Lightwave volumetric Items, clouds and sky moods, planet clouds, fog, godrays Part 2.

prometheus

REBORN
btw, nice Terragen examples you guys posted.


From my part, my first planet clouds in terragen, yesterday...and I didn´t read any tutorials on how to at all, just raised level and applied clouds, global dense, and tweaked and tweaked.

Rtp sometimes takes some time to refine to a good level I think (cpu dependent) that actual final render at 1 minute, also 3 minutes for 1020x450 isn´t that bad considering the light and atmosphere complexity, but I skipped on major part, no terrain micropolyrefinement than a standard surface ..apply that and it will probably go up quite high.

As I said before, I need to solve why the final renders in cloud scattering turns up completely different than the RTP, could be the tone mapping, could be amount of GI or some other settings, I think I had that issue before and solved it, but forgot about it.

Easy clouds are so nice just to slap in there and you don´t have to do much to tweak noise and scale really, just adjust position and hight mostly...the bad part is that they are world position objects unlike the other cloud layers, and isn´t easy to adjust rotation on, I did try some nodal setups there with shader masks to do that, semi ok results.
Not sure but I think Kevin mentioned something about rotation control in his vids, but not sure if he covered that for the Easy clouds, the other clouds ..no problems with that.

I need to bring forth some suggestions with mockups to planet side, I don´t think it´s acceptable how sliders, values works, not in respons, not in UI ..it has no rotational controls as mentioned and you have to do that with nodes, and easy clouds harder, it has no x-y.-z visual indicators for the values, that is bad.
The distinction between active shaders, sunshade, terrain, and atmosphere activation, on and off is very hard to see since it´s almost the same color tone, with a darker interface and a proper highlight very different from what there is today, it would be so much easier to get an overview of what is active and not.
RTP is also a bit strange ..it has some glitches, not updating atmosphere after pausing, despite a new generation of the whole scene by turning RTP off.


And for you Erik, you have a faster cpu than mine, and you could try the demo yourself as I did, but with faster results, given time and interest..but I think you mentioned it wasn´t something for you right now, pricing and all perhaps?

For me I could deal with the pricing I think, the UI and workflow and some other glitches, including need for faster cpu or gpu for the skies..and vdb import, that is what needs to happen before will consider it...I Think, and rtp and final renders must match exactly..can´t have different previews and end up with a final render that isn´t matching what I initially have tweaked.

For planetary clouds, I may think that terragen has some good starting fractals, but is more limited in direct control and amount of fractals and fractal control than if I set it up in lightwave ..which almost feels it has no limits with all those fractals and controls available, it´s a matter of good eye, know how to tweak and time of course.
The difference is in the lighting model and the atmosphere of course, while you can get very nice results in Lightwave, it is still kind of beaten by the more physicly accurate atmosphere model inside of Terragen.

I checked some nebula samples in Terragen, but boy they were slower than nebs in Lightwave I would say.
 

prometheus

REBORN
afaik it can be adjusted to look very good. the Falcon example kind of shows the potential.
but might indeed not be to a Terragen level.


might very well be. will probably take quite a few years.
was thinking the same yesterday.

unless i had to do it for a production.


Yes, Had I decided to get a new machine with much better CPU this year, I may have continued, but right now I am very unsure in any investment, that means everything from
focusing on volumetrics, and clouds for my projects or training material, since it´s almost required to produce good material in decent times, and it´s uncertainty about my economy partly due to job situations, corona and some other economic issues.

When times get tough I tend to go in to cheap mode and don´t want to invest, but unless you don´t invest you may find yourself in a limbo, and if you do invest, it´s either probably a good investment or a waste if not taking advantage of it the right way...and then you may be in a worse state then the limbo :)

It´s to much shilly-shally mode from me, I know..I need to change that and make better more faster and be more decisive, I was sort of conned by someone that ows me a lot, and that put some obstacles in front om in terms of what I intended to invest in.
 

prometheus

REBORN
afaik it can be adjusted to look very good. the Falcon example kind of shows the potential.
but might indeed not be to a Terragen level.


might very well be. will probably take quite a few years.
was thinking the same yesterday.

unless i had to do it for a production.

yes, as for the Falcon flight sample, It looks like he´s not really using VDB cloud sets for that, I could be wrong, but more that it is octanes volume items with it´s noise, and that is where it sort of breaks, too many small puff noise fractals for each volumes surface so to speak, and then the genearl main volume shape is too round as well and it seems like they are placed together.

So that could be a bit telling if it is volume noise only, and fractals only from Octane, if it can not access all LW standard fractals and not the Dpont Rman collection either.

Cloud volumes need to be more larger and none uniform in that cloud volume surface, and for the main volume shape it should be more fused continuous in volume.
It´s not easy though, so it´s just a reflection on how hard it is to acheive realism for anyone really.

To bad the demo of Terragen will not allow me to render animation samples, they should allow that, with watermark.
 

prometheus

REBORN
And to bump again, Lino´s Grandi´s youtube demonstration of octane in blender is quite good, with the noise control and render speed..so with such noise settings and a flight through, I think a Falcon fly through sample would look better than the Shaun sample.

but what Lino demonstrates is hard to grasp how it would work in larger scale, he doesn´t showcase how big the volume is, perhaps only some meters, so that means he would have to scale down meshes like the falcon, animation and get all that in sync with a meter cube or something, or..Actually work a large real scale, wich could be several km´s of cloud layer for it to look decent with a fly through, and how that would impact rendertime, scaling of noise and scattering..that is hard to say.

But octane for blender was just to messy for me to try and install, will get back to that later perhaps.
 

I was sort of conned by someone that owes me a lot, and that put some obstacles in front om in terms of what I intended to invest in.
not cool, hope you get it back. i've had situations like that also, not fun.
in my case i chose to take distance to the guy, since i knew getting it back was a no-go, since it wasn't a "priority" to him.
but if your case is different, make sure to discuss it with family members or friends.

now, dinner,... 9.00pm

To bad the demo of Terragen will not allow me to render animation samples, they should allow that, with watermark.
yes, a shame, i take it rendering to an image sequence won't work.

When times get tough
the rules i use >
- how much does it cost
- how long will it last
- how much $ can i spend

 
Last edited:

prometheus

REBORN
not cool, hope you get it back. i've had situations like that also, not fun.
in my case i chose to take distance to the guy, since i knew getting it back was a no-go, since it wasn't a "priority" to him.
but if your case is different, make sure to discuss it with family members or friends.

now, dinner,... 9.00pm


yes, a shame, i take it rendering to an image sequence won't work.


the rules i use >
- how much does it cost
- how long will it last
- how much $ can i spend


Well..I think I will get it back, but it´s taking longer time than agreed, so I get montly payments, but it´s gonna take some years probably.

It´s a relative , and Unless the person get´s injured, death etc..I think I will get it Back, but the person isn´t prioritizing it and spend on other things than to pay me off, not
following the plan we agreed on which makes me mad, but it is what it is and I will probably just shut down any future relation with this person ..when it´s all over.
No need to go in to this further though, to personal.

as for terragen, a bit of a shame, testing animation capabilities is crucial for evaluation, and not just stills, that I think vue does right..yet smaller size and watermark, and that is what Terragen needs to do as well to gain more interested clients (among other things)

Limitations are..
The free version may only be used for non-commercial purposes, or 30 days of commercial evaluation, and the following basic limitations:
  • Render Size Maximum: 1280×900.
  • Render Detail Maximum: 0.6.
  • Antialiasing Maximum: 6.
  • Maximum Number of Rendered Populations: 3.
  • No Animation.
 
Last edited:

prometheus

REBORN
A tip for using the Lightwave volumetrics a bit faster...

If you can get away with it that is, using a scalar layer and a fractal like weather for clouds on a plane or sphere volume, will be so much faster than it´s 3D nodal fractal relative.
makes vpr refresh so much faster compared to plugin the 3D fractal in there, but you may loose out on more advanced nodal stuff to further adjust it.

but otherwise it´s faster, and easier to simply layer smack an image map for density distribution, then just another factal layer on top of it and the result is fractal height distortion and some slight edge distortion, but using a different procedural or image map for density.
 

+was also thinking about putting point lights inside clouds. might work, but would be very time consuming.

scalar layer and a fractal like weather for clouds on a plane or sphere volume
noted.


looked a bit more into it, and it stands between Terragen or Octane
... or [.B] as an alternative.

 

prometheus

REBORN
+was also thinking about putting point lights inside clouds. might work, but would be very time consuming.


noted.


looked a bit more into it, and it stands between Terragen or Octane
... or [.B] as an alternative.

Blenders native fractals aren´t good enough for this, octane with blender perhaps, or probably required.
Ivé seen and have a scenefile someone did ..with a special cloud shader with a lot´s of noise nodes connected, that is a problem since the renderer has to process all those, it´s a horrible mess of nodes as well, compared to the ease of using one, or two..maybe 3 fractals for the actual cloud fractal noise nodes (additional nodes, clamping, math etc excluded)

Haven´t gotten in to blender planetary clouds, maybe I should, I think I posted a sample of someone who made that nicely in the star wars destroyer thread you started.
But time, it´s lightwave, terragen, houdini, blender...in a way.
Maybe I should be glad bentley and vue managed to screw some things up for my interest in that software.

Cinema4D, max, maya, , modo...not even looking at those...though I have tried them all out as general software, some years agoa.
And then I have to continue with davinci resolve, fusion, krita, inkscape, terrain editors, some other design tools..and when it´s time for bed, a terrorist sings for me..blah ;)

Davinci Resolve, the free version isn´t taking me to the level I wan´t though, so many things after effects and premiere does as it should..so I need to look in to video editors more closely again, I mean..resolve is quite nice in some ways, but for some other (simple stuff) I just go..argh, reloading of clips in the timline, saving to gif, cropping to a frame crop tool, and text animation..all that is done much better in Adobe tools.

May be that I need to check in to prescriptions anyway...but only if I decide too, or get a job where I need to work with it, same for photoshop and Illustrator, many years ago I went through some minor education on those tools, and then used them at work as well, that was some time ago though the demands is still there in the job market for people knowing how to with that product.

I may have a chance of sustaining, enhance my skills on Adobe photoshop, illustrator, indesign...by going some free courses, recently sponsored by our government, 9 hours a week, a total of 40 hours...will look in to that soon.
 

I mean..resolve is quite nice in some ways, but for some other (simple stuff) I just go..argh, reloading of clips in the timline, saving to gif, cropping to a frame crop tool, and text animation..all that is done much better in Adobe tools.
true, i actually use the free KdenLive editor for this.
Maybe I should be glad bentley and vue managed to screw some things up for my interest in that software.
yeah, 3D is extremely time consuming, that's 4 sure...
enhance my skills
always a though choice, what to chose that is. so many fields, so little pay...


 

prometheus

REBORN
I would suggest looking a bit in to Terragen, especially now with the new Kevin tutes that feels quite proffessional, and that it´s a workflow between Lightwave and Terragen.

RTP elapse time feedback, on my "slow" cpu, yours should be faster, how much I don´t know..I am also using an older terragen demo, I need to check if the newer demo loads my older demofiles.

Workflow suggestions is to disable shaders of sun, shading, and atmospher and start with the basig terraing generation in low res, when that looks aproximated as you want in main scale, raise or move your viewpoint with the controls, hen turn on HD for refining the detail, when that is finished, turn on RTP for "raytrace previews" then activate the shaders for the sun, atmosphere, and shading.
The next step is to copy the viewpoint to your camera and save the scene, or you will loose all the views.

After that you have to keep in mind that ever camera view change will just render the atmosphere and clouds and the previous views on the terrain, but the terrain will not update since it only generates terrain within the view, so each time you change view, you have to turn off RTP and the generation of the terrain in the new view will start again.

Sample images is the easy clouds, which are the nicest, but a type of clouds you can not tweak that much in terms of shape and rotation and density fractal, it´s also the slowest to compute.

See sample images of the RTP speed, elapsed time...

terragen elapse RTP.jpg
 

prometheus

REBORN
true, i actually use the free KdenLive editor for this.

yeah, 3D is extremely time consuming, that's 4 sure...

always a though choice, what to chose that is. so many fields, so little pay...

virtual dub someone recommended that, for the cropping, what I ment is that whatever frame crop I do in davincie resolve´s edit page, it should be able to transcend to deliver the video in that resolutions, but it´s not that simple, simple static image cropping though, is not what I need.

Yes, get a life and live in the real world, not suspended in 3D Limbo...unlucky me that I picked that as a hobby and passion then :)

About the course I may take, it´s free of cost, so all it would take is to pay attention for 40 hours total, 3 hours/day, at mondays, wednesday, and thursday, until the end of june I think, some of the sections in the course I already know very well, but maybe good to sustaining the skills, and also have that in my CV, if they provide a decent diploma, that´s kind of essential as well.
These courses is something I like to do, unlike that network education which I kind of rushed in to, and it was full time as well.
 

courses are usually cool,
however working in a specific field can be quite boring. (at least for me)
sometimes it is best to just through the education out the window, go with a "simple" job.

 

prometheus

REBORN
thanks for the Terragen Workflow suggestions btw....

saved.



Perhaps Terragen tips and workflow should be posted here...I just added this thread with another tip as well.

And I made a general question in the Lw community forum as well to avoid irritation.

If there is a specific question or comparison between render speed, quality ..workflow, that could probably fit in there though, if related to the topic.
 
Back in Lightwave.
Here is a simple test patch where I shaped a volumetric object (cyclon) with a texture. Sometimes useful to achieve a desired shape.
Setup: a box as the earth's surface, a volumetric object as the cyclone, and a volumetric object as the atmosphere.

Cyclon_016.jpg

ciao
Thomas
 

prometheus

REBORN
Back in Lightwave.
Here is a simple test patch where I shaped a volumetric object (cyclon) with a texture. Sometimes useful to achieve a desired shape.
Setup: a box as the earth's surface, a volumetric object as the cyclone, and a volumetric object as the atmosphere.

View attachment 150068

ciao
Thomas

It´s so so Thomas,

I think the problem is stacking 2d elements and with an emulation of getting depth, it gets step artifacts sort of, and lacks any full 3D volume on the sides, as we can see in your sample there is repetitve stacking.

For better realism I would actually just use the full 3D approach, though granted, takes quite a bit longer time to render.
And of course you would need to know how to tweak the procedurals since using 2d images will not get you there as explained above.
We have coriolis, cyclone etc, or warping with twist, vortex modifiers in various node inputs.

You can do this with a full real scale shape sphere, sphere volumetric atmosphere, and sphere volumetric item and reference the cyclone or twist nodes to a null to get a position and fall off like a true cyclone as well.

Cyclone and coriollis by the way, do not exist as true 3D fractals in Lightwave, they are only available through the scalar, bumt and color layer inside of nodes.
 

prometheus

REBORN
By the way, the vortex cloud content can be a starting point, you would have to scale down the depth/hight though and change fractal layers to use a more puffy cloud style, like fbm, or gardner clouds, and adjust the distance to null ref to be not so much rotational, so much you can do with this ..given time and passion to tweak it, it will give you a much more fuller 3D depth on the cloud, than stacking images up, it will take more time to tweak though.

Another experimental thing you could do, if only using partial earth views, that is to convert a small part of a sphere by cutting it out, convert mesh to volume and apply atmospheric vollume to that.
 
Top Bottom