloud scenes from the LW content, saved it to a different file name, animated the texture frequency and amplitude over frames, changed some emission parameters, animated the whole thing moving over frames, and made a duplicate and changed the animation and VOILA... beautiful, gorgeous clouds rolling across the sky in about an hour.

I never could have achieved that with hypervoxels.



I'm not sure what you are referring to as hypertextures in connection to the new volumetric primitives. I just used a 3d texture node (turbulence) and that seemed to work quite well.

I also tried it with several other 3d texture nodes to see what types of effects I could achieve. They were all interesting and pretty fast.... add a 3d texture node, connect to the texture input on the material, new volumetric effect to play with.

So since I am part of "everyone" and I personally feel like screwing around with the old hypervoxel setup was a major pain in my backside compared to the far simpler process of using nodes, I guess I disagree on the "everyone" statement.

I am certainly not a master of the old hypervoxels (like you) or even the new volumetrics... but to me, the question seemed aimed at getting good clouds quickly. Volumetrics certainly does that, especially for those of us who may be hypervoxel-challenged.



I have no problem working at a smaller scale because in almost every case I am compositing the end result with a 1080p or 720p composition that includes other effects. So really, the actual scale is irrelevant to me, as long as the rendered scene composites well with all of the other elements.

I remember you mentioning this in another thread and I think I asked you why it would be necessary to render an animation at real-life scale. I mean, my first test was at a very large scale, but I honestly can't see myself using the same scale for any future renders.

Anyhoo... I think the OP should definitely load a scene from the LW content and give it a spin around the block. It certainly can't hurt, and they will definitely get clouds.

Sorry if I didnīt answer before, It Isnīt necessary to render at real-life scale, but you need to put in more work by rescaling meshes, other scene elements, as well as rescale dynamics, fluids, and also light behavior and shadow size..ultimately there is so many things to consider that the end result may not be realistic if not properly set up, while a full real life scale wouldnīt give you all that hazzle, you work on the scene animation and donīt let these workarounds get in your way, so itīs definitely doable to approach it by re-scaling if you want that..

Hypertextures, well that it the "old" term for adding a texture noise in to the voxels, I was just refering it to that to make it easier to distinguish from just texture..but obviously that didnīt work out so well :)
when I compare I mean that you are Required to click enter nodes, search or click in the menu for any type of texture, adding it, hover over that texture node output and feed it in to itīs texture output....now compare that process to simple pick a texture in the hypertexture drop down list...DONE much easier and faster than having to enter nodes...and again, donīt bring ..you can do presets up(not that I Have seen you do).

Yes..I may be quite good at hypervoxels and know it well, but even if everyone find it easier to get good looking clouds, clouds isnīt the only thing it needs to do...and you will eventually find it more painstakingly working with several
volumetric items if you really compare to how you can copy and paste, turn it on off in One single windo module..as said you have to stand on a volumetric item in order to change itīs properties, you can not move around other items without then having to switch back enter nodes again etc and tweak, unlike hypervoxels ...where you can move camera lights, or objects, or other voxels and with a small click change any given null voxel, it is much more faster and slick to work with that interface when it comes to that workflow...do not mistake that for " I donīt understand..I can get more realistic looking clouds now than with hypervoxels" That is Not the point..thereīs more to it than that.

As for you disagree on my statement on adding hypertexture would be faster for EVERYONE (Me sounding now like Gary Oldman in Leon the proffesional) You can disagree how much you want but...
I dare to assert that your disagreement is Very Very subjectiv, while I dare to state that a recording and comparing of the time it takes to set that up would prove it more objectively... that adding hypertextures in old hypervoxels is faster than entering nodes, adding pluggin it in to the nodes, and I am quite confident that No one can show the opposite, I am no newbie in nodes..as well as I am very well versed with hypervoxels...I feel I can do a good judgement on these to ways of workflow...and mind you, do not bring in I feel more comfortable with nodes I can acheive more and faster in to the equation, the issue is One thing only when I talk about hypertextures, itīs adding that texture and get a noise going, not how fast you can get good clouds or shading.

As for your scene comparison, you simple can not compare One already setup cloud scene in the new volumetrics, ( with the exclusion of some frequence settings or so) You do not have any similar already set up cloud scene for hypervoxels to compare at.

I could mix up good looking clouds for you in the old hypervoxels system (not that I will) and add a null or simple let you change velocity motion on particles, and voila you would get clouds moving, donīt think you give a fair comparison based on you using already setup stuff, try from scratch and add a null..make it volumetric, change to non pyroclastic, enter nodes, add texture in nodes, feed it to the texture input..go back to the properties and change scattering etc..those are the essential stuff.

Donīt get me wrong..I love the look of the clouds in the new volumetrics, as you may have seen I have posted some samples, itīs just the workflow..and that the step size need to be at a level that makes large clouds too slow to render, if not you will get very uggly grainy clouds, with the same scale with hypervoxels..I can get a way with certain settings and lower quality, but still looking smoother in quality than the new system, and that with a faster render speed at larger scale...Now, I do recognize that the old system isnīt as realistic in shading and lighting as the new system..so maybe itīs not fair to compare those two that hard..but if tweaked right you can get nice ones.

Sample video below is close to real scale clouds, and I can definitely say to you that I have tried this with particles and the new volumetric system...doesnīt work with the same splendid in speed and quality.
By all means, do a scene like this on several thousands of meters in the distance.

Check info on settings here...Oops wrong description initially..fixed now...
"I used lightwave... well...here we go again, more cloud testing with particles and hypervoxels, I have to mention that this was only a VPR draft renderer ..otherwise it would have taking to long time for initial testing, the VPR makes it at aprox 46 seconds per frame, 5000 particles, hypervoxels,fog. a final render would probably take around 7 minutes, but that also with radiosity. This draft vpr render has some stuttering due to vpr inefficiency when refining or the vpr fps settings, and it also has some flickerings due local density I think, but that is applied to smooth out hypervoxels. will post a thread on newtek forums for it with some image comparison of final render with radiosity vs the vpr draft. particles ar moving at 35 meters per second in the particles velocity motion(to fast) the hypertexture is turbulende, and the hypertexture effect is velocity translate, that gives a fluctuation and transform of the texture, but I do recognize it isn.t good enough since that is based on each particle. Instead I would also need to complement that with actually changing the birth rate texture I have in the particles birth rate tab, thatīs where the overall particle density is set. Software used-lightwave no postprocessing..direct vpr draft render, thus some stuttering and lower quality."


https://www.youtube.com/watch?v=N3i7v1b7kSQ

Marander
06-01-2018, 08:53 AM
Yes, yes, we all know that they are idiots ... we've all heard this from several repeatedly so it must be true. I was just wondering, have you submitted this great idea as a feature request? Or do you prefer to just flog the devs instead?

Please Steve - I have never called anybody names.

Do we really have to submit feature requests for things so obvious. Sorry look at the Volumetrics in 2018 and how they are represented in OpenGL UI, it looks horrible to put it mildly. There is no 3D application or plugin that doesn't get that right.

prometheus
06-01-2018, 09:03 AM
Please Steve - I have never called anybody names.

Do we really have to submit feature requests for things so obvious. Sorry look at the Volumetrics in 2018 and how they are represented in OpenGL UI, it looks horrible to put it mildly. There is no 3D application or plugin that doesn't get that right.

I was disapointed too with the openGL Representation of the new volumetrics..it doesnīt give you a volume feel At all, Unlike the old dynamite plugin...Many years ago, to bad no one has reached that level in developing this.
it could also show temperature colors, which new new volumetrics also can, but only as flat representaion, not as an estimated volume.



https://vimeo.com/9127056

SBowie
06-01-2018, 09:05 AM
Please Steve - I have never called anybody names.Fair enough, but the implications are clear in many if not most of your posts. It gets old.


Do we really have to submit feature requests for things so obvious.And yet, I don't recall ever seeing it mentioned in these discussions before ... so maybe the value of the idea is not so obvious to all, or its absence in the implementation would have arisen previously.

Tobian
06-01-2018, 09:05 AM
If you want it to ever happen, submit it as a feature request...

Because of a lot of terrible behavior and personal insults (not singling you out here) the devs just don't come on here, and I can't blame them...

SBowie
06-01-2018, 09:08 AM
Because of a lot of terrible behavior and personal insults (not singling you out here) the devs just don't come on here, and I can't blame them...This is true. I think the tone has generally improved, but there are still times when a little common courtesy would go a long way to make the forums more effective.

prometheus
06-01-2018, 09:11 AM
Fair enough, but the implications are clear in many if not most of your posts. It gets old.

And yet, I don't recall ever seeing it mentioned in these discussions before ... so maybe the value of the idea is not so obvious to all, or its absence in the implementation would have arisen previously.

depends on what you mean, these discussions? definitely not this thread, not the latest months...years ago we had topics covering this, included my showcase on the vid on the opengl from the dynamite plugin.

prometheus
06-01-2018, 09:14 AM
Here is a quick test. It's "Cloud_04.lws" from the LW2018 Content (without fog). First the original scale form the scene with a approx. 11m cloud (I didn't find any 1m cloud in the whole cloud content...). Second a approx. 1100m cloud with a similarly framing. This is done without much tweaking any setting, so not all aspects are equal. The resolution of the big one is 100 times lower than the resolution of the small cloud (=step size is 100 times bigger). The big one renders slightly faster (only seconds). I don't see any huge difference in quality.

I will look in to that more closely again..not today though, what update version of 2018 did you use for that? I have only installed 2018.01 discovery edition...so itīs not something that has changed over the latest updates I mean.
If you are correct...It may be a case of VPR not capable of finishing, refreshing properly...cause that is where I need to see the tweaking before I commit to a final render.

SBowie
06-01-2018, 09:18 AM
...years ago we had topics covering this, included my showcase on the vid on the opengl from the dynamite plugin.I'm not sure about the LW team's policy on such matters, but on the video side it would be likely that a suggestion that has gone without being implemented for years has likely fallen to by the wayside. Especially if a new feature means there might be renewed interest in it, it would be good to re-enter the feature request. (This doesn't necessarily mean it will be fulfilled, but there's no harm in bumping the topic.)

prometheus
06-01-2018, 09:23 AM
This is true. I think the tone has generally improved, but there are still times when a little common courtesy would go a long way to make the forums more effective.

Understandable, itīs easy to go and lash out when you are frustrated over functions in Lightwave or some part of the development, think I also have mentioned..what did the dev team think of ..etc, from time to time, didnīt find Maranders post to be that offensive, but then again..itīs not me doing the development either and too many re-marks referencing the development team may not be serving any good purpose here...if you use other non newtek forums, thatīs your choice and maybe better suited to vent your own frustration over something.

I agree that One could narrow down remarks and frustrations to the actual Technical issue, and maybe provide a suggestion on improvements etc..without the need of referencing the development team, better climate I guess.

prometheus
06-01-2018, 09:27 AM
I'm not sure about the LW team's policy on such matters, but on the video side it would be likely that a suggestion that has gone without being implemented for years has likely fallen to by the wayside. Especially if a new feature means there might be renewed interest in it, it would be good to re-enter the feature request. (This doesn't necessarily mean it will be fulfilled, but there's no harm in bumping the topic.)

Actually I did a thread when lightwave 2018 was quite new, called hypervoxels VS the new volumetrics...I did cover that openGL issue there..
Se here....end of the paragraph on post 98
http://forums.newtek.com/showthread.php?155488-Hypervoxels-VS-New-Volumetrics&p=1533818&highlight=dynamite#post1533818

or directly readable here...
"two features from the old dynamite plugin I would especially would want, that would be multi editing of several nulls, neither hypervoxels or the new system allows for that, in dynamite you could set several cloned nulls as slaves, so when tweaking the master, the other null volumes would change as well, but they are independent in the sense that you can move them seperately.
And of course, the openGL preview that actually shows a polygon volume representation, currently neither the new system or old hypervoxels provides a good enough openGL presentation. "

A better volumetric feel representation similar to that of the old dynamite plugin, would be helpful directing motions over and passing through clouds for instance, and some more stuff..without having to refresh vpr and slow the motion down.

Thomas Leitner
06-01-2018, 09:35 AM
I will look in to that more closely again..not today though, what update version of 2018 did you use for that? I have only installed 2018.01 discovery edition...so itīs not something that has changed over the latest updates I mean.
If you are correct...It may be a case of VPR not capable of finishing, refreshing properly...cause that is where I need to see the tweaking before I commit to a final render.

Hi,
I used LW 2018.2 (paid version of course) for this. However, I did another quick test in LW 2018.1 and it's the same (is even 11% faster). VPR works normal here.

ciao
Thomas

prometheus
06-01-2018, 09:43 AM
Hi,
I used LW 2018.2 (paid version of course) for this. However, I did another quick test in LW 2018.1 and it's the same (is even 11% faster). VPR works normal here.

ciao
Thomas

Uhmm..so you say 2018.1 is 11% faster?
I heard that some folks are experiencing some things suddenly going slower than faster, weird...

Ok, great that you test and great that you use the paid version, I use the paid version of the discovery edition..that means in paying the price of checker renders :)

I am to busy for some days, but will have to try and seriously verify your tests, if I am wrong..then I am wrong, but it is about the final renders in such case...canīt see that vpr is updating to finished render equally fast, though you might think there shouldnīt be any difference since itīs the same renderer albeit an interactive iterative refreshment render.

SBowie
06-01-2018, 09:43 AM
Actually I did a thread when lightwave 2018 was quite new, called hypervoxels VS the new volumetrics...I did cover that openGL issue there.I stand corrected, then, that it has not been mentioned recently. This said, however, I still don't see any mention of a feature request, and we've pretty much established that the devs have come to view the forums as a hostile environment*, so the chance that your post came to their attention is pretty small.

* Which situation, imho, could be improved if a small handful of people could self-moderate, by offering criticism without feeling the need to add insult to their posts.

prometheus
06-01-2018, 09:45 AM
Sorry for going a bit off topic, guess we need more post by the thread starter and what he thinks may work for his cloud projects, so he needs to bump in..
meanwhile regarding this..

And for open GL in 2018, color representation is a step forward though compared to old hypervoxels OpenGL, as seen in this image..I still wish it had a more volume look as seen in my dynamite sample vid.

http://forums.newtek.com/attachment.php?attachmentid=139706&d=1516819003

prometheus
06-01-2018, 09:51 AM
as a hostile environment*, so the chance that your post came to their attention is pretty small.


Most likely the case sadly, If I only had the time, have been thinking of making a website with Lightwave improvements from my list, and that means covering the topics I have done previously on threads, but more slick on a webpage with comparisons etc, though it may actually contain reference to other software on areas I think they should look at...just for reference on how to do things great, not a page I may post here and propagate for.

Tobian
06-01-2018, 09:58 AM
Prometheus, give it time, it's a brand new system with far superior volumetric shading and lighting, it just is missing a lot of the particle simulation/fluid systems to do more than blobs, but it's version 1. If you load and render openvdb files generated from software which can do such things (such as Houdini) you can see the power is there to be exploited soon. (or now if you want to use LW instead of mantra) I've also seen users using the volumetric light scatter in preference to the Octane lights, for speed and simplicity to set up (on facebook).

SBowie
06-01-2018, 10:12 AM
... though it may actually contain reference to other software on areas I think they should look at...just for reference on how to do things great, not a page I may post here and propagate for.There is a qualitative difference between a real suggestion along the lines of "I would love it if LW's feature X had something like this screenshot from application Z, because it makes it a lot easier to manage/visualize, what-have-you" and one which has an evident agenda and runs along the lines of "LW's feature X sucks, total crap compared to application Z; clearly the LW team is not paying attention to their users or the competition. Abandon all hope ye who enter here!". We've had quite enough of the latter, but really have no problem with the former.

prometheus
06-01-2018, 10:50 AM
Prometheus, give it time, it's a brand new system with far superior volumetric shading and lighting, it just is missing a lot of the particle simulation/fluid systems to do more than blobs, but it's version 1. If you load and render openvdb files generated from software which can do such things (such as Houdini) you can see the power is there to be exploited soon. (or now if you want to use LW instead of mantra) I've also seen users using the volumetric light scatter in preference to the Octane lights, for speed and simplicity to set up (on facebook).

Yes I know..though we have been giving it time :) and I am not getting younger...but yes, I do suspect the new volumetrics will improve over time, That is the mportance of Us actually raising criticics where needed, not silently go meh etc.
And yes..you should do your best and refrain from suck etc, but at some point you need to put a value in to how it is working...such as not good, bad, poor, horrible..before reaching suck or it stinks.
And of course..if you say it is bad, do your outmost to describe why...and maybe in reference to something that worked is working better.

And me not upgrading and bringing forth critics on some areas, just mean that itīs not at the level I want right now...I may or may not upgrade later on, a lot of factors depending on what I choose, including what kind of work I am occupying at the time etc.

My guessing is that I will not see myself upgrading this year, maybe not early next year either...maybe after that ..and after substancial changes or new features are introduced.






There is a qualitative difference between a real suggestion along the lines of "I would love it if LW's feature X had something like this screenshot from application Z, because it makes it a lot easier to manage/visualize, what-have-you" and one which has an evident agenda and runs along the lines of "LW's feature X sucks, total crap compared to application Z; clearly the LW team is not paying attention to their users or the competition. Abandon all hope ye who enter here!". We've had quite enough of the latter, but really have no problem with the former.


I believe the Lightwave team needs to be able to take critics..provided it is done so with respect.(granted that many threads has been out of line)..and I assume they also posses a certain degree of distance when someone may critic the software or the development, even hockey stars can not be on their top and will from time to time do a lesser good efort, or simply do mistakes, you simply do not cheer them on if they do not enter the ice and do their best or have no interest of winning.

As for the time of recent...It seems Newtek and moderation officers... have decided to be harder and not tolerate certain things, and fully understandable so...I just hope they have the sense of accepting some degree of critics, otherwise they simple wouldnīt be able to improve on things on the product or how they develop it with a certain respect to how users perceive their software to work.

I am no perfect horses..a, I need to be corrected from time to time, maybe I am wrong about certain things on the cloud issues..I am only human and make mistakes, so are peopler behind the lw team.

Great Steve that you make it clear, so it is ok for me to mention that software smoothie mixer from time to time then ;) as long as I say"I would love if Lightwave would have that kind of skin modifier..and preferably under a proper thread section.

Unfortunately I think due to the many agenda post of negative form, critics from people who only wants Lightwave to evolve to a more Amazing software..may have been seen as a part of that group that brings on such posts, where in fact there is quite a difference.

SBowie
06-01-2018, 12:41 PM
...I just hope they have the sense of accepting some degree of critics, otherwise they simple wouldnīt be able to improve on things on the product or how they develop it with a certain respect to how users perceive their software to work.Agreed. And there should even be room for expression of a level of frustration at times. I believe we've been tolerant of such in the past.


Great Steve that you make it clear, so it is ok for me to mention that software smoothie mixer from time to time then ;) as long as I say"I would love if Lightwave would have that kind of skin modifier..and preferably under a proper thread section.Of course. Pandering has never been expected ... just civility, and some reasonable limit to nagging and demeaning one another, whether directly or by obvious inference.


Unfortunately I think due to the many agenda post of negative form, critics from people who only wants Lightwave to evolve to a more Amazing software..may have been seen as a part of that group that brings on such posts, where in fact there is quite a difference.The difference in tone between passion or even temporary frustration and remarks intended to give offense is usually not that hard to discern, especially when it is part of a constant pattern.

infinice
06-01-2018, 01:17 PM
Thanks to all for answering! Helluva lot there to digest so i'll keep my fat ignorant mouth shut, except to say marander's remarks hit home.
This may sound off-topic but it sure affects my purchasing decision of LW etc: what drives me nuts is online only/non-downloadable manuals. I'm not an internet lover, and at $80 a month it's a lot of money for a technical requirement. (LW isn't $1000, it's $2000!, and $1000 year thereafter.). Sob.
PS: and thanks for volumetric primitives remarks... sound like a mixed bag- a lot like LW 2018.

jwiede
06-01-2018, 03:18 PM
Which situation, imho, could be improved if a small handful of people could self-moderate, by offering criticism without feeling the need to add insult to their posts.

Please point out where Marander's post contained such an insult?

SBowie
06-01-2018, 04:12 PM
Please point out where Marander's post contained such an insult?I did. It requires no imagination to read "Again, poor UI design by NewTek" as a dig. If this were an isolated incident, it would pass without notice, but the inclusion of "again" is an open admission that it does, in fact, form part of a pattern.

Not only was this 'addendum' in no way helpful, but quite the contrary: this sort of thing is without question one of the main reasons most of the developers refrain from even reading the forums, much less participate openly here. In addition, many members have PM'ed me over time to the effect that they are sick and tired of being assailed with a consistently negative drumbeat. To the extent possible, I resist such calls for preemptive moderation activity, but I certainly understand where they are coming from.

jeric_synergy
06-01-2018, 04:54 PM
I disagree: to me, there seems to be a pattern of consistent poor UI design that should be addressed. It's like whatever process the devs are using when in comes to implementing UI does NOT include thing like "consistency" and "low hanging fruit" and "multiple entity functionality", and I've been seeing this for like, what, twenty years?

Examples:
Consistency: "a" versus "fit" in the Node Editor.
Low Hanging Fruit: the entirety of the Layers Panel.
Multiple Entity Functionality: anyplace where it's impossible to affect entities en masse.

It's a systemic fault with the LW development process. Implementing big features doesn't seem to be a problem, but where the point where the code meets the user certainly does.

++++
to whit: (from 2015)
http://forums.newtek.com/showthread.php?139547-Rename-multiple-selected-surfaces-in-the-Surface-Editor&p=1419714&viewfull=1#post1419714

SBowie
06-01-2018, 05:26 PM
... and the beat goes on.

SBowie
06-01-2018, 06:52 PM
Alright, I've done a quick review of his most recent posting history, and belatedly realize I may have been a bit hard on Marander, for which I apologize. I'm just really, really tired of the tendency of a small chorus to pile on with passive aggressive sneering at every opportunity. I could go on about how discouraging it is for the devs, but I don't want to dilute the apology by doing so.

RPSchmidt
06-02-2018, 07:43 AM
LW isn't $1000, it's $2000!, and $1000 year thereafter.

Huh? LW 2018 full version is $995 and it's a perpetual license.

141917

Qexit
06-02-2018, 07:54 AM
Huh? LW 2018 full version is $995 and it's a perpetual license.

141917I think he is taking into account how much it costs him per month for internet access, $80 if I read him correctly, which adds $960 to the cost if that is the only reason he needs/uses the internet.. As the documentation is only available online, no hard copy or downloadable manuals, you can only get to it if you have internet access...so it becomes a necessary additional cost for him.

shrox
06-02-2018, 08:38 AM
I think he is taking into account how much it costs him per month for internet access, $80 if I read him correctly, which adds $960 to the cost if that is the only reason he needs/uses the internet.. As the documentation is only available online, no hard copy or downloadable manuals, you can only get to it if you have internet access...so it becomes a necessary additional cost for him.

One can save webpages as PDF files.

jeric_synergy
06-02-2018, 12:38 PM
I think he is taking into account how much it costs him per month for internet access, $80 if I read him correctly, which adds $960 to the cost if that is the only reason he needs/uses the internet.. As the documentation is only available online, no hard copy or downloadable manuals, you can only get to it if you have internet access...so it becomes a necessary additional cost for him.

Good point: it's like a "stealth subscription". :(

Marander
06-02-2018, 01:35 PM
Thanks guys for your feedback and backing me up!
Steve - apology accepted, thanks.

I'd like to explain where the word "Again" came from (UI issue with Properties panels) but I rather create a separate topic for that with a corresponding bug report.

Marander
06-02-2018, 01:41 PM
you can only get to it if you have internet access...so it becomes a necessary additional cost for him.

Yes, I miss offline html help (and context sensitive for that matter) too for LW. My 3D workstations are usually offline.

Signal to Noise
06-02-2018, 07:16 PM
Wow! Who'd have thought this thread would turn into 5 pages of debate?

Can't wait to see what happens when "Rainbows?" subject pops up.

50one
06-03-2018, 02:59 AM
Wow! Who'd have thought this thread would turn into 5 pages of debate?

Can't wait to see what happens when "Rainbows?" subject pops up.

If you're so curious why not create this thread and see?

prometheus
06-03-2018, 07:19 AM
Wow! Who'd have thought this thread would turn into 5 pages of debate?

Can't wait to see what happens when "Rainbows?" subject pops up.

The only plugin I know of that could do Rainbows, that was the old NaturefX plugin, which also did nice oceans, and it also had a way of using polygon shape to design clouds, that is all gone now for the Lightwave platform anyway...And the ogo taiki plugin, also not possible to use with 64 bit 2018, and the devolopment seem to be no more on that plugin.

http://www.asahi-net.or.jp/~pq1a-ogs/rainbow.jpg

I still await for that Full Global sky tool that can do both atmosphere with fog, haze, turbidy, glow...and infinite volumetric clouds.

Ozone plugin...never liked it because it misses so many things from the main application, and some quirks when used with lightwave, I am not sure if it had the option of rainbows, though I am pretty sure the main app did it.
Results could of course be great with this Ozone plugin, but obtouse as I am..I rarely jump in to something that doesnīt provide almost full potential.

Below sample..Lightwave 11 and ozone 6, I have been waiting so long for ozone to improve..but each new version seem to have very little new improvements...and the status of how a new ozone version will work with 2018, I do not know.


https://vimeo.com/39122792

gerry_g
06-03-2018, 12:40 PM
Ozone didn't work in 2015 (on a Mac at least) so I would say less than zero hope for 2018, still E-On had a 50% off sale last week so on the upside you cold have found this on the cheap, maybe thing are looking up

jwiede
06-03-2018, 01:20 PM
Ozone didn't work in 2015 (on a Mac at least) so I would say less than zero hope for 2018, still E-On had a 50% off sale last week so on the upside you cold have found this on the cheap, maybe thing are looking up

After everything they put owners through in last couple years, I'm pretty much done with E-On Software. Shame too, as CarbonScatter is still among best at distribution-instancing I've seen. Ozone's a bit of a money-sink for a couple years now, alas, very little added mostly paying for OS/pkg compatibility. Certainly not enough work done to justify upgrade costs, IMO. They also pulled some nastiness to maintenance subscribers when they went dark, claiming maintenance didn't include upgrades, etc.

Marander
06-03-2018, 01:23 PM
Carbon Scatter and Ozone have been discontinued last year unfortunately.

http://www.cgchannel.com/2017/07/e-on-software-to-can-carbon-scatter-cloudfactory-ozone/

jwiede
06-03-2018, 01:46 PM
Carbon Scatter and Ozone have been discontinued last year unfortunately.

http://www.cgchannel.com/2017/07/e-on-software-to-can-carbon-scatter-cloudfactory-ozone/

Yeah, "anyone whose maintenance was active as of June 2017" my hiney. As if most rational customers continued paying maintenance after receiving nothing in return for 2016 maintenance subscriptions. Riiiiiiight.

Anyway, getting off-topic, sorry.

Signal to Noise
06-03-2018, 07:47 PM
If you're so curious why not create this thread and see?

I would but I'm still working on unicorns.

erikals
06-04-2018, 02:59 AM
for Cumulous clouds and some Cartoony styles, consider VUE...


https://youtu.be/6JuQumMXgtE



from what i was able to figure out, VUE Esprit ($200) should work fine

i think it might go on my next "to buy" list...

(the VUE render will then either be composited or used as a LW background image/video)


Houdini can also kick it - https://www.youtube.com/watch?v=wGgglh4_HOI
however VUE has 150 preset clouds, depends on the time you have available.

not sure about LW2018 yet

Marander
06-04-2018, 05:01 AM
...from what i was able to figure out, VUE Esprit ($200) should work fine

i think it might go on my next "to buy" list...

(the VUE render will then either be composited or used as a LW background image/video)


Vue Esprit is the low end version that has serious limitations and almost everything needs to be bought as addon (Animation, Nodes, Ecosystem, Weather effects, Export etc.). It is also limited in CPU cores, output resolution, library etc.

Believe me, for Vue you will need as many cores as possible (even for good stills, your render will take long with 12 cores, not to think of animations), as far as I remember Esprit is limited to 2 cores. Backplates require high resolution output.

If you want to composite the output you want to use EXR which is only available in Infinite and xStream as far as I know. The same for camera sync. I wouldn't be surprised if there are limitations in the quality settings, shading and lighting models. Note that there is no upgrade path from the Hobbyists line of products to the Professional packages.

Best features for the money you will get from Houdini but the landscape tools are still limited and the learning curve is steep.

My advice: get Vue Infinite or xStream (but LW2018 is not supported) or get Terragen Professional. But all depends on your requirements of course.

Note also that all communication with e-on is via mail and there are no online services (no download page to get updates or licenses, no online store, no forum, no bug tracker) since half a year now.

CaptainMarlowe
06-04-2018, 05:22 AM
Yeah, I’m done with this software and still have to get the royalties that were pending on cornucopia3d when they shut it down. I will certainly fall for Terragen in the future, though.

Marander
06-04-2018, 06:02 AM
Little correction to my previous post:

- Vue Pioneer is limited to 2 CPUs, Vue Esprit can use up to 12 (virtual) CPUs
- Render size limitation only for animations (720p), but requires animation addon

But:
- no AVX support (!) - that makes rendering much slower
AVX optimizations for dramatic increase in rendering speed (requires an AVX compliant CPU)
- even many UI / viewport and navigation limitations
- some camera limitations
- some material limitations
- rendering limitations
- lighting limitations
- no render passes, alpha channels and serious import / output / export limitation
- no scripting
- many functions require addon modules

So if you only want nice looking background skies ok, but for good looking and custom clouds you would also require the Nodes / Functions addon.

But: You get Rainbows ;-)

RPSchmidt
06-04-2018, 07:11 AM
I think he is taking into account how much it costs him per month for internet access, $80 if I read him correctly, which adds $960 to the cost if that is the only reason he needs/uses the internet.. As the documentation is only available online, no hard copy or downloadable manuals, you can only get to it if you have internet access...so it becomes a necessary additional cost for him.

I'm still not sure I understand his point; I mean, not to be rude or anything, but how much did he just spend on posting here?

Or, stated another way, is he saying that if he didn't have LW, he wouldn't have internet access? Hmmmm.

erikals
06-04-2018, 10:13 AM
Little correction to my previous post:

- Vue Pioneer is limited to 2 CPUs, Vue Esprit can use up to 12 (virtual) CPUs
- Render size limitation only for animations (720p), but requires animation addon

But:
- no AVX support (!) - that makes rendering much slower
AVX optimizations for dramatic increase in rendering speed (requires an AVX compliant CPU)
- even many UI / viewport and navigation limitations
- some camera limitations
- some material limitations
- rendering limitations
- lighting limitations
- no render passes, alpha channels and serious import / output / export limitation
- no scripting
- many functions require addon modules

So if you only want nice looking background skies ok, but for good looking and custom clouds you would also require the Nodes / Functions addon.

But: You get Rainbows ;-)

thanks Marander, great info   :king:

i need to look closer at it, it seems. maybe Houdini is the way...
or even LW Turbulence...

https://i.imgur.com/ycgfVzI.png

Marander
06-04-2018, 10:38 AM
thanks Marander, great info   :king:

i need to look closer at it, it seems. maybe Houdini is the way...
or even LW Turbulence...

https://i.imgur.com/ycgfVzI.png

Welcome Erikals

Maybe you could compare the two Terragen packages too... it has LWO support and a better feature / price ratio. TFD is great too for clouds (or LW2018).

prometheus
06-04-2018, 12:29 PM
thanks Marander, great info   :king:

i need to look closer at it, it seems. maybe Houdini is the way...
or even LW Turbulence...

https://i.imgur.com/ycgfVzI.png

I would recommend checking out Lino Grandiīs youtube page and look at his latest uploaded vid, some interestin stuff for clouds there indeed.


Welcome Erikals

Maybe you could compare the two Terragen packages too... it has LWO support and a better feature / price ratio. TFD is great too for clouds (or LW2018).

Indeed, though the latest demo I tried with 2018...ment that the color shading for smoke was screwed up, it doesnīt show up properly in the OpenGL, That is set to have a texture previous or fluid channel for the smoke color..so you get
a completly black result in the openGL, you need to remove it ..and when doing so you can not get a proper color depth in that smoke color, in lw that used to work...though I am not sure which part it is screwing this up..the latest turbulence build, or 2018 lacking something.

but using a simple polgyon plane ..and use texture in the emitter tab of that fluid emitter, you can set a cloud texture density fluid emission, so you get some air/space between some parts of the clouds, then use various gradients to control height..so you will have different height pillars within that cloud, easy to create a cloud section that way, then you could save out cache files and reload, make another cloud set and be able two load various cloud sets.

But beware as if using TurbulenceFD, you can not use the new global scattering system, unless reloading as VDB format, then it will not be the shading volumetrics of turbulenceFD, rather the new volumetrics shading system, and that will work with volumetric scattering if you need godrays as well.

Unfortunately I think TurbulenceFD during its whole development...Still crashes in too many places, too often...not good, and it continues in 2018 for me.
I just loaded a simple scene some minutes ago, run the simulation a couple of times, probably just 4 minutes, then crash.


https://vimeo.com/38310716

prometheus
06-04-2018, 12:36 PM
for Cumulous clouds and some Cartoony styles, consider VUE...


https://youtu.be/6JuQumMXgtE



from what i was able to figure out, VUE Esprit ($200) should work fine

i think it might go on my next "to buy" list...

(the VUE render will then either be composited or used as a LW background image/video)


Houdini can also kick it - https://www.youtube.com/watch?v=wGgglh4_HOI
however VUE has 150 preset clouds, depends on the time you have available.

not sure about LW2018 yet

Actually for cumulus clouds, I would rather use the new volumetrics in lightwave, or use simulated fluids, e-ons cumulous cloud could be done best with metaclouds, but that was tricky to get the noise right I think, and a pain to setup with all that moving of several spherical metacloud objects, with lightwave and cloning only a couple of nulls, the results with right noise functions would be better...But the actuall reaction to sun and atmosphere would be lesser good, and much harder to set up a light for godrays if you want to than in Vue.

A standard issue with vue clouds is that they are distinctively noticable as vue clouds quite often..(if not tweaking them right) at least the many presets have too bad noise setups for the clouds.
The Very best fractals for clouds are Terragen hands down, and feathering and dissolve of smaller cloud parts breaking loose from the main cloud density ..that his handled very nicely with the new cloud models in terragen.
Terragen renders decently fast also.

If Terragen had the vue interface though..and workflow..I would have bought it in a second, I do not like the interface at all.

erikals
06-04-2018, 03:12 PM
Thanks for all info.

it does certainly look like i have to do more research.

very much appreciate the help


https://i.imgur.com/A3LuMaG.gif
back to the scope

jwiede
06-04-2018, 04:02 PM
I'm still not sure I understand his point; I mean, not to be rude or anything, but how much did he just spend on posting here?

Or, stated another way, is he saying that if he didn't have LW, he wouldn't have internet access? Hmmmm.

It's quite possible that for him, getting internet service to the location where the computer is location requires different/special service or connection. Different places, different market environments.

MonroePoteet
06-05-2018, 07:44 AM
It's quite possible that for him, getting internet service to the location where the computer is location requires different/special service or connection. Different places, different market environments.

My primary rendering system doesn't have any Internet access at all so I don't have to run a firewall, antivirus or other background "cycle hogs". I "vett" any kits / content on my laptop through an antivirus before moving them via shared drive to the render system. In addition, when I'm on the road I'm usually accessing the Internet through a hotspot, which is adequate for my business needs but has limited bandwidth. It'd be nice to have a downloadable version of the documentation, even if it doesn't have the fancy animations, etc.

mTp

RPSchmidt
06-05-2018, 08:38 AM
It's quite possible that for him, getting internet service to the location where the computer is location requires different/special service or connection. Different places, different market environments.

True, true.

I guess I'm just looking at the other available software that would generate usable, realistic clouds and among the commercial offerings, they all cost more than Lightwave and a large portion of them (if not all) have moved their user guides / help files to the net.


My primary rendering system doesn't have any Internet access at all so I don't have to run a firewall, antivirus or other background "cycle hogs". I "vett" any kits / content on my laptop through an antivirus before moving them via shared drive to the render system. In addition, when I'm on the road I'm usually accessing the Internet through a hotspot, which is adequate for my business needs but has limited bandwidth. It'd be nice to have a downloadable version of the documentation, even if it doesn't have the fancy animations, etc.

mTp

My primary work system has no internet access either. We have systems that do have internet access, but they are completely separate from our production systems.

I emphatically agree that it would be ideal to have a downloadable version of the documentation; but again, going back to the statement I was referring to, I don't think he would have significant cost savings using any other commercial software.

Marander
06-05-2018, 10:37 AM
...they all cost more than Lightwave and a large portion of them (if not all) have moved their user guides / help files to the net.

Not quite true.
- Terragen is cheaper than LW
- Vue (xStream) ships with a giant 1200+ pages pdf manual

Ztreem
06-05-2018, 12:28 PM
Don’t forget Blender. :)

sadkkf
06-05-2018, 02:21 PM
I haven't used Terragen in years. Is it capable of exporting skies to use as backdrops? Vue can export spherical and box type maps as HDR.

erikals
06-05-2018, 02:52 PM
Terragen looks neat, however, 50 minutes rendertime per frame seems a bit much ?


https://www.youtube.com/watch?v=87TkK1_avO0


Is it capable of exporting skies to use as backdrops?
yes/no, only as a image sequence, not vdb

prometheus
06-05-2018, 03:25 PM
Terragen looks neat, however, 50 minutes rendertime per frame seems a bit much ?


https://www.youtube.com/watch?v=87TkK1_avO0


yes/no, only as a image sequence, not vdb

That is what to be expected, same in vue around 45 minutes for decent cloud quality at maybe hd resolution.

and you would probably reach near that with lightwave, you may however get even faster renders I suspect with lightwave...But a big but is that lightwave do not calculate the rest required for the very very best realism and that includes sky and volumetric scattering, if you add volumetric scattering it would raise the render times as well in lightwave.
Terragen offers a more complex lighting within the clouds and more realistic ones including sky and fog.

I dare to say that the cloud and scattering shading in Terragen is more complex and realistic for clouds than the new lightwave volumetrics.
simulating fluids in a certain free software can also yield nice clouds.

erikals
06-05-2018, 04:13 PM
hm, so still high rendertimes  https://i.imgur.com/bcwLfNX.gif

yep, it certainly looks like LightWave can kick those rendertimes,
agree though that the quality of especially Terragen looks a tad better.

for Cartoon-style i think VUE Esprit ($200) should work out alright
for Cumulus realism Terragen seems to be the way
for Cirrus i recon Photos / PhotoShop is good

prometheus
06-06-2018, 04:59 AM
hm, so still high rendertimes  https://i.imgur.com/bcwLfNX.gif

yep, it certainly looks like LightWave can kick those rendertimes,
agree though that the quality of especially Terragen looks a tad better.

for Cartoon-style i think VUE Esprit ($200) should work out alright
for Cumulus realism Terragen seems to be the way
for Cirrus i recon Photos / PhotoShop is good

Photos is good, and sometimes all that is needed, or at least filmed clouds for animations, it requires a bit of planning though...and actually taking notice of where the sunlight is originating from at pitch and heading angles.
I guess the "real pros" is doing research properly and also have the resources to capture photage at several hours a day, to cover all the range of how the sun is located.
Going at it backwards, jut pick som photoge and then having to adust your light source after how the cloud photoge is created, that means you do not get full freedom to direct the light in the scene as you would want, it is governed by the actual cloud photo..otherwise a missmatch in lighting and realism.

Cartoon, well you can do cartoony cloud stuff with Lightwave as well, you can use various lighting models with old hypervoxels, and you could most likely change the new volumetrics to be cartoony as well...if that is what you aim for that is.

Otherwise with Lightwave 2018, there is now many options with interchange with other software such as houdini and blender, where you can do a blender smoke sim, and reload either the full sequence, or only single cache frames as static clouds and render in lightwave, that gives you True smoke raising behavior and vorticy that isnīt created based on procedural noise, that is a plus in realism.

The shading of such sim is all handled by Lightwaveīs new volumetric engine, and you can of course boost scattering, emission and color in various ways with nodes.
Also houdini (even in apprentice Version) allows for doing fluid sims in houdini, or use modelled/sculpted geometry and turn it to a cloud object and save out as openvdb, you do need to take measures of adding a certain smoothing vdb filter when exporting that vdb file, or it will be very blocky despite which interpolation mode you set in lightwave openVDB tab.

For blender I am not sure if there is a similar smoothing filter, what I do know right now is that you most likely just need a high res simulation and do not use linear interpolation for the volumetrics in lightwave when you load such vdb, and do not export compressed blender fluid sims.

And to note..I havenīt completely verified it, but the old volumetrics seems faster than the new one in 2018, which means that cloud fly throughs isnīt that bad in rendertimes it seems, (just done some initial tests)
and you have the option of reducing realism by unchecking textured shadows, lower some quality etc to gain a bit faster cloud fly throughs.

And finally there is an option to use the baked hypervoxels techniuqe, requires some workarounds and you do not get instant feedback in vpr when you tweak the clouds..so itīs a process of guessing, and baking, guessing and baking, and you need to work on low scale items.
The result can be quite nice clouds ..though it may tend to get more softer thinner as a volume and sharper in the edges ..rather than the opposite which is what you mostly want, but it can probably be tweaked to a different result with some gradients.

See various youtube tests from others..
https://www.youtube.com/results?search_query=lightwave+clouds+baked

The cheapest method is to use the cheapest software out there...no surprise right, the surprise is that it yields so good results being that cheap.
But if you do need..and have lightwave 2018, I recomend take a deeper look at various fluid simulators and importing the VDB exported file to Lightwave 2018 and render there, as mentioned..houdini also has itīs cloudFX tools allowing for a very nice OpenGL preview and allows for modeling your cloud with polygons, or import a pre-modelled shape, or sculpt in in some some software, or use particles, points as cloud shapes...not necessary to simulate though that is perhaps to prefer.

A side note, I think the VDB suppor is great they have added, though I find it a bit unstable when you tweak some shading settings, it is a bit proned to crash too often than I like on my system.

RPSchmidt
06-06-2018, 07:00 AM
Not quite true.
- Terragen is cheaper than LW
- Vue (xStream) ships with a giant 1200+ pages pdf manual

True, if he is just looking for a terrain / sky generator.

I guess I read into it that he was looking for a 3d solution that was also capable of generating great clouds. Otherwise, why would he be looking here, when Terragen is so much more affordable?

prometheus
06-06-2018, 07:35 AM
There is also some realtime volumetrics with realtime clouds and godrays for some of the s.k game engines out there, looks semi dcecent, I noticed one of particular interest, but I am not sure it is fitting to link or name it here.
it may be better to p.m me if someone is interested to know exactly to which engine I refer too.

Edited..Just found out about a realtime simulation engine, that seems to knock the door out of some of the others, probably the best Realtime clouds and skies Ivé seen up to this date.

jwiede
06-06-2018, 11:32 AM
hm, so still high rendertimes  https://i.imgur.com/bcwLfNX.gif

yep, it certainly looks like LightWave can kick those rendertimes,
agree though that the quality of especially Terragen looks a tad better.

So you've seen animated render results out of LW2018 where the clouds/sky looked "only a tad bit worse" than, say, that turntable anim from Terragen? Please provide some pointers or references to them?

I've not seen any LW2018 animations that showed cloud/sky lighting quality anywhere close to the quality shown in that turntable rendering. I've seen some decent stills with nice cloud/lighting interactions, but most were also post-processed. Generating a consistent, similar-quality turntable animation as that Terragen example is much more difficult, so please provide refs/pointers to where you saw such example from LW2018?

Asticles
06-06-2018, 12:10 PM
https://www.youtube.com/watch?v=ObNJoJIAx4s

Just saw this, awesome.

erikals
06-06-2018, 01:38 PM
a realtime simulation engine, that seems to knock the door out of some of the others
shouldn't be a problem mentioning it, since it is not a direct competitor of LightWave

i did check out both Unreal/Unity, to my surprise Unity had some cool features, but not quite what i (personally) was looking for


jwiede
many should know by know what LW2018 can / can't do.
the biggest minus about LW2018 regarding clouds is how long it takes to set up the actual scene.
the word "a tad" is relative, you might feel the word should be "much"



https://www.youtube.com/watch?v=xiaagpK6mCQ

prometheus
06-06-2018, 02:09 PM
shouldn't be a problem mentioning it, since it is not a direct competitor of LightWave

i did check out both Unreal/Unity, to my surprise Unity had some cool features, but not quite what i (personally) was looking for


many should know by know what LW2018 can / can't do.
the biggest minus about LW2018 regarding clouds is how long it takes to set up the actual scene.
the word "a tad" is relative, you might feel the word should be "much"



Unity had some interesting stuff, while I think unigine is looking better, but apart from that the cloud motion is mostly happening on the outer rim of the cloud for the cloud fluctuation, in real life and with other velocity changes of a cloud, it also happens more on the whole cloud volume.

Realtime clouds...

https://www.youtube.com/watch?v=tui5-x2i2zw

I am also revisiting ogo taiki again for lightwave 2015 (only 32 bit..thus not compatible with 2018 since 2018 is nowadays only 64 bit), and found some old scenes I had and retweaked, and updates reasonable fast in VPR, donīt think I have tried it in 2015 before actually. as always..checking the shadow layers for godrays will be a bit slower..need to check if I can optimise it.

erikals
06-06-2018, 02:34 PM
yes, Unity ain't bad at all, just not what i'm looking for.
edit; Unity can do similar stuff with an addon >
https://www.youtube.com/watch?v=DEgb-R7uqFQ

i'm mostly curious to Cumulus clouds in LW2018,
Turbulence can somewhat do it.
Procedurals seems to be kinda alright
...Cumulus in LW looks to only "kinda" get there though ?

in regards to Turbulence, i'm far from an expert, here is what others did using VDB

https://img1.cgtrader.com/items/860653/53d139c636/large/realistic-3d-cloud-vdb-format-3d-model-max.jpg

i saw some Procedural Cumulus in LightWave some time back, can't find the link though.


ogo taiki
did revisit it too days back, wasn't quite the quality i hoped for.

edit; checking, for Cumulus, it looks to be DP Gardner clouds procedural.
perhaps i should make further tests, however, i worry those test-runs will take a loong time.

MonroePoteet
06-06-2018, 04:56 PM
Well, as always, I'm no expert but here's an interesting Cumulonimbus made from a single Volumetric null using a pair of procedurals (Puffy Clouds and Turbulence) for the texture and Gradients for the shape. Sample scene attached.

141935 141936

141937

The fly-by was taking about 12 minutes per 640x480 frame before I terminated it at frame 145, and frame 180 took 12:49 on a quad-core (eight thread) Intel i7 2.3Ghz laptop. I didn't modify the textures (moving Turbulence or Puffy Clouds procedurals) during the animation.

141939

The shape of the cumulonimbus is controlled by three Gradients on the X, Y and Z coordinates, the primary vertical shape being a fairly complex gradient on Y. Note that the Gradient alpha output is used as a plain scalar to control the Texture Density and Texture Amplitude. The Volumetric "object" itself is only 1m in diameter, so the gradient(s) range from -1.0 to 1.0 to cover the entire object. The Null is then scaled up to be "life size" (4000x, 2000x, 1000x) in X,Y and Z and moved to 6km from the camera and 2km above "ground".

141938

I thought I could get away with an Envelope on the Step Size to have shorter render times farther away (a la MipMaps), setting it high (e.g. 20m) at the beginning of the fly-by and reducing it down to say 5m as the Camera gets closer, but the higher step size produces a repeating surge-type pattern in the procedurals, so that didn't work.

I also CAN NOT get volumetric scattering to work at that scale for the "glory" or "god rays" type effect.. I'm probably just missing something.

Anyway, perhaps a starting point. I've got an "growing" animation to change the position / scale of the procedurals rendering now - we'll see how far it gets! So far, right at 12 minutes per frame on a 3.4Ghz quad-core i7.

mTp

jwiede
06-06-2018, 05:22 PM
https://www.youtube.com/watch?v=ObNJoJIAx4s

Just saw this, awesome.

It's nice but (all IMO) the clouds01 & clouds02 examples certainly doesn't compare to the breadth and complexity of lighting conditions in the Terragen turntable video. They're both (clouds01 & clouds02) relying almost entirely on diffuse indirect lighting, without any (significant) visibly-cast shadows/godrays, spectral dispersion, etc. Even where the clouds do change form over time, the resultant shading remains quite simplistic -- in the Terragen video, the newly-forming/-formed areas offer some of the most interesting results.

Overall, the Terragen turntable video demonstrates a much broader array of cloud lighting conditions, and performs quite well in dealing with them. Seriously, go back and watch the Terragen video a couple more times, and then go find something equivalent from LW2018. Those examples you posted are ample for showing off LW2018 can do detailed OpenVDB import of volumetric data. As demonstrations of volumetric shading capabilities (esp. w.r.t. clouds), however, neither of those examples was particularly interesting or impressive.

erikals
06-06-2018, 06:26 PM
True, it is certainly a difference between Terragen Cumulus and LightWave Cumulus.

just did some tests by the way, and finally got quite alright Cumulus clouds using LW11.6

they should look even better using LW2018

--- an important note though, even Terragen has problems rendering certain types of clouds.

though, certainly the atmospheric effect of Terragen is quite beautiful.


(i'll have to wait a bit posting the LW result, not 100% there yet)

prometheus
06-07-2018, 04:38 AM
True, it is certainly a difference between Terragen Cumulus and LightWave Cumulus.

just did some tests by the way, and finally got quite alright Cumulus clouds using LW11.6

they should look even better using LW2018

--- an important note though, even Terragen has problems rendering certain types of clouds.

though, certainly the atmospheric effect of Terragen is quite beautiful.


(i'll have to wait a bit posting the LW result, not 100% there yet)

You can get nice one..I am quite convinced of that..dpontīs rman textures and gardner clouds are very nice for that as a starting point, you need to know how to tweak each parameter though, fbm noise and turbulend noise, even dented can alsow work nicely.
Otherwise just do a fluid sim in blender without the cost of turbulenceFD, and you will have a good simulation as well...for rendering directly in blender, you could just use one of the cached files, itīs a bit tricky to simulate and then bake, reload with the right index number for static clouds in blender, that will change in upcoming versions though.

This image..also posted previously...I am not sure what procedural I used, but the new volumetrics and nulls, is blending much better than old hypervoxels, you get softer edges, you get more realistic shading..it does take longer time to set up, and rendertimes as I mentioned before can be an issue...though leitner showed that there shouldnīt be any difference in quality and scale, I do get slower vpr iterations..but have to verify what he says with final renders.

new volumetrics..

http://forums.newtek.com/attachment.php?attachmentid=139902&d=1517614108

Old hvīs...


https://lh3.googleusercontent.com/BV-K2BddAfHGEboVeVo_yOIKwwitPPvTNOdPIWTWjtiIIMxGKjjcI Ngq5Eo-yprNZw9rZO_EaI_DzQ=w460-h220

erikals
06-08-2018, 03:45 AM
thanks, very nice renders  https://i.imgur.com/bcwLfNX.gif

i'm experimenting with particles instead of nulls.
what do you think in regards to nulls vs few particles ?

btw, running 11.6 at the moment over here, and surely the new 2018 renders are better
it would be nice with Gi in 11.6

prometheus
06-08-2018, 03:56 AM
thanks, very nice renders  https://i.imgur.com/bcwLfNX.gif

i'm experimenting with particles instead of nulls.
what do you think in regards to nulls vs few particles ?

btw, running 11.6 at the moment over here, and surely the new 2018 renders are better
it would be nice with Gi in 11.6

Well..Generally I use particles mostly on large cloud areas and moving clouds, but also previously in 2015 for point cluster sculpted or filled object with the fill solid tool in modeler, point clusters because of how bad it was when using several nulls in 2015 with hypervoxels, as soon as they overlap it would get lightwave to a halt ...very slow render and impossible to work with in that way..with 2018 you can now use nulls that way with old hypervoxels..they have changed that so it doesnīt halt and you can mix several hv nulls with the old hypervoxel system, unlike 2015..thatīs a good thing.

Depends on you refer to 2015 or 2018, I do not like how the new volumetrics work on particles, again VPR seem to iterate refresh very slowly, While I can not tell for certain if final render actually is doing a better job, but using VPR and tweaking the new volumetrics on particles isnīt a nice experience for me, to that comes the fact that you can not control random size directly in the way you could with old hypervoxels, you have to add a couple of nodes and connect them in node editor, Or use the particle emitters -+ particle scaling, and that somehow isnīt yielding the same nice random particle scale...and using random particle scale is Crucial for clouds, you do not want each particle voxel volume to be the same size.

To note about blending, the new volumetrics blending between nulls is much better than old hypervoxels blending when using nulls, but the new volumetrics on particles and particle voxel blending isnīt as good, seem almost the same as old hv blending, wich is a bit odd.

Oh..forgot, the volumetrics with old hypervoxels is faster in 2015 than 11.6, and since you are using 11.6, do not use nulls..since you can not have them overlap unless you use lightwave 2018, so it would be simple points, point clouds or particles, you can also use line pen and draw tree structures, or why not dpont tree (in skelegon mode) and either apply hypervoxels on itīs vertices (result depends on the dp tree segments) or use line pen or dp tree at lower segmented points, and add particles to those tree branches..line pen segments, and you could control how many particles is emitted on those line segments by using surface, or vertex mode, or objec line and various frame rate birth.

otherwise the 2018 lightwave and moving around nulls is to be prefered I believe when it comes to cumulus clouds, otherwise using a lot of particles, the volume result with somehow reflect each particle and with lesser good blending, while only a few nulls will let the texture itself do the work of bulging volumes,..but that is only with reference to 2018 since 2015 or previous versions do not have that nice blending between nulls.

erikals
06-08-2018, 10:46 AM
To note about blending, the new volumetrics blending between nulls is much better than old hypervoxels blending when using nulls, but the new volumetrics on particles and particle voxel blending isnīt as good, seem almost the same as old hv blending, wich is a bit odd.
that's indeed a bit odd...  https://i.imgur.com/0MfIYXF.gif

seems i'll have to test more blending modes in 11.6

i can get nice Cumulus effect in 11.6, but not the soft-blend part,
kinda forgot about that challenge

so close, yet so far away...

prometheus
06-09-2018, 04:41 PM
Had some mailing with Bryphi77, and just want everyone to know, since he also wondered about any suggestion to make volumetrics faster in 2015 ..And in special with reference to using hypervoxels clouds, with another null that works as a volumetric fog layer for godrays, to my experience and knowledge, it isnīt worth it, you can cut down by disabling volumetric aliasing, or set a higher value of near clip distance in shading advanced tab so you cut off volumetrics closer to the camera, that may ruin the scene though if you want volumetrics quite close to the camera.

Turning of volumetrics and textured shadows completely for the volume fog item will also help out for faster renders, but as mentioned..using overlapping volumetrics like this in 2015 is a render hog and not worth it.
Using the new volumetric scattering with the new volumetric clouds is much faster in 2018, at least for lower scale, larger real cloud scale and scattering may be trickier too.


http://forums.newtek.com/attachment.php?attachmentid=139817&d=1517168381

I do hope the VDB import may be a bit more stable in next releases, found it to be too crashy currently, when stability is there..It may be good to use houdini cloudfx and use any of the tools there, either fluids, or the cloudfx tool ..maybe on modeled shapes, and then export to VDB for lightwave rendering, you need to add a VDB smoothing node at the end of houdini exports for it to be smooth when importing to lightwave.

Check this older thread for some of that..
http://forums.newtek.com/showthread.php?155880-Houdini-OpenVDB-exports-cloudFX-and-pyroFX&styleid=14


This image was actually taking form of a threesome..
Blender metaballs to form the main cloud shape, imported to houdini as obj, then used houdini cloudfx, then exported to VDB and rendered in lightwave.
Mind you it was just my first test with that workflow and those pipelines, had I interest of really doing it great ..then I would have worked much much more with the main shape in blender.

http://forums.newtek.com/attachment.php?attachmentid=139818&d=1517180501

erikals
06-09-2018, 05:39 PM
a late night reply here...

https://i.imgur.com/aS9u5JU.png


yes, HyperVoxels in 11.6/2015 is just too tricky

2018 at least has the possibility and renders are nicer


metaballs / vdb
quite cool, realism is so-so, but i like the look  https://i.imgur.com/bcwLfNX.gif

---------------

sidenote,
looks like i'm going the Vue Esprit route for Cartoon style clouds

prometheus
06-10-2018, 05:45 AM
a late night reply here...

https://i.imgur.com/aS9u5JU.png


yes, HyperVoxels in 11.6/2015 is just too tricky

2018 at least has the possibility and renders are nicer


quite cool, realism is so-so, but i like the look  https://i.imgur.com/bcwLfNX.gif

---------------

sidenote,
looks like i'm going the Vue Esprit route for Cartoon style clouds

yup..it wasnīt worked for realism, just more a check on how the vdp and shape workflow would work, I jumped in yesterday firing up houdini (which I use too rarely) and did some basic shapes with houdini metaballs, then added cloudfx, added convert vdb nodes and smooth nodes then exporting to lightwave VDB, I need to learn more about the smoothing process since I sometimes still get to blocky volumetric files, depends also on how you set the scattering scale which sometimes pronounce that too much.

The Legacy hypervoxels illumination mode CelShade, that one can be quite nice for clouds, especially cartoony ones, you may need to use some gradients to avoid too much opacity or dark edges though, but otherwise that can be an option in 2015 if you use only particles and/ or points, since those you are applying hvīs as one volumetric instance of the whole point object or particles, as opposed to 2 or more nulls so it wonīt slow down the renders in the same way.

If not using 2018, and if you are desperate..maybe go for investing in turbulenceFD and simulated cloud sections...I think the workflow of doing that is pretty nice with turbulence, and you could use textured based density for the fluid emissions...as well as adding gradients and procedural textures on other channels so cloud formation is raised at various height level ( can not really see how to do the similar with blender smoke)
And if not investing in anything other than time and wrestling with the UI, doing the fluid sim in blender works as well..if you know how to, 2.79.3 has a new principled volume shader so you do not have to set each node up for volumetrics.

erikals
06-10-2018, 05:57 AM
yes, i think i'll end up wit various techniques in the end,
so first Vue for cartoons,
then more experiments with LW2018 once bought an turbulence

the second challenge now is Cirrus clouds, i've got some ideas, but need to test further.

none of the apps (LightWave Vue Terragen) are any good at that...

prometheus
06-10-2018, 07:21 AM
yes, i think i'll end up wit various techniques in the end,
so first Vue for cartoons,
then more experiments with LW2018 once bought an turbulence

the second challenge now is Cirrus clouds, i've got some ideas, but need to test further.

none of the apps (LightWave Vue Terragen) are any good at that...


You have two main functions..
1. the main density (shape)
2. secondary noise, often and mostly distorted by wind turbulence.

The main density you can create by various noise/ procedurals, then you can try and distort with displacements etc, but the most realistic would be to approach it as realistic in simulation as possible..meaning actually doing a fluid simulation with wind turbulence and vorticity perhaps.

Cirrus clouds with long stretched..but still very undulated with turbulence is really sweet to watch when it covers the whole sky.
The new volumetric system in lightwave 2018, has a very nice softness to it, so I think some stuff can be made there, though and exact dynamic turbulence isnīt possible, it would be all distortions of the procedural texture.

Sample images is a very flat plane and just using texture fluid emision for the main shape, then some directional movement and distortions, didnīt really take it to that required level of getting realistic cirrus clouds since I never bought turbulenceFD, might consider doing that Once I get my long standing request implemented someday in the years to come, and when I know what direction the new turbulenceFD will take as well as Lightwave Next..Next :)

I did a little mini tute, not narrated..with how to use sprites on a larger scale and some noise functions, including hypertexture displacement effects, for some type of cirrus styled clouds..not as good as those stretched ones with lots of turbulence..but may be working for some type of soft cirrus style cloud layers..


https://www.youtube.com/watch?v=LxmCg6dPHGc

http://forums.newtek.com/attachment.php?attachmentid=141950&d=1528636709
http://forums.newtek.com/attachment.php?attachmentid=141948&d=1528636602
http://forums.newtek.com/attachment.php?attachmentid=141949&d=1528636639


141948
141949
141950

prometheus
06-10-2018, 08:19 AM
As I mentioned before, I think you can work out cirrus clouds with the new volumetrics very nicely, one favourite procedural is the fbm noise ( is inverted by default weird though, you have to invert alpha to have it as your preview texture otherwise show it)

If carefully set right..this is where the magic happens, you simply got to know what parameters is needed..learn to tweak the values and just a single little value can change the look dramaticly, itīs the same as any other procedural texture on hypervoxels, or displacements.

In the image the final distortion is made by a turbulence procedural fed in to position input of the fbm noise.

Dpontīs weather procedural can also be nice since it has that special dedicated cloud weather function, with various cloud parameters and wind parameters, it is however more complex and tricky to set up, so for now this fbm and turbulence texture will do...no fluid simulation here, just new volumetrics and procedural textures.

The lower image shows just the fbm noise with the turbulence distortion disconnected..not as nice of course.
These was more aimed at the feather type cirrus clouds, there is also the type of smaller puffier round stacked cirrus clouds, but those are actually easier I think.
Below images and techniques can be translated for great nebula clouds in space as well, with a softness And volumetric feel that previously would be hard to acheive.


http://forums.newtek.com/attachment.php?attachmentid=141951&d=1528640185

http://forums.newtek.com/attachment.php?attachmentid=141952&d=1528640203

141951

141952

prometheus
06-10-2018, 09:29 AM
And for godrays also mentioned before, forget trying to work with the double hv volumetric tricks if you use fully volumetrics in clouds (unless using transparent cloud planes, then 2015 and such technique would work fast) but mixing two volumetrics one for the cloud object and with an Hv volume fog item is just toooooooo slow.
The new gobal scattering seems a lot faster ..itīs a bit tricky set the lights and scattering values correct, itīs not by default setup to represent a physicly correct atmosphere, and you would need two lights for the clouds to be illuminated good enough, you could use the environement light, but I prefer additional distant light...

The distant light however doesnīt work with the new global scattering volumetrics, so you can not use sk sunsky motion unfortunately, that one was great if you set itīs position to use manual overide, and the sk suncolor motion modifier used that properly, so it is sort of harder to get color properly on the light as well as the cloud color..which was sort of easier with old hypervoxels.

Sample of godrays at two different Light volumetric intensity, scale of cloud or sun distance is just set at lower scale and they are inaccurate for a true real global scale atmosphere.

We really could use a full spectral physical model of the global volumetric scattering system, and make distant light work properly with that as well, that may improve on realism for cloud color and lighting as well, next step would be implementing infinite procedural cloud planes..that would help shaping things up, the actual amount of procedural textures for clouds is quite good.

for getting volumetric godrays, ergo..we can not use distant light and the sk sun motion modifier which had an option for manual pitch and heading control, we have to use the other lights, in this case a point light and use the sunspot modifier instead, which only offers parameter time and longitude/lattitude controls..itīs a bit harder to set your scene lighting up exactly as you want.

difference in the images is only the light volumetric intensity at 200% and 1000%
Scattering weight affects the amount of glow and godrays as well, and also the color of the scattering where pure white will increase the amount or bloom so to speak.

http://forums.newtek.com/attachment.php?attachmentid=141953&d=1528644428

http://forums.newtek.com/attachment.php?attachmentid=141954&d=1528644444


141953

141954

Marander
06-10-2018, 10:55 AM
yes, i think i'll end up wit various techniques in the end,
so first Vue for cartoons,
then more experiments with LW2018 once bought an turbulence

the second challenge now is Cirrus clouds, i've got some ideas, but need to test further.

none of the apps (LightWave Vue Terragen) are any good at that...

Hi Erikals

I wouldn't reduce Vue to Toons, it was used for many Blockbusters (Terminator, Jupiter Ascending, Avengers etc.). But for hobby use the problem are the long render times on my opinion (that actually applies for most cloud renderings except maybe when using TFD).

prometheus
06-10-2018, 11:11 AM
Hi Erikals

I wouldn't reduce Vue to Toons, it was used for many Blockbusters (Terminator, Jupiter Ascending, Avengers etc.). But for hobby use the problem are the long render times on my opinion (that actually applies for most cloud renderings except maybe when using TFD).

Not sure, but I do not thing the Vue usage in terminator had any cloud renderings really...not sure, but what I read about the vfx usage for that was the instancing system and using pre-made cornuccopia buildings.
for jupiter ascending and avengers, donīt know.
But granted, if you use the right procedural noise, and tweak them good, you can do nice stuff as well...apart from that the main noise function is not as good as lightwave procedurals and terragen noise, vue will only render quite fast with the cloud types of spectral layer 2 ( I think) the newer layers are Sooo slow, I couldnīt find any value in tweaking and render those.

Terragen though slow as any volumetrics, is still quite acceptable I think...though rotating and moving clouds, navigating is a horror I think..do not like the interface at all, and sliders respond poorly unfortunatly when tweaking.
But nothing beats the cloud lighting and shading and the noise fractal terragen has.

Canīt we have just one, like terragen living inside of lightwave...with possible fluid sims, openGL of houdini..and the terrain editing and presets of vue ..oh, back to dreaming, had some cloud sessions in 2018 today with some nice stuff, unfortunately after a while, changing cloud fractals to fbm noise made it crash ( the fbm noise often crashes lightwave) and since I am using discovery edition, no saving and ability to go back, have to set things up all from scratch again almost....thatīs the price I pay of not having upgraded...then again I am not satisfied with it yet so I am not too sorry.

erikals
06-10-2018, 03:23 PM
thanks for sharing,
i think i'll go a dynamics way or something similar for Cirrus, i feel the other ways are just not quite there.

started some manual tests that look alright, however, automating it would cut down the workload x40 times

or mix it with procedurals somehow.

agh! so darn time consuming this cloud stuff...  https://i.imgur.com/pWmMPnl.gif

jwiede
06-10-2018, 05:03 PM
the second challenge now is Cirrus clouds, i've got some ideas, but need to test further.

none of the apps (LightWave Vue Terragen) are any good at that...

Odd, doing a search of images of Cirrus clouds in the Terragen forums (https://www.google.com/search?q=cirrus+cloud+images+from:https://planetside.co.uk/forums/&source=lnms&tbm=isch&sa=X&ved=0ahUKEwi98cfdmMrbAhWmq1QKHZzFByEQ_AUICigB&biw=1378&bih=881) yields quite a few very nice results from users (often including info on how it was done). Definitely seems to strongly challenge your claim that Terragen can't do decent Cirrus clouds.

https://planetside.co.uk/forums/index.php?action=dlattach;topic=5219.0;attach=1376 2;image

This particular thread (https://planetside.co.uk/forums/index.php?topic=5219.0) demonstrates and discusses an interesting method for Cirrus clouds which might even be transferable to other sky sims given available, similar-enough fractals, etc.

prometheus
06-10-2018, 05:15 PM
Odd, doing a search of images of Cirrus clouds in the Terragen forums (https://www.google.com/search?q=cirrus+cloud+images+from:https://planetside.co.uk/forums/&source=lnms&tbm=isch&sa=X&ved=0ahUKEwi98cfdmMrbAhWmq1QKHZzFByEQ_AUICigB&biw=1378&bih=881) yields quite a few very nice results from users (often including info on how it was done). Definitely seems to strongly challenge your claim that Terragen can't do decent Cirrus clouds.

https://planetside.co.uk/forums/index.php?action=dlattach;topic=5219.0;attach=1376 2;image

This particular thread (https://planetside.co.uk/forums/index.php?topic=5219.0) demonstrates and discusses an interesting method for Cirrus clouds which might even be transferable to other sky sims given available, similar-enough fractals, etc.

I think thesample image is too soft however, they need to be more thick and thin at the same time, my sample with the cirrus and lw 2018 fbm noise is like that, though the global cloud density isnīt that good.

compare...right image is the one you posted..unfortunately I am not in the mood to improve mine, since I could not save the scene, and Ivé been doing clouds all day today, ranging from houdini, terragen, blender, lightwave .
As mentioned, the global scale and stretching isnīt good enough on my image, but I think the density and feathering is nice enough, maybe a tad bit softer in various parts maybe.

http://forums.newtek.com/attachment.php?attachmentid=141957&d=1528672720

141957

prometheus
06-10-2018, 05:24 PM
thanks for sharing,
i think i'll go a dynamics way or something similar for Cirrus, i feel the other ways are just not quite there.

started some manual tests that look alright, however, automating it would cut down the workload x40 times

or mix it with procedurals somehow.

agh! so darn time consuming this cloud stuff...  https://i.imgur.com/pWmMPnl.gif

unfortunately Extremely time consuming, I canīt see how any software can compete with terragen when it comes to full cloudscapes with that kind of nice lighting overall, from haze, mulitple scattering in clouds etc, might get time to compile some mood printscreens, donīt have the full version either of terragen though.

jwiede
06-10-2018, 07:22 PM
I've yet to find anything that seriously competes with Terragen 4 when it comes to skies, they did a really amazing job with their new spectral model. Vue latest is close (though still a ways behind), though I never was much of a fan of Vue's UI, frankly. TG4's has issues too, no question, but at least it doesn't suffer from the same extent of "eight-million-sliders-jammed-together" issue common with Vue (and CS & Ozone too).

OTOH (and all IMO, obviously), Vue basically _owns_ environment instancing/scattering, can do everything CarbonScatter can do and a bunch more, with as much control as you could ever want. Even Terragen isn't quite in the same league there. The way Vue allows driving the environmental scattering with terrain generation params and details can't really be done elsewhere, Terragen's probably got the best chance of catching Vue in that regard, or perhaps Houdini, but both still need to put some serious effort in to catch Vue in that regard.

I've been using C4D + CarbonScatter + DEM Earth, and that works "okay to decent" for existing terrain given the right veggie population collections, but I'd still love to see some real competition for Vue in those areas. Perhaps Gaia will make inroads there, who knows.

erikals
06-11-2018, 02:46 AM
Definitely seems to strongly challenge your claim that Terragen can't do decent Cirrus clouds.
https://planetside.co.uk/forums/index.php?action=dlattach;topic=5219.0;attach=1376 2;image

that above is not quite the style i'm looking for...
now, try this >

http://typesofclouds.net/wp-content/uploads/2013/06/url-1.jpg

no app can do Cirrus in a nice way, however others might feel different about it, and that the current Terragen solution works for them.


Terragen cerainly does create nice clouds, and often does a better job than Vue.

for this current project however, i'm personally going for Vue + KronosFX

...as for Cirrus, i'm looking into other techniques

erikals
06-11-2018, 02:55 AM
test One >

https://i.imgur.com/clvMyAZ.png

RPSchmidt
06-11-2018, 07:21 AM
I think thesample image is too soft however, they need to be more thick and thin at the same time, my sample with the cirrus and lw 2018 fbm noise is like that, though the global cloud density isnīt that good.

compare...right image is the one you posted..unfortunately I am not in the mood to improve mine, since I could not save the scene, and Ivé been doing clouds all day today, ranging from houdini, terragen, blender, lightwave .
As mentioned, the global scale and stretching isnīt good enough on my image, but I think the density and feathering is nice enough, maybe a tad bit softer in various parts maybe.

http://forums.newtek.com/attachment.php?attachmentid=141957&d=1528672720

141957

I agree that that version Prometheus produced is more realistic... as he has pointed out, it does have some problems that perhaps an alpha falloff texture could help with, but they have the spread, pile and push density that the Terragen image lacks.

IMO, it's very obvious that the Terragen example is fractal-based; it literally looks like a fractal at the edges where the clouds are closer to the camera.

I think it would be a huge benefit to the community at large if, while you are experimenting, you shared settings that you like as you achieve them, Prometheus. Of course, it's up to you.

prometheus
06-11-2018, 12:18 PM
test One >

https://i.imgur.com/clvMyAZ.png

We need a tiny bit larger image:D
the sweep is interesting, but I think it may be in reverse..at least it may look like that, maybe if you could get that clump more scattered , generally wind is pushing cloud parts in a long long stretch..which is smooth, and when the wind drops in force or take a direction turn, or if it is met with another force or pressure, at the end it will become scattered in to undulated feathering threads.
The initial sweep though is looking very good though.

kadri
06-13-2018, 06:19 AM
By the way this was just posted new in the Terragen forum. Seems OK to post here?
If anyone is interested in VDB cloud export from Terragen.

https://planetside.co.uk/forums/index.php/topic,24626.0/topicseen.html

prometheus
06-13-2018, 06:35 AM
By the way this was just posted new in the Terragen forum. Seems OK to post here?
If anyone is interested in VDB cloud export from Terragen.

https://planetside.co.uk/forums/index.php/topic,24626.0/topicseen.html

Good option to have if that arrives, however...I wonder if it not would be better to have OPEN VDB import in Terragen, so we can do fluid sims and then use the wonderful renderengine of Terragen, also being able to design your clouds more manually which is harder for hero clouds in terragen..either using lightwave procedurals on nulls or particle or point clouds, then again that means OPEN VDB export is needed for Lightwave as well.

kadri
06-13-2018, 06:56 AM
Yeah import to Terragen would be nice. Looks like the next step whenever it arrives.

prometheus
06-17-2018, 09:38 AM
More fast fiddlings, just some more tests without really doing that much effort of making it perfect.
Would there be of any interest for someone who has the full update to compile my settings as close as possible ..then save it and re-upload for everyone here to share and learn from since I only have the discovery edition.

It may of course be of more use for you guys than me ...with reference to that I didnīt get the upgrade...so I am not sure why I would do that though :)
Two different noise fractals to the position slot of the main fbm noise, you can tweak this endlessly for a variety of undulated clouds, many settings yields simular result..and sometimes with a slight difference.
Here it is flow noise and turbulent noise fed in to the positionslot of the fbm noise, the main fbm noise is also very sensitive when tweaking contrast ..that sort of determs if the cloud density should be more spread or look like thin tendrils.
Flownoise procedural from dpont Rman collection textures, well that only works now if you add a nodal color layer and acess textures from there, All the direct nodal textures from Dpont is broken in 2018

I May also record to showcase some settings..maybe tomorrow, I got five weeks of vacation now so itīs about time to start recording again,unless this Swedish record warm pre-summer continues, then I need to be out a bit more, and look at cirrus clouds if any.

https://forums.newtek.com/attachment.php?attachmentid=141974&d=1529249379

https://forums.newtek.com/attachment.php?attachmentid=141975&d=1529249473


141974

141975

erikals
06-17-2018, 04:28 PM
it is definitely on to something.

i believe this would be possible in 2015 too ?

btw, flow noise = a lovely procedural 

particles with motion blur also crossed my mind, however, would take lots of time to get it right
...doing some other test projects now... agh, where is time and iQ 180 when one needs it.

again, it's a very cool look, makes me ponder how it could be tweaked further...

prometheus
06-18-2018, 04:42 AM
it is definitely on to something.

i believe this would be possible in 2015 too ?

btw, flow noise = a lovely procedural 

particles with motion blur also crossed my mind, however, would take lots of time to get it right
...doing some other test projects now... agh, where is time and iQ 180 when one needs it.

again, it's a very cool look, makes me ponder how it could be tweaked further...

Forget about particles and motion blur..itīs just overdoing it, you would anyway have to freeze it and use it as an image, which makes it just as easy to just paint something.

Yes, somewhat possible with the old hvīs, but it may not be that smooth, maybe with sprites as I have used before, but then you miss out a bit on a certain density depth..mixed with that nice thin look. and using
full volumetrics will not be soo smooth I think.
By old hvīs nature..you do not have that box mode, so you have to trick stretch it along with using gradients to get a flat....sort of what I did here in some old ..Old hv test, not cirrus clouds ..but the principles could be used and tweaked to something more of a cirrus cloud look.

https://forums.newtek.com/attachment.php?attachmentid=121398&d=1397273737

prometheus
06-18-2018, 04:53 AM
found an old post I did with 11.6, with something that was a bit closer to some sort of windswept thin clouds.
just a single sprite for that cloud layer...with dp sk sun motion and applying that to the color of the cloud as well, it reacts very nicely and changes color when sun goes down, unfortunately the sk sun motion is broken in 2018 ..that is a big deal for me.

https://forums.newtek.com/attachment.php?attachmentid=128148&d=1430856951

prometheus
06-28-2018, 12:47 PM
Well you can find a lot in google image searching for cloud referencing, but I find a site from this amateur photographer, he seem to love clouds..taken in Colorado, lotīs of nice clouds..if you find it useful for reference or want to use his stuff, notify him and sponsor him if you can somehow.

https://coclouds.com/

And if you are searching with google for what else he got on his site..use this string in google image search..
site:https://coclouds.com/


It was midsommers eve over here a few days ago, had some barbeque with family, went home a bit early around ten at evening, and the sun about to go down, and there was these nice windswept midsommer clouds, and the temperature and breeze in the air was so lovely, so I got a bit inspired to make my own.

I am busy trying to be out in the summer now during vacation...but I really want show how you can tweak some minor things to get this with 2018 for softer whispy clouds, hope to record something soon, some other priorities first.
dpont sk suncolor in the lights color envelope is applied... so you can have either whiter cloud patches if the sun pitch is higher, or if lower you will get pinkish, orange sunset tinted clouds, not that is a bit of faking, unlike a true atmospheric enginge like terragen ..in this images you will get the color tint on the whole cloud patch, while in reality and in special landscape renderers, it is the atmosphere colors it more correctly, and only a sundecay at lower levels while som higher cloud patches may still be more white.

Even if we with 2015 can use sprites as I published above this post, the new volumetrics in 2018 does this much nicer and easier.
And setting the y-scale of the cloud item just enough in height, you can controll if itīs gonna be just very thin threaded feather clouds, or if itīs gonna have a slightly more volume feel to it..at certain areas while still being very thin in others.

https://forums.newtek.com/attachment.php?attachmentid=142039&d=1530211143

https://forums.newtek.com/attachment.php?attachmentid=142040&d=1530211194

https://forums.newtek.com/attachment.php?attachmentid=142041&d=1530211219





142039

142040

142041

erikals
06-28-2018, 01:04 PM
...you are definitely on the right path here prometheus   :king:

prometheus
06-28-2018, 01:10 PM
...you are definitely on the right path here prometheus   :king:

I did this quite fast, but I donīt dare to try and work more on them...donīt have the full upgrade and canīt save them.
So why didnīt I upgrade then? many other things to consider and though it can perform nice stuff like this, there are too many workflow issues I do not like about the new volumetric system, so those were huge parts of why id didnīt upgrade.

Ill try and point those remarks out to the development team..for them to consider, perhaps improve on for a better lightwave, and mabye then ..I may consider upgrade if they do something about it.

erikals
06-28-2018, 01:13 PM
why didnīt I upgrade then?
yup, one can always wait for 2019/2020. think i will.


It was midsommers eve over here a few days ago, had some barbeque with family, went home a bit early around ten at evening, and the sun about to go down, and there was these nice windswept midsommer clouds, and the temperature and breeze in the air was so lovely, so I got a bit inspired to make my own.

got the same ones over here some around the same time   :hey:

prometheus
06-28-2018, 04:42 PM
yup, one can always wait for 2019/2020. think i will.



got the same ones over here some around the same time   :hey:

Donīt know your exact location in Oslo I think it was?
Itīs around 40 miles from stockholm I reckon, not impossible it was the pretty much the same cloud front.
Wish I had the time to go there this vacation..will see.

jwiede
06-28-2018, 04:45 PM
unfortunately the sk sun motion is broken in 2018 ..that is a big deal for me

I've basically relegated DP Sunsky to the "broken in LW2018" pile altogether -- even the parts that "still work in LW2018" appear to significantly compromise overall LW2018 stability here. I really hope Newtek can do something to get all (or even most) of the currently non-working DP plugins working in LW2018.

kopperdrake
06-29-2018, 02:04 AM
Agreed - we desperately need some sort of atmosphere/real light set up, it's intrinsic to so many projects.

prometheus
06-29-2018, 06:35 AM
I've basically relegated DP Sunsky to the "broken in LW2018" pile altogether -- even the parts that "still work in LW2018" appear to significantly compromise overall LW2018 stability here. I really hope Newtek can do something to get all (or even most) of the currently non-working DP plugins working in LW2018.

I agree, itīs a shame and pitty.

for some textures you can use color layer and acess the textures to some degree ..as I have done in these images for feeding transforms in the position slot, thatīs one thing...but if I would like to use Dponts rock textures, there is no node that can properly deform a subdiv mesh as it could be done when they are in node textures, color layer rman rock textures will not have as good output and not the same.

Then you got all the other nodes of course, If possible..Newtek should buy Dpont out..(if he doesnīt have time nor some interest of updating, make sure what the status is..that is the intial step to take) and donīt be cheap about buying him out..I think Newtek will have that investment payed off later, but make sure there are skilled developers who can correct the nodes to 2018 and push them even further.

The Lw community and customers are suffering too much without these plugins, without a full replacement from their own native tools.

It starts with a Good question, like Hi Denis, we are willing to give you this offer...can you consider, and make him an offer he canīt refuse in the best Italian mob way :)
Another alternative would be a crowd funding campaign again to make him update the stuff, but then Denis himself need to verify and make a promise if it reaches a certain amount, he can do it, there is also the possibility
that he may want to, but just dont have time..then he should sell it if he wants to and can do that.

JohnMarchant
06-29-2018, 06:56 AM
I would love an update to OGTaiki and Hikaro as well.

prometheus
06-29-2018, 08:08 AM
I would love an update to OGTaiki and Hikaro as well.

Wouldnīt I , but I think the probability of that is about zero, that plugin has left the building.

Otherwise that is the only one that could perform atmospherics in a realistic way, it was just messy with quality settings, and renderspeed..but at least it had full controll over cloud noise, and you could tweak all settings and get feedback in VPR, Unlike the only other atmospheric plugin ..Ozone, which was too limited.

With all these tools not(in my opinion) working adequate, I had hopes for the lw team to make a new sky generator with atmospherics..and also throw skytracer in the garbage can, as it is now..the options are worse than in 2015 for sky atmospherics I think, the new volumetrics doesnīt seem to include spectral atmospheres or something close to ozone or ogo taiki, unless you go to the extremes with nodes somehow..that is out of my knowledge though.

I do wonder if ogo taiki if rewritten to 64 bit, would benefit from the new render engine in 2018, as being faster...maybe not.
With ogo taiki and some improvements in quality setups, making it easier for various good, medium, high quality results, we could get the full sky and cloudscape, with infinite clouds unlike the per scale Item dependent volumetrics as of now.

Some old ogo test in ligthwave 11.6.

https://lh6.googleusercontent.com/-5ZDd2fp9IxU/UvZ_Jo6ltPI/AAAAAAAACME/e6egeBtKGAI/w1280-h720-no/ogo%2Btaiki%2B5%2B21min%2B21%2Bsec%2Baa%2B3.jpg

https://lh5.googleusercontent.com/-Ht2WEv6Fdno/UvZ_TqYsLyI/AAAAAAAACMk/gwy3Q1u3JcQ/s908/world%2520machine%2520terrain%2520cloud%2520and%25 20shadow000_002.jpg

JohnMarchant
06-29-2018, 08:22 AM
Funny because both of them are still available to download. I do wish there was a 64bit compile. I have sent him an email to see if he would A) Update it himself or B) release the source code so someone else could have a go.

Still the best cloud plugin for LW in my opinion.

Well thats strange, the mail came back with unknown user but his plugins are still available.

prometheus
06-29-2018, 08:25 AM
Funny because both of them are still available to download. I do wish there was a 64bit compile. I have sent him an email to see if he would A) Update it himself or B) release the source code so someone else could have a go.

Still the best cloud plugin for LW in my opinion.

Well thats strange, the mail came back with unknown user but his plugins are still available.

Yes..best cloud plugin for lw, if worked on right...it will outperform the new volumetrics in 2018 for clouds, except maybe for GI illumination on clouds.

Maybe his mail has changed, his site is still active but he hasnīt updated that, dunno.

JohnMarchant
06-29-2018, 08:47 AM
There are a few of them i would love to get the source code for but OGO tops my list.

erikals
06-29-2018, 10:24 AM
did check OGO not long ago, but it doesn't quite stand up to other cloud generators these days. (imo)

prometheus
06-29-2018, 10:59 AM
did check OGO not long ago, but it doesn't quite stand up to other cloud generators these days. (imo)

Well..that is only with certain aspects in my point of view, the actual procedurals you can retreive from Lightwave itself, is better than vue procedural functions for clouds, with that in mind and if
the rest of the sky is set right, I think it outperforms vue in cloud noise realism, if vue would have the same type of cloud functions as terragen, I think I would have purchased it already, the functions and many presets in there for vue is in my opinion quite bad compared, I have many many...Many hours of tweaking all kinds of noise in vue, but could never be satisfied with it to the same degree I can with lightwave noise, or terragen.

Then we got the spectral thing and fog, vue and terragen does it better .and easier and I guess more physical correct, but you can reach good atmospherics with ogo as well, it takes you longer to get there though.

Then it is about quality...you can adjust so many settings...just too many in ogo, but if set right the quality output may be able to match vue for instance.

Not sure where you think it doesnīt stand up to the other cloud generators? and I would with that refer to sky and cloud generators, not only cloud generators...it needs atmosphere so we canīt just compare to simple volumetric per items, or a fluid cloud..it needs the full spectre of volumetric atmosphere and full global infinite clouds.

prometheus
07-06-2018, 07:15 AM
Sharing some old 2007 cloud photos..not much but a few, use it as you like under condition that you give me some credit for the origininal photos.
The camera wasnīt the best, canon powershot 3, and I wasnīt really cloud hunting that day, but there was some nice stuff every other day.




Album..
https://forums.newtek.com/attachment.php?attachmentid=142109&d=1530882146

Use the link to google photos for larger images..
https://get.google.com/albumarchive/100944643113557837045/album/AF1QipO2WobjRcGgcuWFL90SFwSuSJ9Dh36_HVJlGFCB

Sample...

https://lh3.googleusercontent.com/-rR8Q9GS7gYI/Wz9q4hVddBI/AAAAAAAAC74/2626YrFaz3YDhKYuqauLIor5ccWZYnWCwCJoC/w1600-h1200-n/IMG_0189.JPG

erikals
07-06-2018, 10:15 AM
some really nice ones there...   :)

prometheus
07-06-2018, 06:07 PM
some really nice ones there...   :)

Thanks erikals..


I did a recording on the soft whispy stuff going on, though I was sloppy and after re-installing new drivers for nvidia, the settings might have changed and I happened to record in the wrong resolution, the quality was a bit of as well, and I didnīt plan how to narrate it, which I should do more properly when explaining stuff in the future.

Just a mini showcase..not a full tute, but I hope it may be of use..


https://www.youtube.com/watch?v=DSZGztj3WEc

erikals
07-06-2018, 06:48 PM
honestly,... that is a great tutorial!  https://i.imgur.com/tJGL61i.png

i'll have to look more closely at this at a later time... :king:

busy days over here, but thank you very much for sharing this.  :)

prometheus
07-06-2018, 07:33 PM
honestly,... that is a great tutorial!  https://i.imgur.com/tJGL61i.png

i'll have to look more closely at this at a later time... :king:

busy days over here, but thank you very much for sharing this.  :)

and check the thread on the disney asset as well, not sure I will be able to record something on that as well, itīs a bit slow to render, maybe...between the breaks in the world soccer championships tomorrow between Sweden and England.
Ivé been tweaking the cloud vdb a bit, but there is still a lot to do to get close to the hyperion render, I also need to be more familiar with all the scattering options, how they balance each other, it would probably take a year or so until I reach the level of knowing it very well, at that time we should of course have a new Lightwave out with more goodies, that is a must right :) and I will open my purse again.

erikals
07-06-2018, 08:27 PM
yeah, it does look to be a LW2018 challenge to get that "cumulus" look right.

the clouds just don't illuminate that well. they kinda look like out of paper in away.
it might (unfortunately) very well be a LW2018 limit...


the world soccer championships tomorrow between Sweden and England
must see!  :D :)

prometheus
07-06-2018, 08:35 PM
yeah, it does look to be a LW2018 challenge to get that "cumulus" look right.

the clouds just don't illuminate that well. they kinda look like out of paper in away.
it might (unfortunately) very well be a LW2018 limit...


must see!  :D :)

I think it may have to do with multiple scattering, where these images are done without radiosity..while I have tried with GI, I do not experience any improvements withing cloud scattering really..but I may not be that familiar with it all yet.

and you may be able to use the scattering data, but itīs really difficult to tweak right I think, one small value wrong in gradients, or color or scale intensity, and it goes from almost realistic to cartoony.
Could also be that the distant light isnīt suitable for it, or simply adjust the size angle abit to spread the light a bit more before hitting the clouds.

jwiede
07-08-2018, 10:56 AM
Here are some more RW alto-/cirrostratus cloud images...

142147

142146

142148

prometheus
07-08-2018, 01:11 PM
Here are some more RW alto-/cirrostratus cloud images...

142147

142146

142148

Thanks...always appreciated with reference photos, I got some nice feather cirrus as well, but I have been disorganized, when transfering photos, Have to locate it first and structure it up.

After giving up on the new volumetric clouds at larger scale ..and trying to get the global scattering to work decently..I just had to
to go back to Lightwave 2015 try ogo taiki some more, may be able to record in the end of the week, if I am satisfied with both quality and speed, I can in fact get faster godrays going on than in terragen, since the vpr at a certain size which is bigger than terragen or vue preview, will render draft much faster.
Need to work on quality a bit more, but the previews and general atmospheric with nice light shafts is really sweet..not something that the new volumetrics can come close to.

Marander
07-08-2018, 01:37 PM
Here's a cloud with god rays I took with my mobile this afternoon around the house here in Switzerland - might be a challenge to render ;-)

142149

142150

142151

142152

prometheus
07-08-2018, 02:03 PM
Here's a cloud with god rays I took with my mobile this afternoon around the house here in Switzerland - might be a challenge to render ;-)

142149

142150

142151

142152


Those are great Marander, Thanks..
And yes, itīs a challenge for any software to match that, no problems really with godrays, you can do that in vue, terragen...use the new volumetrics may work to some degree since this is snapshots on a small part of a cloud, and I have posted an image to show that you can do at least a hero cloud then using a spherical light to backlit it and put it close to the cloud and then enable global scattering.

The problem for any software is the actual cloud, especially getting the realistic density and scattering in there.
As I mentioned, you may be able to get somewhat close with the new volumetrics, but I think that only pertains these kind of focused hero cloud shots...and only working on smaller scales, when you make your clouds really large..close to real world scale, the global scattering seem to be very hard to produce something at all..despite how much you crank the values up..at least very hard to do.

If I can get ogo taiki cloud density to reach in my mind good quality enough, then itīs a no brainer with the rest of the atmosphere and godrays really, itīs just about the cloud illumination and quality it all depends on for that..and what ever final rendertime it could impose of course.

jwiede
07-08-2018, 02:22 PM
Is it just me, or does LW2018's global scattering (medium) require significantly "denser" values to yield visible godrays than you'd expect (for real scale)? IRL, at real scales, even minuscule densities of airborne particles create very visible crepuscular lighting effects.

prometheus
07-08-2018, 03:49 PM
Is it just me, or does LW2018's global scattering (medium) require significantly "denser" values to yield visible godrays than you'd expect (for real scale)? IRL, at real scales, even minuscule densities of airborne particles create very visible crepuscular lighting effects.

not just you, It seems to be reported from others as well, and I just canīt get it going either with real scales of clouds.
Besides that, you can not use the most common sunlight, the distant light...it doesnīt work with the new volumetrics as I am aware of, we need a proper light for the sun that works with global scattering.

Then I really have my doubts about the global volumetric scattering being created or adapted as to be working in a full real world scale, that is why I turned to Ogo Taiki again.
Unless there is a way in nodes to extend the spherical lightīs volumetric falloff to work at larger more real world scales?

No problems with smaller meter hero clouds, and if you put the light very close to the cloud..behind it, that is however not very physical correct.
linking to this image again with the new volumetrics and spherical light for the godray effect.

https://forums.newtek.com/attachment.php?attachmentid=139902&d=1517614108

Thomas Leitner
07-11-2018, 04:45 AM
Because we mainly do VFX we rarely have to render clouds.
Although we do sky replacement, but therefor mainly photos are used.
Where we need 3D clouds are shots with a flight through the clouds or Heroclouds with special shapes.

Just a fun test for vdb import in LW 2018:

142182

142183

A few things about clouds I noticed in LW 2018:
OpenVDB:
I would like an easy way to remap the channels.
We need a way to time the VDB sequences as needed (envelopes), especially for explosions.
Asymmetry input seems to be broken (nodal does not work nor envelopes and texture)
A better documentation about the useage of the openVDB node (I dont understand it...).
OpenGL preview with adjustable quallity settings for faster preview.

A test with a vdb from terragen, with distant light at real world scale and dp sunsky:

142184

Volumetric:
We need special procedural textures for clouds. Some cloud types with simple, "real world" parameters and animated motion/evolution.
We want to use geometry to shape clouds.
Real OpenGL preview with adjustable quallity settings for faster preview.

ciao
Thomas

prometheus
07-11-2018, 05:48 AM
Because we mainly do VFX we rarely have to render clouds.
Although we do sky replacement, but therefor mainly photos are used.
Where we need 3D clouds are shots with a flight through the clouds or Heroclouds with special shapes.

Just a fun test for vdb import in LW 2018:

142182

142183

A few things about clouds I noticed in LW 2018:
OpenVDB:
I would like an easy way to remap the channels.
We need a way to time the VDB sequences as needed (envelopes), especially for explosions.
Asymmetry input seems to be broken (nodal does not work nor envelopes and texture)
A better documentation about the useage of the openVDB node (I dont understand it...).
OpenGL preview with adjustable quallity settings for faster preview.

A test with a vdb from terragen, with distant light at real world scale and dp sunsky:

142184

Volumetric:
We need special procedural textures for clouds. Some cloud types with simple, "real world" parameters and animated motion/evolution.
We want to use geometry to shape clouds.
Real OpenGL preview with adjustable quallity settings for faster preview.

ciao
Thomas

Youre a pro Thomas, so of course, that workflow of using backdrop images is effective and to use volumetric clouds for fly throughs or special Hero clouds as you say.... is the way to go.

My first priority "Fixes of the new volumetrics" are...

1. Random scale parameter slider directly in the interface, I do not like to have to use the particle emitter -+ settings for size and switch tabs all the time, neither do I like to set it up through nodes...Hypervoxels 3.0 Was..Is superior in that regard.

2. Make it work directly on points, we should have to do unecessarey work with adding pfx and set that up.

3. Bring back the direct acess to some main textures and make it connected when entering nodes for editing and changing.

4. Master slave settings...as I think you also wanted? a way to set One volume null as master and the others as slaves, so changing emission and scattering etc..will affect the others the same, dynamite plugin had that option.

5. I had hoped for a similar UI where I just could check either hypervoxels 3.0 or New volumetrics 4.0, cause I really find it annoying not being able to uncheck various items, make them active or not and acessing them from there, I find it more
easier to go wrong when you switch between several null volume items, apart from that ..It is Required that you select a volume item to change itīs properties, meaning you can not move other items around at the same time as we could with hypervoxels 3.0

6. Bring back some texture animated effects such as velocity translate.

7. As you say...we want modelled or sculpted geometry to work with volumetrics, not just cone, spherical or cubes, both Modo and Blender can use any modelled shape, that said...Lightwaves volumetrics with textures and softness I think may be better than both of the others though, Itīs just that it is harder to do special shape designed clouds.

8. Yes..new procedurals perhaps, my favourites is Turbulent noise, Rman gardner clouds, Rman Weather, and dented, unfortunately Dpontīs rman textures is broken as nodes and only available as textures and appliance through the color layer nodes, Itīs not the same and you would loose some nodal connection transforms and inputs.

(oh...I have a hellish nightmare when changing textures in the layering system..itīs been like that from 11.6 and up to 2018, where changing a texture with vpr active just crashes Lightwave very very much, depends on certain textures, may be something with dpontīs textures..and some others, also depending on if I have used a texture as displacement layer )


I forgot...

9. Full global infinite cloud layer system.
10. Full Global spectral volumetric layer.

I agree on open GL enhancements.

What...is there openVDB export now in terragen? and is that also in the free demo version?

Positives.. new volumetrics is much more realistic ..at least for clouds, they have made improvements in the sense that the falloff settings actually works now..which it never did with hypervoxels 3., the softness of the edges are much much better than hypervoxels 3.0, we could use local density gradients in various channels to soften ol Hvīs out, but it also affected the general volume quite a bit.


And for erikals....Feel free to use and photoshoping, I do my fixings in Krita nowadays, I forgot to crop away for you only the cloud image, but you can fix that yourself.
see the images, One image smooth repaired with a single filter, the other one is direct raw render..but print screened, added some Light info node to the emission input slot, put on some GI and increased rays, increased volumetric intensity for the two lights, and used environment light and distant light, the distant light I cranked up the angle size a bit.

For avoiding too much detail in original render, It seems you either have to lower the texture density and raise scattering scale, or do the reverse..these two are intertwined that needs a good balance.
At the upper part of the cloud, there is small breakaways and puffs, in real world those are rarely there, at least they should be very very thin and dissolved almost, I could have tweaked the actual cloud parameters more ..but time you know, that kind of unrealistic puffs is something that Terragen handles so well, and where Vue is simply horrible to handle that.




https://forums.newtek.com/attachment.php?attachmentid=142186&d=1531309439


https://forums.newtek.com/attachment.php?attachmentid=142187&d=1531309458



142186

142187

erikals
07-11-2018, 07:41 AM
what really makes the effect in PhotoShop is making a 2nd layer with "screen" (+ a mask)
the other stuff has a more subtle effect.

i think it is also important to remember that Disney studio also 100% likely, tweak it in post.


https://i.imgur.com/DhKg3QF.png

prometheus
07-11-2018, 03:22 PM
what really makes the effect in PhotoShop is making a 2nd layer with "screen" (+ a mask)
the other stuff has a more subtle effect.

i think it is also important to remember that Disney studio also 100% likely, tweak it in post.


https://i.imgur.com/DhKg3QF.png

Perhaps, but I am not that convinced as you are about post tweaks, perhaps some smaller color and contrast, but getting those kind of scattering results and absorbtion, I do think that is the result of how it deals with
multiple scattering within a volume, specially designed to produce that result, and I do not suspect that is the result of any post processing really.

Terragen can yield similar shading on the clouds.

I did a little guide on some settings of the multiple scattering in Terragen, the cloud is an asset from Pablo Del Molino for Terragen anyone can use for free, download it here...
https://gumroad.com/l/dZvdb

I did some minor cloud coverage and altitude changes, and changed the sun heading and elevation.
The rest was just a process of testing multiple scattering from 0-6, and also one color curve adjusted image, and finally the last image I decreased the Param C, which gives a more silverlining to the cloud.
Watch in higher resolution.


https://forums.newtek.com/attachment.php?attachmentid=142204&d=1531343972
142204

erikals
07-11-2018, 05:37 PM
Perhaps, but I am not that convinced as you are about post tweaks, perhaps some smaller color and contrast, but getting those kind of scattering results and absorbtion, I do think that is the result of how it deals with
multiple scattering within a volume, specially designed to produce that result, and I do not suspect that is the result of any post processing really.
oh yes, Absolutely :)
don't get me wrong here, just saying they use it to tweak further :)


Terragen can yield similar shading on the clouds.
very true, NewTek/LightWave should aim at / look towards the Terragen shader.

Gungho3D
07-11-2018, 10:57 PM
...
Terragen can yield similar shading on the clouds.

I did a little guide on some settings of the multiple scattering in Terragen, the cloud is an asset from Pablo Del Molino for Terragen anyone can use for free, download it here...
https://gumroad.com/l/dZvdb


https://forums.newtek.com/attachment.php?attachmentid=142204&d=1531343972


Nice work. Thanks for sharing the cloud freebie model link, and the images work-up, in addition to being informative, looks great.

Thomas Leitner
07-12-2018, 04:11 AM
....2. Make it work directly on points, we should have to do unecessarey work with adding pfx and set that up...

Yes of course , esspecially as we had this since years in the old HyperVoxels.


....What...is there openVDB export now in terragen? and is that also in the free demo version?...

The OpenVDB is from this site:

https://planetside.co.uk/forums/index.php/topic,24626.0.html?PHPSESSID=dqnkfcj8otbevkvv6sa5q tsvi6

It is for testing the VDB workflow in terragen.
Here an other one with this vdb:

142206

ciao
Thomas

prometheus
07-12-2018, 07:42 AM
Yes of course , esspecially as we had this since years in the old HyperVoxels.



The OpenVDB is from this site:

https://planetside.co.uk/forums/index.php/topic,24626.0.html?PHPSESSID=dqnkfcj8otbevkvv6sa5q tsvi6

It is for testing the VDB workflow in terragen.
Here an other one with this vdb:

142206

ciao
Thomas

A question Thomas, I see you used distant light here, I thought it was broken...I have never gotten distant light to work in 2018 with global scattering, have you added some nodal correction for that light or may it have been fixed in
the more recent 2018 versions?
I am still using 2018.01 here.

I see you have too much glow intensity in general sun flare, a tip would be to add another distant light at the same angle and location, make it a large angle size, that way you will illuminate the bottom side of the clouds so they arenīt so dark...thatīs what Ivé been doing to avoid to dark clouds when backlit and doing godrays, you may also be able to use environment light perhaps.

And while it is cool with VDB output from terragen, it sort of looses a lot of what makes terragen render looking nice, the only thing you are left with is some cloud layer, and you can actually do cloud layers with
decent fractals in lightwave as well, but the shading you loose, as well as the sky.

I hope for VDB the other way around, to import fluid simulations to Terragen, and we may see even more fabulous cloud renderings with a new way to create the clouds (hero clouds especially) together with the rest of the goodies in
Terragen.

Thomas Leitner
07-12-2018, 08:18 AM
A question Thomas, I see you used distant light here, I thought it was broken...I have never gotten distant light to work in 2018 with global scattering, have you added some nodal correction for that light or may it have been fixed in
the more recent 2018 versions?
I am still using 2018.01 here......

Distant and Environment light still dont work with volumetric scattering. I use a volumetric primitive (cube) as medium for light scattering.


....And while it is cool with VDB output from terragen, it sort of looses a lot of what makes terragen render looking nice, the only thing you are left with is some cloud layer, and you can actually do cloud layers with
decent fractals in lightwave as well, but the shading you loose, as well as the sky....

Yes, I would also render in terragen. It's just a vdb file I found on the internet to test in LW.


...I see you have too much glow intensity in general sun flare, a tip would be to add another distant light at the same angle and location, make it a large angle size, that way you will illuminate the bottom side of the clouds so they arenīt so dark...thatīs what Ivé been doing to avoid to dark clouds when backlit and doing godrays, you may also be able to use environment light perhaps...

I think you can improve it much. I did playing around with the vdb...did not come to an end.
In other renderings I used an environment light too (like in the one I posted before).

ciao
Thomas

prometheus
07-12-2018, 08:46 AM
More knowledge for the Lightwave team to dig in to, regarding Disney research on AI neural algorithms, aiming to avoid long rendertimes when heavy mulitple scattering may otherwise be required.

It explains that typical darkness shadows accouring as we can see in the cloud asset image from disney..
"We synthesize multi-scattered illumination in clouds using deep radiance-predicting neural networks (RPNN). We combine Monte Carlo integration with data-driven radiance predictions, accurately reproducing edge-darkening effects (left), silverlining (right), and the whiteness of the inner part of the cloud. "


http://drz.disneyresearch.com/~jnovak/publications/DeepScattering/DeepScattering_teaser.png

The paper "Deep Scattering: Rendering Atmospheric Clouds with Radiance-Predicting Neural Networks"
http://drz.disneyresearch.com/~jnovak/publications/DeepScattering/
Pdf...
http://drz.disneyresearch.com/~jnovak/publications/DeepScattering/DeepScattering.pdf


also look at the video which explains in a very easy way what it is dealing with..




https://www.youtube.com/watch?v=7wt-9fjPDjQ

prometheus
07-12-2018, 08:52 AM
Distant and Environment light still dont work with volumetric scattering. I use a volumetric primitive (cube) as medium for light scattering.



Yes, I would also render in terragen. It's just a vdb file I found on the internet to test in LW.



I think you can improve it much. I did playing around with the vdb...did not come to an end.
In other renderings I used an environment light too (like in the one I posted before).

ciao
Thomas

Ahh...I was fooled there, Yes...that trick Is something I used too, though I think it will become slower than actually just using the new global scattering system if you instead of distant light use a spherical light which works.
Environment light I also know doesnīt work within volumetric scattering, just mentioned it for illuminating the clouds beneath, but as I also said..it may be better with a spherical light or some other light with large angle size.
In principle the trick is similar to this vid..but since you can not mix the two legacy system, you have to add the new volumetrics in a large enough size to cover the scene.


https://www.youtube.com/watch?v=DbkvT_HgKRs

When it comes to terragen 4, I like the new lighting in the new easy clouds, but I miss density fractal control changes as with the older ones, and tweaking the new clouds in rtp realtime preview can sometimes become to slow.
The older cloud models can look nice too..and is significantly faster to tweak with rtp preview, I may record some demonstration of that later this week.

Installation of the latest vue build is on the way this week too, for test driving of partly the new metaclouds.

prometheus
07-12-2018, 03:23 PM
Some experiments on faking that darkness, by feeding the same texture in to main texture background color which then is fed in to emission scale, there are probably better ways to acheive this or set up nodes with gradients and other slots, but this was an easy first try test.

various stages where the first top image has the node disconnected ..where the others below is connected, with various contrast settings and also various multiply values before it is fed in to the emission scale.


There is a bit of falloff errors in the cloud, nevermind that..can be fixed if I want to.
This isnīt the ideal cloud to try and create that darkness in the volume, it should be used on larger cumulus clouds..not this midlevel cloud types really.

I do not have acess to any textures in the disney cloud asset, so I would have to try other ways there..if I got the time, I could perhaps try and set up a more cumulus styled clouds with lightwaveīs internal fractals and use the above techniques to see how it looks.

I would like to record this and put it on the tube if I got the time, some really interesting things going on if you invert that second textures alpha, and adjust multiplier.

https://forums.newtek.com/attachment.php?attachmentid=142224&d=1531430361

https://forums.newtek.com/attachment.php?attachmentid=142225&d=1531430399



142224

142225

jwiede
07-12-2018, 04:18 PM
Yeah, I see what you mean about it not really being the right kind of cloud for the effect, a big cumulonimbus hero cloud would definitely serve as a better test for that effect.

Maybe try exporting the Pablo Del Molino cloud asset from Terragen as FBX, then use DeepFX's VDBSplicer (https://www.deepfxworld.com/vdb-splicer.php) to convert it into VDB for LW volumetrics? It's commercial, but for the work you're doing, it seems like a highly useful addition (w.r.t. converting meshes to VDB).

prometheus
07-13-2018, 05:15 AM
Yeah, I see what you mean about it not really being the right kind of cloud for the effect, a big cumulonimbus hero cloud would definitely serve as a better test for that effect.

Maybe try exporting the Pablo Del Molino cloud asset from Terragen as FBX, then use DeepFX's VDBSplicer (https://www.deepfxworld.com/vdb-splicer.php) to convert it into VDB for LW volumetrics? It's commercial, but for the work you're doing, it seems like a highly useful addition (w.r.t. converting meshes to VDB).

Itīs all hobbywork for now, and I wonīt invest in any commercial plugin at all until I see what the next Feature version brings for lightwave, I would also need to upgrade from Lightwave 2015, I would also need to buy Terragen full.
Currently neither Terragen or Lightwave satisfies me enough, where I think both software still Lacks to much for my liking, thereīs always a fear that I have to high demands and the software may not be able to provide what I require..and I would keep on just working with demoīs and not getting anywhere to boost personal projects.

So VDB slicer, well yeah..I noticed that ..but it is put on a lower section on my want list.
Itīs a bit sick that I invest so much time working with these experiments..considering the above mentioned excuses, but itīs a little passion for me trying to acheive nice landscapes.

I may need a better plan on wether or not to continue with this, as personal projects..or take it to another level, or reduce the amount of time I work on it, or set up a good plan on how to elevate it all to another level.
Currently I am on vacation one week more, and I am also planning to move away from my current job which doesnīt include any 3d work and too much of administration work for a too low salery, so I should be workin more on
finding something else.

prometheus
07-17-2018, 07:42 AM
Puh..this very hot summer balances the lousy summer previous year, it all started in early may with record temps at 28 + celsius ..and through almost all may, a slight drop during midsummer eve, but it has been boiling up ever since again, and now we have constant temps from 25-32 degreeīs around the country here in sweden, hardly any rain at all..except for some small local areas with thunder.
Yesterday around 31 + celsius and the same today...theres a around 40 forrest fires around the country..
and a drought for the farmers so the Government have to assign extra money to them to ensure food for the cattle.

I was out at sea in the coastline archipilago of stockholm ...and there was some nice clouds piling up.
I will post some photos of some nice clouds during that boat trip, though only in video hd size since I used a sony handycam and not a better camera, but it may serve as reference.

I Was taking a stab at the volumetrics again,(see below image) a sample draft vpr render here, no post processing.
I think I may be able to get the illumination and shading a bit better now, and some thing I may discard in the future is the environment light, it takes to long to render if you want good quality...and the environment light does nothing for scattering insided a cloud volume, for that we need GI, and then I see no point in using additional environment light.

To get the GI to work with volumetrics, you need to check affected by volumetrics in the GI tab, furthermore..in the volumetrics tab you need to enabable indirect sampling..otherwise the GI wonīt affect anything in the cloud.
I also raised the GI intensity slightly, though I think we in the future would need a seperation where you can set GI intensity for volumetrics and for the rest of the scene seperately, not sure that individual light volumetric intensity is the same?


https://forums.newtek.com/attachment.php?attachmentid=142276&d=1531834854

I also used a new nodal setup for some of the channels, may use that more often if I know exactly what I am doing :)

142278

142276

erikals
07-17-2018, 08:17 AM
what i really don't like, is that the render result looks like it is using a shader from 1990

only NewTek can fix this.

the Subsurface effect looks like paper.

so, someone needs to pick up that phone and call NewTek...


:phone_cal

prometheus
07-17-2018, 08:37 AM
what i really don't like, is that the render result looks like it is using a shader from 1990

only NewTek can fix this.

the Subsurface effect looks like paper.

so, someone needs to pick up that phone and call NewTek...


:phone_cal

I actually think this one looks far better than some others I posted, and a bit more accurate in the shading, can smooth it out..or change the multiplier settings for absorbtion etc.
As for terms..I wouldnīt speak of or refere to subsurface effect, itīs volumetric scattering.

Not exactly sure of the specific area you say looks like paper, to hard shadows..or to flat..not mulitple light scattering within the volume?
It may be that I exaggerated the darkness and sharpness within the volume just to see the more pronounced effect, I am sure I can smooth it out..but getting that nice illumination is a challenge.
But I do agree that it doesnīt seem to be capable of producing multiple scattering with true algoritms in the way Terragen Or hyperion does.


And for reference, a cloud image from my boat tour...not the best quality, extracted from sony handycam video.

https://forums.newtek.com/attachment.php?attachmentid=142281&d=1531838105

142281

erikals
07-17-2018, 08:51 AM
...I do agree that it doesn't seem to be capable of producing multiple scattering with true algoritms in the way Terragen or Hyperion does
yes, pretty much this NewTek needs to fix.


:phone_cal

Thomas Leitner
07-19-2018, 05:26 AM
...I Was taking a stab at the volumetrics again,(see below image) a sample draft vpr render here, no post processing....

I really like the shape of your clouds. I never get them so well (that's why I want better cloud - textures).



...I think I may be able to get the illumination and shading a bit better now, and some thing I may discard in the future is the environment light, it takes to long to render if you want good quality...and the environment light does nothing for scattering insided a cloud volume, for that we need GI, and then I see no point in using additional environment light...

Unfortunately, the manual is not very informative about this. My tests seem to show that (unlike shading of normal surfaces) GI with sampled Backdrop does not distribute any light to direct shading of volumetrics. Only indirect shading gets lit by GI (scattering). On the other hand enviroment light with sampled backdrop contribute to both, direct and indirect shading. But indirect shading looks complete different to indirect shading of GI (and I don't like it very much).

A cloud lit with Backdrop (Sunsky) only, to compare enviroment light and GI:

142305

Also LW 2018 volumetric scattering seem to use only one bounce, regardless of what you set in Render Properties. From volumetric rendering tests with TurbulenceFD and Octan I saw that more scattering bounces leads to brighter lit clouds (so I also tend to use higher GI Intensity).
An other way to fake this would be to use Emission to lit the clouds. Therefor you have more control over the effect (but also more settings to get lost in).

ciao
Thomas

prometheus
07-19-2018, 07:39 AM
I really like the shape of your clouds. I never get them so well (that's why I want better cloud - textures).




Unfortunately, the manual is not very informative about this. My tests seem to show that (unlike shading of normal surfaces) GI with sampled Backdrop does not distribute any light to direct shading of volumetrics. Only indirect shading gets lit by GI (scattering). On the other hand enviroment light with sampled backdrop contribute to both, direct and indirect shading. But indirect shading looks complete different to indirect shading of GI (and I don't like it very much).

A cloud lit with Backdrop (Sunsky) only, to compare enviroment light and GI:

142305

Also LW 2018 volumetric scattering seem to use only one bounce, regardless of what you set in Render Properties. From volumetric rendering tests with TurbulenceFD and Octan I saw that more scattering bounces leads to brighter lit clouds (so I also tend to use higher GI Intensity).
An other way to fake this would be to use Emission to lit the clouds. Therefor you have more control over the effect (but also more settings to get lost in).

ciao
Thomas


I really like the shape of your clouds. I never get them so well (that's why I want better cloud - textures).




Unfortunately, the manual is not very informative about this. My tests seem to show that (unlike shading of normal surfaces) GI with sampled Backdrop does not distribute any light to direct shading of volumetrics. Only indirect shading gets lit by GI (scattering). On the other hand enviroment light with sampled backdrop contribute to both, direct and indirect shading. But indirect shading looks complete different to indirect shading of GI (and I don't like it very much).

A cloud lit with Backdrop (Sunsky) only, to compare enviroment light and GI:

142305

Also LW 2018 volumetric scattering seem to use only one bounce, regardless of what you set in Render Properties. From volumetric rendering tests with TurbulenceFD and Octan I saw that more scattering bounces leads to brighter lit clouds (so I also tend to use higher GI Intensity).
An other way to fake this would be to use Emission to lit the clouds. Therefor you have more control over the effect (but also more settings to get lost in).

ciao
Thomas

Thanks Thomas,

The cloud I did was actually based on my boat tour, straight from memory..I could have used that cloud video image I took and also posted here, and I would have landed even closer to a real cloud, 4 nulls strategicly placed
to form a lift up and drift to the right.
In lightwave...shaping up cloud formations with nulls is actually easier to get this kind of raised clouds than doing it manually in Terragen..( the controlls sucks)
If we only had the same illumination..and a spectral sky to go with it.

Yes..we could use new special designed cloud fractals, the old sky tracer designed clouds is out of date and not working properly in the new volumetrics,
gardner clouds and weather was quite nice within hypervoxels..but is harder to use in the new volumetrics,
a part of that may be because the rman nodes is broken and you can only use the rman gardner clouds or weather within the additional color layer node,
so it isnīt wrapping properly to the volume.
For this cloud I used 4 nulls, and the turbulent noise..you have to move around and tweak a bit to get something that looks good, the noise governs both main shape form and smaller noise.

For houdini in some cases.. itīs easier to model a cloud shape, and add minor noise and advection noise, a different approach where it is easier to predict and design the main shape of the cloud faster, but there are cons such as being stuck with that form, passing it on the real-time opengl may take a couple of seconds for it to cook depending on resolution, once there you can navigate with camera and lighting in realtime with great opengl preview, but a con is that it takes longer to cook and send to realtime preview. than what it takes for lightwave VPR iterations for any tweak you do to the actual cloud noise.


I made a complete guide for this particular cloud with all the settings, with everything from null positions, scale and rotation, camera setting and ligth setting, volumetric settings, and texture settings..so I can re-create it when the next
feature update comes..and I may be able to save that scene and perhaps share, it was good that I did that..since 2018 actually just crashed and I was about to record some notes with that cloud, will have to recreated it in such case with the help of that guide.


Thanks for the feedback on the scattering with environment light and only gi lighting, didnīt now that about the environment light.
I thought the GI brute force settings with more rays did increase the scattering, but I am not sure..will have to check again.

As you say about octane..and what I see when I raise volumetric scattering bounces in blender for smoke..it definitely shows a good scattering illuminating more of the volume, I think both octane and blender would need a way of doing algorithms to scatter according to some true physics inside a cloud volume...as researched for the Hyperion render and Terragen.

Foremost there need to be a fix for the scattering bounces in 2018 volumetrics...if we havenīt missed something here?

Yes..I did some nodal enhancements for some of the channels, but for the scattering scale..you could also use emission, but itīs alot of experimenting without actually knowing how it works truly...and as a result you may end up with more fake looking paper looking cloud as erikals mentioned, but you may fake increase the look of the lighting and shadows that way.

Thanks again for your contribution and research of this Thomas, next would be to forward the issue to (Janti?) wether or not multiply illumination bounces actually is taking place within the volumetrics or not? and a request for them to look in to how the hyperion render or terragen is approaching the advanced cloud illumination..I have linked to several white papers on that matter here on this forum thread, or the disney cloud asset page.
Edit...See post 167 in this thread.


https://forums.newtek.com/attachment.php?attachmentid=142306&d=1532007335
142306

prometheus
07-19-2018, 11:41 AM
I really like how lightwave 2018 at least can do midlevel low clouds, or whispy clouds...
Hereīs yet another one, this one based only upon the turbulent noise fractal, also another turbulent noise fed in to rotation slot..I couldnīt use the position slot this time since it was referenced to a null..and thus constantly fixed with that no matter what you try to feed in to the position slot.

it could use a little more smoothness and quality boost, but I am only willing to tweak up til a certain point ...considering I can not save in discovery mode.
Again, do not forget to add sunsky suncolor modifier in the lights color envelope channel, and make sure to balance scattering scale properly to get enough depth for a slight cloud shading..even if the clouds are thin.


https://forums.newtek.com/attachment.php?attachmentid=142308&d=1532022054
142308

prometheus
07-20-2018, 12:59 PM
was fiddling yesterday with houdini..mainly to test cloud shapes, modeled in lightwave..and then export back to lightwave through openVDB.

as far as working with clouds in houdini VS lightwave..both have their cons and pros, once you have created your cloud in houdini...you can get full realtime feedback with something else flying and where it should fly etc...but, you have to cook the volumetrics in order for the houdini opengl to work in realtime...and that is valid for changing lights, or any little tweak you do to the cloud..which means depending on resolution it can take a bit of a time..maybe 20-30 seconds for it to cook before you can use the viewport, that can be a bit annoying.

For lightwave there is only VPR iteration update speed to worry about..it means you will never get realtime viewports..but constant refreshments withing seconds, then a bit more before it has refined..I think I prefer that way (if it werenīt for that 2019 volumetrics needs to render faster..GPU please)

Ideally I would want Lightwaveīs way of VPR updating for any cloud or light tweak..so you can tweak faster without having to wait for a cooking of the volumetrics, but only when you are pleased with the cloud and light and shadows, start cooking it to realtime openGL like houdini does.

Cloud object modeled in lightwave..one of the most advanced models ever made:D

The spatial scale in the volumetrics in houdini i set to 2, which means it doesnīt maintain the original shape that much in this case, otherwise ...applying volumetrics to any model or object is of course an advantage over
Lightwaves simple cone, box and sphere shapes, to note..even though both blender and modo can use any shape as volumetric..those doesnīt come close to how houdini deals with objects to volumetrics.

The actual cloud noise, well ..Lightwave have more fractals to choose from, but the provided fractals in houdini is sort of split in to various special features, spatial scale, main noise, advection noise..together if you know them well, it may actually be easer to get realistic cloud noise compared to how much tweaking you need to do with Lightwave procedurals.

Houdini render could be improved with better light scattering and resolution, just testing..
Lightwave renders..well, itīs hard to get a good match..completely different lighting system, and lightwave doesnīt seem to have a good way of setting light exponentional smooth curves that scatters light in a certain way..as houdini does.


https://forums.newtek.com/attachment.php?attachmentid=142319&d=1532112846
https://forums.newtek.com/attachment.php?attachmentid=142317&d=1532112817
https://forums.newtek.com/attachment.php?attachmentid=142318&d=1532112831
https://forums.newtek.com/attachment.php?attachmentid=142320&d=1532112867


142319
142317
142318
142320

jwiede
07-20-2018, 02:30 PM
More local morning cirrus pictures for reference. Enjoy!

142321

142322

142323

142324

Funny thing about local weather, we frequently have lots of these in early morning but they typically burn off by around 11am. Unfortunately, the first has a semi-dispersed contrail mixed in among the clouds, sorry.

erikals
07-20-2018, 04:10 PM
Art!   :)

Dan Ritchie
07-21-2018, 11:06 AM
Dang lightwave, you've grown!

prometheus
07-21-2018, 05:50 PM
More local morning cirrus pictures for reference. Enjoy!

Funny thing about local weather, we frequently have lots of these in early morning but they typically burn off by around 11am. Unfortunately, the first has a semi-dispersed contrail mixed in among the clouds, sorry.


Hi jwiede, did you use a cellular phone for the shot? what phone? or camera?

I took some snapshots during shopping etc, with my samsung 3 ..so not the best quality, I need a new phone with better camera, or always keep my canon powershot with me for better photos.
I should actually think of getting myself a real pro camera to take the shots to another level of quality.

Today itīs been as hot as the latests weeks..around 28-30 celsius+, Itīs gonna be a record warm summer I think (I am a bit scared that a predator may show up on my front door :D)
I am having problems sleeping, and I really would like to have some nice soft warm rain during summer, but nada so far, Itīs unusual and unfortunate for me who really loves rain to some degree.

Some thunder frontīs are arriving, but all it has done mostly is actually causing a lot of lightnings, and a lot of the huge forest fires we got these weeks are caused by that..but no rain or very little rain.
The forest fires this year is also record breaking...se map

https://www.sosalarm.se/contentassets/0ecb95670905422f95306fc8486d1295/karta-20180721-20.png


Snapshots of today...loveley evening, crickets playing constantly all day and night..summer night is the best.
the sunset images (almost sunset) are taken around 21:12 in the evening (24 hour clock)
Letīs see if the link is working properly, I hated when they dropped picasaweb for google photos.
I also wish I had a better camera with me than my lousy samsung S3, itīs about catching the moment...the moment.

Album..
https://photos.app.goo.gl/VT2g2NsEgKSnPHQG9

Some of the snapshots..more in the Album link above..
https://lh3.googleusercontent.com/aFM1-2mlVVC8zkH7DADLZpHh6eBcXScPiSWCEgFhYBvHg5bdZgjzeZu 2hP3x5jXzi97wnzJV48PZW7hqwTHvMbeJ2MpS7BbwTuGEzGuPx DKwDKOsTuq-MQQ8nd_jHmDYK4zydSNnqYKPYB3xv5imjsheiIeOzloqYCJB63 sYGb7hx3RviCN3sL9gC0QO4I9hcVJ48Kkg-Zx99O7NPK-E1gQrrQZs-IrZm-5E4844_RZLryQ9wQFVvTkryl563NCgKoas4VLeYpObseo36t1G G4gjH951OSCuCKhy3Nepc5QQcL5XRF6tt_Osl7CXwuL9j-8j5uSiNu3vecG1TgPv0JTitIHSpF6bPtUDGuXstx6I7pX9fNL_ Uwtrsvtg8cj54RZh81ta8xlSF9tqLR8gfKhvzyWU3ik2WBXQCb SM3-PAeDkAOX2_Elp1t5WOIAgzAg-kdPcPrUChT370ZfjE66LE48DgotmFdcmRXq7k9ipVc4UJxDwL9 ZVUDYLfiVz10a9FSVbmaYPwXRgNvXR6B1xQ9a2Xf5CMCH06oz3 0l7Oe7th-w9wb9oHb7KmaKsCTuJs30hSlf31x47sjswhmVAKkUJWklSQEUA hL-ljZKI_ZIlkEKpqSJmiJAQmR56tRbm452mEsUPCHYZ1Cg2poawV DIJbOZXUi=w1380-h1035-no

https://lh3.googleusercontent.com/G42mbTfRguWooLpiQq_vLZOkZynFUBiAf3qKZLWjvOZMkN5TBU Mp7YTNads9boIq9t1eBW6i7tQ75XvnDRcMc7-9kE09LAif5rZ_Z7VjJiPoRxYRodxOltXT0TniItoQ4QJnhd0yS oH3lc2UPUdeZscaWqsAIiObkOwbTCVgVXd3_YqfLh9P2EY-tU1UBF_GkxWv_jPGLxaI1IIHt0-6mqWkttDYc9aYuEFBO-8VAjDav4zqLBGz5QI1nAtXPtzt8Gyly-Pk1YqHso3rDajOFQxtSNf5JQXx-iJZoaO1vAnLlMoGOld1QEzPj1FKKBmvTdr_8ojZj3kh8_G9Ral 7DDFAIaXuT-ICRQLxen3mZfeDgh980gfy5owbB3SDSj9bw_tnyPUDKT64hgKH GTTb-7Kv3hHtRhQBrga_-BaQmOIvTsDmFw0u22IbA0otTUvlW2pzMfbap9hBtKpLtxm0CFa p1BTjgAt_omhm6fLB7Fc0PgBOIVhuAXFZaAxK6gGBViFl8RpPD wrSSL5QcmPkxW8xsDFSLiqtvEJVumFABft8i0Rw9v84LhHZd2X c_F-gl542HF2wEEh3Ru815qbyzrbPFoeRV_8TcfE-lNJmMdZ6NbfDBcfSsr0NKd_KGhBHRPBNRvyNc3DqVhH62UdKjW wVjNaUSukT=w1380-h1035-no

https://lh3.googleusercontent.com/arL0rNZnqoR7KDcN30n-15NG0uxaCTlI7lO_lCkF4nMnT1EuYsp6NoT9oRfK3tQFLI5nf6 NiwxE8cNJUng7u77XEZzGahV2ZReot5ubdopqsrX2n2rhgR9FV 7OTbuDR9uqwNmVuXPSCeIe4SmAs79brmY7wR4ZgdcrwmbzPTUA agRpfVDUI_kWikJP_9RJc3JdUQ_bkyKjMl3fk--XOJ-d13_XhzRBvv3Rxerzv4OfqCCh4d4u78DYmPZITBhbhJV3dhDgx g86Wg36-ZWhumBiAAVhTWz-a2dsMR7333rqKydCk6OKk-_yQ7E7QzS51222dLHT9scYjnFNHL7SEFETeBxRGdBqeVnsxXCd jYmc0FbZnrWqzv3mJEE1NL6LXPYOFmNc1nvkr7Zy92rdB1P2bM-oD6oeJFuq5z71DFryev9n52LQkiTOcIUfq0tMhwWaR4Q1BBIDZ Y0JDGa_85B7EyJWRP0AVnrIAVVe7aTgrXp6Zh9QPSNZvGffafr jsGNn9MJa55SJdVeRHidbq4VTAvj9MWMqRD9qe5_J0bpOj2M4M iY0NB-OnuAHScSWkaCiDFwlgBsJ8UVO4e8iI5ah3NJiA-FhREAQ05je4=w1380-h1035-no


https://lh3.googleusercontent.com/rciDXXGo98bg3y5FGxGQ87-q46ZI7rLWLXBBd7LLl5fEXDYwonLZwin_p1GyeNirgS1ckcOgB UU9fmaYMrXVGAY-aGchoAtS7Y2dEvR2SSOoCHMs7ykds-gtSQ0BOZJjXJMorcV1G-eKlQrvsCYBYb10peDKhHxPgixSVNQu2Vcw3eBZDz4i4ah_GgG7 Zr5NI1kHfF_DiAggMgoqy5mAphtVHrI6jGQvYPnbAgAJHggc5-LK7zA7h-8Dx27Zz9Dwq-JqNllN3fiiiVm32-0BBQH67zUMghPgj3J609aFWDePR970eBBzNyEY5TVwPMec-O3XdHtMsayuSEO6E_3ab5OBN9u-A_fE5QlWCEUXj7xJG9RhcFtkS_ZXLKNVsbOJxgZwM9TTb14T3_ pJEGXe0k2JlJ8aBg-Q5pv_3VJoKerp0xLG2_sCzFyHHXqE2iLRXFLNdBPDlnMtkrlSX wHOJ5gx19ItvNJbCY7uL9Rcgg69vll6Kr-OqZrLXzw_Nw2ZeFNbuHWTeaRBaF-IS1Snbehrqssg5iS7I__mv2vE_mXUdGzGFgsgZcrKtoS8xu_S0 toOf6CmwTka1AMS3drXgPKKav_I-IQvhqUnp_mBVmaqghQzPDgddm-emVX6Kxjrecoj0s64usxqNK56YJ1TFZWyviXarSI9=w1380-h1035-no


Next in line for this thread, I would like to discuss various ways of modeling clouds, or not..use particles, points..or just procedural textures and nulls, use fluids etc..I have some more experiments to do before I engage in that though, and I would like to start my trial with vueīs latest ple version upcoming days.

jwiede
07-21-2018, 08:12 PM
Hi jwiede, did you use a cellular phone for the shot? what phone? or camera?

Just standard iPhone 6Plus camera here. I usually take em while grabbing coffee in the morning, never have my real camera/lenses handy.

Those cloud shots of yours are great!

While I don't miss summer in the midwest (hot & humid), I do occasionally miss summer thunderstorms and weather just a little. Those cirrus clouds are about the closest we get in summer.

prometheus
07-22-2018, 05:48 AM
Just standard iPhone 6Plus camera here. I usually take em while grabbing coffee in the morning, never have my real camera/lenses handy.

Those cloud shots of yours are great!

While I don't miss summer in the midwest (hot & humid), I do occasionally miss summer thunderstorms and weather just a little. Those cirrus clouds are about the closest we get in summer.



Yeah..getting nice shots requires the right weather, being able to be out a couple of hours..and have a good viewfinder kind of eye,
there are circumstances that gives promises of nice clouds, and yesterday was such a day...you know that when itīs been hot a couple of weeks,
and the forecast talks about thunder...it should start boiling up some nice clouds..and along with that comes a kind of humidity in the air that breaks the light in spectacular colors and also yielding crepuscular rays..so you
better be prepared and go out for cloud hunting.

To bad I didnīt have the time yesterday...just a few shots, yesterday apart from these little shots..there was some even more spectacular atmosphere and when the sun was semi red, going down in these clouds with a nice flare shining through..and multiple cloud layers of dense white, yellow, pinkish..and further away darkblue huge background clouds, I was on a bus and missed the opportunity.

Currently the air humidity changes between 60-65-75 and some days 80, the forecast predicts the heatwave of 26-30 celsius plus to be even worse this coming week when I start to work again, and goes on in to next week with possible even higher temps, I do miss a heavy summer thunderstorm now, japan is a bit worse around 38 celsius plus for over a whole week.

Air conditioner? well..we often get by with some fans, last summer not a day over 23 plus..and the summers are often short so itīs often not worth the investment, and living in an rental apartment makes it a bit harder to install it properly.


funny....early this morning before I went to sleep at 4:AM, I noticed these soft clouds...and it somewhat resembled the style of clouds I did some days ago, If I could have saved the scene I could turn of the shadows and depth and soften them to get very close to the real thing here in this shot.

https://forums.newtek.com/attachment.php?attachmentid=142349&d=1532259931

142349

Kryslin
07-22-2018, 09:11 AM
Just standard iPhone 6Plus camera here. I usually take em while grabbing coffee in the morning, never have my real camera/lenses handy.

Those cloud shots of yours are great!

While I don't miss summer in the midwest (hot & humid), I do occasionally miss summer thunderstorms and weather just a little. Those cirrus clouds are about the closest we get in summer.

Agree that all the cloud pictures here are great.

Jweide, I can bet that there's one other thing you don't miss from the midwest... Tornadoes. I should have gotten a picture last thursday of the supercell storm whose thunderheads I could see from 100 miles /161 km away... It dropped three twisters into Bondurant, Pella, and Marshaltown IA. No fatalities, around 2 dozen injuries, but lots of property destruction.

jwiede
07-22-2018, 05:28 PM
Agree that all the cloud pictures here are great.

Jweide, I can bet that there's one other thing you don't miss from the midwest... Tornadoes. I should have gotten a picture last thursday of the supercell storm whose thunderheads I could see from 100 miles /161 km away... It dropped three twisters into Bondurant, Pella, and Marshaltown IA. No fatalities, around 2 dozen injuries, but lots of property destruction.

Hehe, I grew up in Evanston (N Chicago suburb) and mid-state IL, and then later spent 9 months living in Tulsa OK once upon a time. I've fulfilled my "midwestern tornado experience" card a couple times over. :devil: I do miss giant thunderstorms (really, thunderstorms, period), but no, I do not miss tornadoes.

Coincidentally, years ago, I was at SGI in their then-new, mostly glass building in Mtn View, when an (incredibly rare) tornado hit Mtn View & Palo Alto. I saw a big crowd out on a big balcony looking up, we'd just had a real thunderstorm so figured it was one or more rainbows, fairly common here. I came out to see what they were watching, and realized they were all watching a funnel cloud that had just been directly above descending towards a couple miles in distance.

I had to yell and cajole them back inside to safer areas because so few had ever seen an actual tornado before. It was just a tiny f1-f2 mostly white one, but still, they were just standing there outside surrounded by glass curtain walls (in an HVAC positive-pressure building) watching from _well within_ "debris-fall danger close" distance. Very, very intelligent people, but many direly lacking in common sense (myself included). ;D

jwiede
07-22-2018, 05:30 PM
(double-post)

prometheus
07-22-2018, 06:41 PM
tournedos? sounds good in the oven, otherwise we donīt have that in our landscape.

erikals
07-23-2018, 04:40 AM
no?   :)
https://www.youtube.com/watch?v=Z1Z6UwfQ3xc

prometheus
07-23-2018, 06:57 AM
no?   :)
https://www.youtube.com/watch?v=Z1Z6UwfQ3xc

Norway...most likely to have more Tornadoes I reckon than Sweden.

I wasnīt exactly truthfully when I said we donīt have that, in the south and along those coastlines, there may be much more than over stockholm.
I have seen some smaller ones, maybe that was the so called gustnado?
It is very very rarely seen..at least over land here, if youré out at sea more often..you may catch a sight of some smaller ones every now and then.

We are often very protected from extreme weather here in sweden, so not many natural disasters happening here...though right now we do have a case of record heat and record forest fires though, we had to call in international help this year and the fires are right now just
contained but it can not be put out..we have to wait for rain, and that will take a while before a rain front arrives.

hereīs a translate of the swedish weather forecast description of tornadoes, with comparison notes of Swedish Tornadoes VS those in the United States.
https://www.microsofttranslator.com/bv.aspx?from=sv&to=en&refd=www.translatoruser.net&r=true&a=https%3A%2F%2Fwww.smhi.se%2Fkunskapsbanken%2Fmet eorologi%2Ftromber-1.3875

Back to cloud topics...I am getting the hang of houdini cloudFX tools more and more, and it may be that I end up investing in houdini indie, I feel that I now can get the main shape of clouds, designe exactly as I want..along with understanding the various noise and advection noise and secondary shapes, I need to learn a bit more about the lighting, but I am on my way on that and recently figured out a bit regarding multiscattering and light resolution in houdini, I will also do some tests soon on how that light data can work within the exported VDB to lightwave.

If I compare to Lightwave..there is no way in Lightwave that I can design it exactly as I want in the same manner, it is guess works when you only can use nulls and texture to positiona and scale the density and texture to something that resembles your desired main cloud shape, you could try particles/ points..use fill solid etc, but that will yield density volumes in spherical falloffs based on each point, and the blending mode is almost the same as old hypervoxels when you use the new volumetrics on particles, on single nulls itīs a different blending result that is much much better though.

The newtek team needs to take a look on how houdini does this, do not look at modo or blender and how they apply itīs volumetrics to geometric shapes..it just doesnīt compete with how houdini does it.

prometheus
07-23-2018, 07:21 AM
itīs extremely easy to set up some spheres, or used even more advanced modelled objects as a basis for your designed cloud and then import to houdini and apply cloudFX.
I will have a look of how to setup various models in easy ways for exporting to houdini and see how that goes, and also various ways of model and shaping cloud geometry in Lightwave and blender.
If I get the time, I would like to record some of that process from setting the density exactly as the geometry is, then go on with adding noise and advection..maybe secondary shapes etc, to get a decent noise in the cloud as well.
Ivé seen some basic tutorials on cloudFX, but frankly...most of those are not looking good in terms of cloud noise.

I must add, the way I can get softer nice whispy clouds in Lightwave, and the outstanding results I get with 2018 for such things......that I need to discover and figure out within Houdini as well.


https://forums.newtek.com/attachment.php?attachmentid=142350&d=1532351911
https://forums.newtek.com/attachment.php?attachmentid=142351&d=1532351924


142350

142351

erikals
07-23-2018, 08:25 AM
though right now we do have a case of record heat and record forest fires though, we had to call in international help this year and the fires are right now just
yeah, same here, hottest / driest summer in 60 years or so...

https://i.imgur.com/IaTKyhG.png

prometheus
07-23-2018, 08:33 AM
yeah, same here, hottest / driest summer in 60 years or so...

https://i.imgur.com/IaTKyhG.png

Just noticed.. a shame that it destroys our great forests...I pity those owners of such forest areas.

erikals
07-23-2018, 09:46 AM
temperatures during the next days...

26°  26°  30°  31°  31°  24°  26°  26°  26°

...it's been like that for months...  :/

jwiede
07-23-2018, 11:41 AM
Just noticed.. a shame that it destroys our great forests...I pity those owners of such forest areas.

Here in NorCal we have two seasons: "Rainy" (~Nov-~Feb) and "Burny" (~Jun-Sep), everything else is transition between em. We're currently well into burny season.

prometheus
08-04-2018, 04:55 PM
bumping the thread again, here we go again..
some more nodal experiments, trying to bring forth a little more light attenuation within the cloud, but I am not satisfied..itīs hard to do, and Lightwave GI doensnīt seem to perfom any advanced multiple scattering or I am simply doing it wrong.

I am leaning more towards dropping lightwave all together for clouds and focus on clouds in Houdini..but both Houdini has their pros and cons, I jump back and forth between them when I get tired of a certain issue in either program.
Houdini is pretty neat in the way that I can just swap out any modeled geometry for the cloud while I am in the cloudfx node setup..so it maintains all other settings but just changes the modeled shape..thatīs awesome, it also has the benefit of
performing scattering in a better way I think than Lightwave currently does, if you want that multiple scattering inside the cloud, if not...Lightwave volumetrics can be very sweet as well for such circumstances a deeper scattering isnīt there in the scene.

I would recommend anyone starting with the provided cloud content in Lightwave to swap the turbulence noise out and instead use turbulent noise, itīs better and with more tweak parameters.


https://forums.newtek.com/attachment.php?attachmentid=142442&d=1533422708


Would be sweet if the lw team could look in to the disney research of deep scattering with radiance predicting neural networks...If they pull that off along with houdini alike apply volumetrics to any geometry..then we will start to have something really strong and then perhaps a Real talk about state of the art volumetrics.

For perfection..if that is to aim for, just do research of a full global spectral volumetric scattering atmosphere and make new type of infinite cloud layers as well...then there wouldnīt really be much to speak of, except for constanly crushing rendertimes.


https://www.youtube.com/watch?v=Hk--xpI9b5U

JohnMarchant
08-05-2018, 12:49 PM
Good Idea that, decided to try it on the LW 2018 content, Primitives, Earth scene, not to bad.

prometheus
08-05-2018, 01:29 PM
Good Idea that, decided to try it on the LW 2018 content, Primitives, Earth scene, not to bad.


My favourite Noise fractals are...

Turbulent Noise
Gardner Clouds (Dponts Rman collection) Can be more difficult to setup..but if used right it may surpass Turbulent Noise.
Ridged multifractal invertet
Fbm noise
Turbulence

And I still have to discover other procedurals, the beauty of Lightwave procedurals that it has so many, not blender..not modo, not houdini seem to have the same amount, though Houdini makes a very very good use of what it got..and especially since you can use Main Noise and then on top of that..advect the noise with some parameter settings.

Unless using particles for Lightwave 2018 ...there is no way to define a main shape of the clouds..there is no other way than experimenting with tweaking the noise to set the main shape of the clouds (excluding fluid simulation here) the main shape of how the cloud will look is defined by the actual noise which is created by experimenting with position and scaling..you kind of guess to get a good looking cloud, but you can not design it fully as you wish.

That is unlike how Houdini can do it..where you can define the main cloud shape with any kind of modeled geometry, and on top of that..swap out the geometry anytime you like and still keep the whole cloudfx volumetrics you previously set up.

I have just started to work a bit on a little website where I focus on Lightwave-Blender Dummy Guide, and Lightwave Houdini Dummy Guide and with a special section for Making Clouds in 3D
Not sure I will continue with it though, will see..If I got the time, I will put up how various software can make clouds...like Terragen, vue, blender, ligthwave and houdini..what impressions I have of them etc.

Also sections that deals with The New volumetrics in LW 2018 VS the old hypervoxels, sections to compare Lightwave-Blender in some areas and how to import Open VDB from blender to Lightwave, from Houdini CloudFX to Lightwave.

If I feel inspired to do so and if I get the time that is.

prometheus
08-05-2018, 01:58 PM
Good Idea that, decided to try it on the LW 2018 content, Primitives, Earth scene, not to bad.

I forgot..You used turbulent noise as it is, for the type of cumulus cloud I did and posted last time, you have to invert it..the alpha in the turbulent noise node, if not..you will have a fractal more suitable for lower whispy cloud layers.

For the planetary cloud layers..I would also consider taking a lot of time and try the weather procedural from the Rman collection, suggest you raise the cloud size parameter to around 2 for a starter on that, but it depends on so many other parameters and default scaling as well.

The earth scene? I donīt have it..must be newer content, donīt know where to find it either..using the discovery edtion.

JohnMarchant
08-05-2018, 02:20 PM
The earth scene is in the original content download "LightWave 2018 Content\Scenes\Primitive_Shapes\Earth_Volume", i did not realise DPont still worked in 2018. Thanks for that i will try later. Oh i would love to be able to use any primatve shape. I already import VDB from blender and also Turbulance4D on LightWave.

prometheus
08-05-2018, 02:25 PM
The earth scene is in the original content download "LightWave 2018 Content\Scenes\Primitive_Shapes\Earth_Volume", i did not realise DPont still worked in 2018. Thanks for that i will try later.

Itīs not in my standard account, not sure..as I recall I had to create an account for the 2018 demo, but I canīt find any account details for that...login name or password.

And as for DPont rman collection..no the nodes do not work, only the textures...which means you have to add a color or scalar layer node to access the dpont rman gardner clouds or weather, they are however not as flexible as the node versions.

prometheus
08-05-2018, 02:33 PM
as for planetary clouds, I had a thread about that..ages ago with the ol hypervoxels, would like to see how the earth content looks like for the new volumetrics, donīt have time to set something new up right now..

In this image..I think that was the weather procedural, as can be seen slightly with the nodes, though exactly this setup wouldnīt work in 2018 due to broken Dpont rman nodes, so no way of correcting with gain functions etc.

https://forums.newtek.com/attachment.php?attachmentid=137596&d=1502041822

The thread...
https://forums.newtek.com/showthread.php?154296-Revisiting-hypervoxels-planet-clouds-single-voxels

Edit...sorry, failed to read your reply properly..found the scene..thanks.

prometheus
08-05-2018, 02:42 PM
And yes..you can just use that earth scene and add dpont weather or other from the rman collection, if you installed those textures that is, not the node textures..the earth scene is already setup with a color layer so just use drop down list and add weather.
Be careful of switching procedural type with those layers, I got a lot of crashes..it maybe safer to just deactivate a certain cloud layer and add a new one instead.

prometheus
08-05-2018, 03:03 PM
and not to forget, but I am not sure these are compatible with 2018, the simbiont shaders..it had some planetary clouds..
http://www.darksim.com/html/simbiontlw.html

if someone could rewrite them to work as standard procedurals...
http://www.darksim.com/Repository/tx_Space/Space_PlanetClouds.dsts

JohnMarchant
08-05-2018, 04:57 PM
Thanks Mike i will take a look later.

prometheus
09-30-2018, 10:34 AM
Got some spare time and was giving it ago with "larger" scale clouds, this one 723 meters in radius, while scaled almost the double by stretching x and z, so that is almost 1446 roughly in wide and depth, stretched down to much more than half the size on the height.

I think I got the hang of the larger more real life scaled clouds now, itīs completely different as to the old hvīs in the sense that some shading attributes doesnīt stay the same, and requires a completely different ratio balance approach..which is a bit harder to get hang of.

Scattering scale at larger cloud sizes may be best to keep very low.
Turbulent noise (not turbulence) used as almost always, since I think that one gives the best ratio between yielding softer puffs and more turbulent ones.

Radiosity is on and indirect volumetric samples and use background, which enhances the cloud with illumination, still not comparable to terragen or houdini cloud illumination scattering, and not as effective as blenders volumetric bouncing either.
Lightwave still has itīs charm when you think about that it is one of the easiest to set up volumetric noise, limitations as previously mentioned...
quite a bit slower as a CPU render compared to blenders GPU, while maybe equal or faster than houdinis renderer, it doesnīt compete with houdinis object to volumetrics and much more in there.

Give me GPU rendering ( wonder if octane renders these volumetrics faster, or does it require itīs own stuff?) and give me deep scattering of illumination, and give me object to volumetric shaping ..
Then it will start to rock.

sk_sunsky, One single light, one single cloud item, cube shape....much more can be done to enhance it, layering more cloud items..push the shading a bit more.
Slight color change towards more blue values in emission channel so we just donīt have dull grey shading on the clouds.

https://forums.newtek.com/attachment.php?attachmentid=142953&d=1538325203

prometheus
09-30-2018, 01:33 PM
And a little house, From tf3dm..not directly appropiate 3ds surfacing, just wanted a quickie in there(shows roughly the scale of it all..and some lighting and shading changes..
So almost the same volumetric clouds.



https://forums.newtek.com/attachment.php?attachmentid=142956&d=1538335945

jwiede
09-30-2018, 02:40 PM
Alas, I suspect few other users are willing to invest the level of time/experimentation you did in order to reach such capabilities. IMO, LW is moving in a really undesirable direction w.r.t. UI/UX and sheer investment of time/experimentation (in an increasingly experimentation-hostile environment) required to achieve acceptable results with features/"improvements".

ConjureBunny
09-30-2018, 07:35 PM
Got some spare time and was <playing god...>

https://forums.newtek.com/attachment.php?attachmentid=142953&d=1538325203

What the hell. That's ... amazing. I didn't know it could do that.

-Chilton

prometheus
10-01-2018, 08:41 AM
Alas, I suspect few other users are willing to invest the level of time/experimentation you did in order to reach such capabilities. IMO, LW is moving in a really undesirable direction w.r.t. UI/UX and sheer investment of time/experimentation (in an increasingly experimentation-hostile environment) required to achieve acceptable results with features/"improvements".

Yes...agree to some extent here, while you actually can get more realistic clouds in 2018, it is not as easy to work with, tweak and experiment as to what we have in the hypervoxels UI, a bit of a shame they couldnīt just introduce the new
shading within that tab, also allowing for fast drop down switch to the old hypervoxels when needed, for surface liquids, sprites etc.

It is a bit cumbersome to make sure you select the right item, and you can not have any other item selectable and move around in order to tweak the volumetrics, that you could do with old hypervoxels Regardless of which item you were selecting or moving around in the scene, you could always just click in the hv menu tab list and deselect, deactivate, select what ever item you wanted to adjust in shading, while still moving around other elements.

Guess they need to fix that appliance thing firstly, meaning volumetrics on to points which is lacking now, and perhaps they will consider the workflow of being able to choose a hypertexture directly as superior in workflow ..rather than the necessary enter nodes, add node, connect the nodes.

Pulling that off initiall, then carry on with making the rendering of volumetrics faster, (probably GPU neede, then an option to make any shape volumetric ala houdini, (not blender or modo object to volumetrics, itīs not good)
and make sure some illumination deep scattering can occour for that extra cloud lighting effect.

Once that is done, most stuff is done for the actual cloud stuff, and they need to move on to global scattering spectral algorithms ..working on global earth scale, with fog levels.

JohnMarchant
10-01-2018, 09:08 AM
Yes NT seem to have given with one hand and taken away with the other.

prometheus
10-01-2018, 10:23 AM
Yes NT seem to have given with one hand and taken away with the other.

Yes perhaps ..but more constructively would be to show them where things in the workflow have gone from better to worse.
I have written about it many times now..in the forums, but maybe more directed suggestions and pointing out would be better served if aimed at for instance Janti or some others.

I could also do video recordings showasing the Beneifits of old hv UI VS the new volumetrics, however...this requires som time effort from me, and with a knowledge or trust in that someone at newtek actually sees it, or listens to
what I and other users writes, ensuring that means a two way communication..and I havenīt seen any responses from the development team regarding these issues..which sorts of reduces my will of trying to explain it better with recordings.

the thing with the old hv worklow, you could have the menu open at all times, or minimized while working ..moving on other elements..moving around, then just check the desired item in the list to edit, that is now impossible and each time you move a light or camera or other objects, you will loose any volumetric item selection AND editing..which requires you to find the right one and open properties, entering nodes again etc..not very good workflow for volumetrics, I have praised Lightwave before with the old hvīs how it was one of the easiest to setup and work with...though it of course takes some training to understand the shading, but workflow wise..one of the best.

They do listen from time to time, fixing surface adding in new 2018 so we do not get double surfaces(which I often pointed out..and now is fixed) so you need to find out ways to attract their attention.

JohnMarchant
10-01-2018, 11:54 AM
So have i. Like i cant use TFD and the new volumetrics together. HV and Partlces still have a place for many things. Yes of course that is as much jascha problem as NT. Volumetric objects are good for some things and situations, but in other times they are not. I think now they are knuckling down to sort out modeler, which im greatful for.

prometheus
11-18-2018, 02:30 PM
Bumping this again,

Since Dpont now updated sunsky to work with 2018, as well as Rman node texture collection is also now working, previously in 2018 we had to acess any such texture with the
extra color layer or scale layer, with this we now have more option to control the rman texture with other nodes.


https://forums.newtek.com/attachment.php?attachmentid=143377&d=1542576480

https://forums.newtek.com/attachment.php?attachmentid=143376&d=1542576461


143377


143376

mav3rick
01-31-2019, 10:30 AM
hey prometheus :) any update on clouds with lw 2019 :)

erikals
01-31-2019, 04:19 PM
should be able to create some cool new effects with 2019.

prometheus
02-01-2019, 06:34 AM
hey prometheus :) any update on clouds with lw 2019 :)

Ivé been sick with pneumonia for some weeks, and have only strength to read som mail and maybe some forum threads, but I havenīt got any strength to work/play with any 3D stuff.
Thus I have decided to not even download and install the new 2019 until I get some strength back...I reckon it may take several days more before that happens.

Sure..I am excited to try it out, but my energy is reserved to take care of laundry and other home stuff first.

CaptainMarlowe
02-01-2019, 09:22 AM
Hope you get well soon. Pneumonia is not a funny experience, I got one a few years ago, certainly the most painful illness I had with Dengue.

shrox
02-01-2019, 11:27 AM
Ivé been sick with pneumonia for some weeks, and have only strength to read som mail and maybe some forum threads, but I havenīt got any strength to work/play with any 3D stuff.
Thus I have decided to not even download and install the new 2019 until I get some strength back...I reckon it may take several days more before that happens.

Sure..I am excited to try it out, but my energy is reserved to take care of laundry and other home stuff first.

Do get better!

mav3rick
02-03-2019, 07:07 AM
Ivé been sick with pneumonia for some weeks, and have only strength to read som mail and maybe some forum threads, but I havenīt got any strength to work/play with any 3D stuff.
Thus I have decided to not even download and install the new 2019 until I get some strength back...I reckon it may take several days more before that happens.

Sure..I am excited to try it out, but my energy is reserved to take care of laundry and other home stuff first.

ofcourse health first ! get well soon!

prometheus
02-06-2019, 06:29 AM
Well..I am well enough now.
just installed 2019..
first test ..checking the build in sunsky..itīs there natively now..how much it differs from dpont sunsky I donīt know..good itīs there natively finally though.

just tested metamorphic for an half hour, cool, but needs faster acess and needs to work faster on denser meshes.

so..then I got around to try some objects to volume, unfortunately I was trying to feed certain outputs to wrong inputs it seems, and lightwave just shut down twice on me without any crash report.
and when I try to restart..layout is missing all the top menus, so something is screwed up with the configs it seems.

the reason for me trying various inputs..I couldnīt get the emission color to react on this vdb converted mesh, so I guess itīs another approach needed, though I think it is odd that should be needed.

prometheus
02-06-2019, 07:51 AM
well...managed to delete the configs and get the menu back again after lightwave just shutting down.

tried out the advection tool, to deform the mesh to volume, got it working sort of..but not from frame one, so I need to figure that out, it is a velocity advection so I need to check that.
also scrubbing the timeline with this, and lightwave just freezes...and I am not getting anywhere in fast pace with this, I may be treating it wrong though.

Generally..trying to advect a mesh to volume to create additionally noise in a cloud volume, the workflow and approach isnīt very initiuive for me, I find it much easier to work with houdini cloudfx tools for similar advection stuff, and it works more as excpected.

That said, it may be to early for me to give a decent opinion on it..but itīs my first experience and reaction on it, right now just frustration on that Lightwave freezes with my initial advection setuy, and why the emission color scale and color settings doesnīt work with mesh to volume..that buggers me.

prometheus
02-06-2019, 08:09 AM
To get the emission color working, we need to also plug the same grid output in to the input emission grid, that will get the color emission working and the scale settings as well.

I think that should be the default setting within a scene of this type.
Se image of nodes, both scattering grid and emission grid is used.

comparing just using procedurals and volume primitives, using a converted mesh is a bit slow for my liking, but it is promising, if I can solve it and get decent speed feedback, it opens up for more designed clouds.
Need to check various ways first to allow advection per first frame..or just static advection.

https://forums.newtek.com/attachment.php?attachmentid=144028&d=1549466180

prometheus
02-08-2019, 06:49 AM
I am glad to see that we finally have a way to convert a mesh to volume, the tech is interesting and will allow for more designed stuff...though I have a reservation for it being of good practical use... as it stands now I feel that the rendering becomes very slow..compared to using just a volume item, it may be that I havenīt learned the tools to optimize them yet.

Otherwise I see two other major important features that may expand on cloud design.

1. clone volume items as slaves, this means that any change I do in scattering, emission or with the texture ..it will propagate through all the other cloned nulls, If I have them set as slaves (the old dynamit plugin had this)
this will allow for cloud consistency directly viewable with feedback once you start to tweak the master.

2. Points...Points, we need this volumetrics to work on points just as old Hvīs does.

prometheus
02-09-2019, 05:39 AM
I forgot...what I see as useful with 2019...not sure it was implemented now in 2019, that is the...

self shadow intensity..those options werenīt there in 2018, and it helps controlling the look of the cloud density a bit, so that is a "hidden" improvement.
itīs located at the very bottom in the volume item primitiv settings.




https://forums.newtek.com/attachment.php?attachmentid=144068&d=1549716228
144068

erikals
02-09-2019, 02:35 PM
self shadow intensity

a good addition, thanks for pointing that out.

jwiede
02-09-2019, 03:26 PM
self shadow intensity

Actually, there are two additions: "Self Shadow Density" and "Shadow Density" (note "density" not "intensity"), and both are quite useful additions.

They're good, useful additions, but it'd also be nice if we had control over the shadows' color, not just the density -- there are definitely situations when you need to adjust/animate the shadows' color as well as the density to achieve the desired result.

erikals
02-09-2019, 03:34 PM
true, more options are welcome, pushes the clouds a bit further.

prometheus
02-09-2019, 04:53 PM
Actually, there are two additions: "Self Shadow Density" and "Shadow Density" (note "density" not "intensity"), and both are quite useful additions.

They're good, useful additions, but it'd also be nice if we had control over the shadows' color, not just the density -- there are definitely situations when you need to adjust/animate the shadows' color as well as the density to achieve the desired result.

I think a cloud self shadow effect (with color) is the result inherent from how you set emission and scattering color, perhaps absorbtion color, not sure if itīs by any is physicly correct to do it another way.
Usually some light emission is transported within the volume, so increasing emission scale in this scene would yield a blue grey cloud, had I raised emission scale a bit.

The new physical sky I am testing, it is of course similar to dponts sunsky...I wonder which one holds the upper hand, I think this new physical sky with the new sunlight type..seem to yield a better color and sundisc setting ..if I lower the sun temp to get a more sunset sky, with big sun..sort of fading if you set a certain amount of turbidity.
Now this physical sky needs to play along with proper setting for the clouds, so when you lower sun etc..the clouds should emulate a realistic case of coloring, but there is quite a few weather circumstances that one setting canīt match it all.

Shadow density doesnīt do anything for the actual cloud, but controls shadow density on the ground, if you need to lower that density shadows cast by clouds...where the self shadow density actually affects the shadows within the cloud.

mav3rick
02-14-2019, 02:57 AM
iv just recently finished TVC with 2019 volumetrics and i have to tell new NVIDIA denoiser come to the rescue as i was rendering scene in full HD...
render time with full screen loaded with volumetrics clouds were 40 sec per frame which i find very fast.... only thing i would like to have is OPEN VDB to work and render at speed of VOLUMETRIC primitive . for now its really pain to design clouds and setting things up with OPEN VDB.. compared to VOLUMETRIC PRIMITIVE its 100 times slower to interact... HOUDINI is blazing fast in that department.. i am sure there is room for optimization for both OGL and render speed with OPEN VDB

prometheus
02-14-2019, 10:53 AM
iv just recently finished TVC with 2019 volumetrics and i have to tell new NVIDIA denoiser come to the rescue as i was rendering scene in full HD...
render time with full screen loaded with volumetrics clouds were 40 sec per frame which i find very fast.... only thing i would like to have is OPEN VDB to work and render at speed of VOLUMETRIC primitive . for now its really pain to design clouds and setting things up with OPEN VDB.. compared to VOLUMETRIC PRIMITIVE its 100 times slower to interact... HOUDINI is blazing fast in that department.. i am sure there is room for optimization for both OGL and render speed with OPEN VDB

VDB slowness, that is my initial experience as well..as I also think I did mention before, it doesnīt feel practical to work with it for clouds.
And yes, houdini handles a mesh to volume much faster in my opinion as well.

UnCommonGrafx
02-14-2019, 04:04 PM
For those of you that have played with Houdini for this, do you know if it's calculated on the gpu or cpu?

gar26lw
02-16-2019, 04:33 AM
Alas, I suspect few other users are willing to invest the level of time/experimentation you did in order to reach such capabilities. IMO, LW is moving in a really undesirable direction w.r.t. UI/UX and sheer investment of time/experimentation (in an increasingly experimentation-hostile environment) required to achieve acceptable results with features/"improvements".

yeah i agree. i though matt was looking at all that, does he not do that any more?

prometheus
02-16-2019, 07:09 AM
I see an issue with clouds and sky, atmospherics being not implemented as a consolidated system.

We have a volume primitive that isnīt infinite
We have a global volumetric system that isnīt setup, or made in a way that it should work realisticly with few tweakings Together with a cloud system (volumetric primitive)
We now have a new sunlight and a new physical sky, which is only a backdrop solution..it is in no way made as volumetric and spectral..thus not really proper to use with the above mentioned systems.

So from time to time...Like yesterday, I fire up 2015 and play arount with ogo taiki, sometimes itīs way ahead of what I could do with 2019 and itīs scattering and volumetrics.
Comparing rendertimes can be difficult, depends on settings and quality and if I were to use godrays o not.

I think the lw team needs to work more and re-think itīs volumetric system, implement infinite cloud layers, and a physical sky that is connected to a volumetric/spectral system....maybe they should disect the code of Ogo Taiki if they can..and if they are allowed to implement similar code within the newer PBR system, Then again...maybe itīs not tha easy?

With such system, Lw could have something that may knock both ozone and terragen out ..in the sense that it would be almost as good But living inside of Lightwave directly with all itīs benefits.
But Ivé been saying that and dreaming for over 10 years I think.

Bernie2Strokes
02-17-2019, 09:41 PM
In a recent livestream with Lightwave users there was brief demonstration of the new Physical Sky feature. But there was no mention of clouds.

There's a scene I've dreamed of doing outside of photography. In the background there are heavy storm clouds approaching as far back as the horizon, but clear sunlight is coming from the opposite direction. It could range from two o'clock to the Golden Hour position. Are you saying Lightwave 2019 isn't capable of simulating that well enough?

prometheus
02-18-2019, 04:34 AM
In a recent livestream with Lightwave users there was brief demonstration of the new Physical Sky feature. But there was no mention of clouds.

There's a scene I've dreamed of doing outside of photography. In the background there are heavy storm clouds approaching as far back as the horizon, but clear sunlight is coming from the opposite direction. It could range from two o'clock to the Golden Hour position. Are you saying Lightwave 2019 isn't capable of simulating that well enough?

That is a bit funny, when the new system of volumetris was introduced in 2018 lightwave version, Lino showcased it against pure black background, at that time neither sunsky worked fully, nor was there any implemented native sunsky.

And when someone shows sunsky ..it is without clouds, now most of the time..a sky and clouds are often together with a few exceptions of absolute clear sky for a possible arch viz project tasks.
And not only that..except sky and clouds, there is atmospherics.

The new "physical sky" is only simulating it to a degree restricted to a background, turbidity isnīt taking place in any algorithms calculating distance in atmopshere or thickness, it is just simulating it in a backdrop, Dpont had (not sure the atmosphere part was fixed for 2019) sunsky atmosphere which helped simulating a bit of atmospheric scattering.

As for giving answer to the question, you can most likely simulate the scene you are talking about, depends on the look.

As for clouds and setting the sun and clouds correctly for stills, not so much of a problem.
As for clouds and setting the sun and clouds correctly for animation..where it all looks good and decently realisticly no matter how you move the sun, that is a bit trickier.

As for clouds and setting the sun and cloud..with atmopsherics...Thatīs a completely differently level and will cause much much more work to get it right, adding global scattering and getting that to work
with the new physical sky, I believe these system isnīt implemented to work physicly correctly by default, you would either have to tweak artisticly or go through quite a bit of research and math to set it up to work in proximity to software like ozone, vue, terragen.



Please check this list of sunsky related topics in my channel, it is for the 2015 version and older using dp sunsky, the new physical sky is a bit different but similar in many ways.
https://www.youtube.com/user/PrometheusPhamarus/search?query=sunsky
Also check this vid where I use old hvīs with particles and use sunsky to change lighting and how that affects the cloud color...

https://www.youtube.com/watch?v=-6PDjQdPyus

But atmospherics..sometimes you may not need much, but sometimes it really makes a difference.
I would like to see a reference photo of what you intend to reproduce in 3d.

And..avoid straight ahead lighting ..doesnīt look good in any 3d software, nor in reality..you said you wanted opposite lighting to the clouds view, which should be straight lighting?

Bernie2Strokes
02-18-2019, 08:02 AM
I can't find an appropriate photographic example. There's a photo in this article that is close to what I'm aiming for, but with the sun further down opposite the clouds.

https://toistudent.timesofindia.indiatimes.com/news/omg/dark-clouds-looming/35511.html

Or perhaps it's the kind of contrast that's best achieved using a polarized filter.

prometheus
02-18-2019, 09:29 AM
I can't find an appropriate photographic example. There's a photo in this article that is close to what I'm aiming for, but with the sun further down opposite the clouds.

https://toistudent.timesofindia.indiatimes.com/news/omg/dark-clouds-looming/35511.html

Or perhaps it's the kind of contrast that's best achieved using a polarized filter.

That should be doable, at least for the clouds, and the new volumetrics, I think you may need to fake detail with feeding the same texture you use for the shape in to emission or scattering, to bring forth some detail, this new volumetrics doesnīt have any deeper scattering occouring...you can activate GI and use inderect samples is a must in volumetrics, and affected by volumetrics..in order to get some multiscattering going on, but itīs not as advanced as terragen...so faking detail an scattering is one way by using the texture in the emission or scattering channel, it will not be accurate in terms of lighting c onditions and direction and ray depth.

And the cloud wouldnīt be dark grey as in the image if you intend to have the sun going down at that level, it should pick up a redish-violet dark grey tone.

prometheus
02-18-2019, 12:20 PM
Just a reminder...

As you can see in your provided sample file, the clouds stretches to the horizon, and as such they are changing look due to the atmosphereīs spectral properties, loosing blue light and picking up a more redish violet tone at the horizon.
Clouds will also be affected on turbidity, fog, haze etc..which drasticly can change the look, in Lightwave we do not have atmospherics that simulates this properly...the global scattering I believe isnīt a model that takes this in account in realistic ways and
the only ones that does is special dedicated software such as vue and terragen.

Setting global volumetrics in real world scale to simulate these properties... may not be worth it.

To note, in my version of 2019...the new sunlight seem a bit buggy, if I load a saved scene..the colors are not affecting the clouds correctly, I have to Re-choose the new sunlight type..only then the proper color is affecting the clouds when rotating the sunlight.

So color in the clouds will become darker and more orange to red if you set it as low pitched angle standing sun, maybe 4-5 degree, if you go lower than that..you loose those colors and the cloud will almost not be affected by the sunlight color, and if you have set emission strength and color, that will shine through...and it does so unrealisticly, we would need a way for the cloud emission channel to calculate that once the sunlight goes down, the amount of emission in the cloud get lesser and lesser, I think vue and terragen treats this much better, you could also reduce cloud emission completely and only go for only scattering and such, but then you have clouds becoming too dark in areas where the sun isnīt hitting it in the right direction, a workaround is to lower the new self shadow density, start by setting it to Zero as no shadows..and increase til you get shadows that doesnīt becomes to dark within the cloud..but that is why I said earlier that itīs not an enjoyable experience to set up more advanced realistic behaviour that works no matter how I set the sun.

Bernie2Strokes
02-18-2019, 06:24 PM
Thanks for explaining that. Having grown up in suburban areas I rarely get to see far into a horizon. The effect I was looking for was mostly introduced to me through black & white photographs and a Maxfield Parrish painting.