PDA

View Full Version : Requests for Hypervoxels, with comparitive note



prometheus
03-07-2012, 01:12 PM
Ivé been posting some questions on luxology forums to find out how the new volumetric system is in Modo 601 compared to hypervoxels, the answers to that seem to be that Indeed, Modo volumetrics can be applied on the whole geometry, not only vertices on a object.
Im not 100% sure this is the case thou, Have to see it first.

Now of course...Modo doesnīt have a particle system yet..but for cloud things and other stuff, a volumetric shader working on full geometry is what Ive been requesting for lightwave a long..long ..long time, but just as for the Ents council...Newtek is taking a long..long ..long time to implement this.

I would even suggest to fix a proper tension blend first between particles or nulls as I have mentioned before..It should have similar blend as the old dynamite voxel system...and while your at it start working on a new Hypervoxel that can work on full geometry.

houdini, modo, vue..they all got something like this.


Michael

Celshader
03-07-2012, 01:17 PM
Ivé been posting some questions on luxology forums to find out how the new volumetric system is in Modo 601 compared to hypervoxels, the answers to that seem to be that Indeed, Modo volumetrics can be applied on the whole geometry, not only vertices on a object.
Im not 100% sure this is the case thou, Have to see it first.

By "whole geometry," do you mean a teapot can be filled with a volumetric, so you can get a cloud shaped like a teapot? That would be cool.

I'm also curious if modo can create the volumetric cloud tunnel in the opening titles for the most recent DOCTOR WHO:
http://www.youtube.com/watch?v=BuVLJRyg2Bg

geothefaust
03-07-2012, 01:22 PM
By "whole geometry," do you mean a teapot can be filled with a volumetric, so you can get a cloud shaped like a teapot? That would be cool.


Yep, sure is this way. It's pretty cool.

prometheus
03-07-2012, 01:26 PM
By "whole geometry," do you mean a teapot can be filled with a volumetric, so you can get a cloud shaped like a teapot? That would be cool.

I'm also curious if modo can create the volumetric cloud tunnel in the opening titles for the most recent DOCTOR WHO:
http://www.youtube.com/watch?v=BuVLJRyg2Bg

well yeah kind of filled or so, not exactly sure on how the method works, so I suspect you can apply the volumetrics directly on sculpted cloud geometry shapes, and it would have a nice filling blend withing that object, not per vertices or particle level as Lightwave have had for several years.

Not sure how similar the shader is to houdiniīs volumetric shader wich works on geometry level too, or if it is similar to vueīs meta cloud (those are limitied to spheres only thou and do not work on sculpted geometry)

In Lightwave 11 we can fill a geometry object with modelers fill solid, so you get point clusters in the shape of that geometry and the can use hvīs

But that is limited..and you are almost certain to get round density around each vertices or bad puff look and poor blending.

Michael

erikals
03-07-2012, 02:15 PM
Have to see it first

yep.

Elmar Moelzer
03-07-2012, 02:33 PM
Not sure about Modo, but VoluMedic CE can fill a fully enclosed polygon object with a volume, but it is currently not an officially supported feature due to a couple of smaller issues (no crashing or anything).

erikals
03-07-2012, 02:44 PM
would it work with for example,

5 enclosed spheres at once?
deformed spheres?
animated spheres?

Celshader
03-07-2012, 02:51 PM
Not sure about Modo, but VoluMedic CE can fill a fully enclosed polygon object with a volume, but it is currently not an officially supported feature due to a couple of smaller issues (no crashing or anything).

This has my attention. We used to have this exact functionality in LightWave with WaveFilter (http://www.wavefilter.com/WFVolumeFAQ.html), and Dave's looking for something to fill WaveFilter's shoes..

Elmar Moelzer
03-07-2012, 02:55 PM
http://www.youtube.com/watch?v=isBvUebQueE

Here is an example that I made really quick with VoluMedic CE. I know it sucks, but it really was just a testcase.

prometheus
03-07-2012, 11:18 PM
volumedic ..yeahh and wow, but this is needed in Native lightwave for improving a long forgotten Hypervoxels that was thrown deep in the pit of underdevelopment.

Huge cost for it to happen with volumedic, unless you really have use of volumedic otherwise, you could as easily invest in modo and get that + a whole lot of other stuff, perhaps
even get houdini apprentice HD wich turns geometry to volumes in a snap.

Michael

jay3d
03-07-2012, 11:35 PM
In fact LightWave itself is under developed, from 2002 till now Luxology managed to build an App from scratch that surpassed LW in almost every way, look at the progress of LW from 2002 till now?

Celshader
03-07-2012, 11:47 PM
Huge cost for it to happen with volumedic, unless you really have use of volumedic otherwise, you could as easily invest in modo and get that + a whole lot of other stuff, perhaps
even get houdini apprentice HD wich turns geometry to volumes in a snap.

Volumedic CE costs $299 (http://www.sharbor.com/products/MSGN0280015.html). That's less than half the cost of a $795 modo crossgrade.


In fact LightWave itself is under developed, from 2002 till now Luxology managed to build an App from scratch that surpassed LW in almost every way, look at the progress of LW from 2002 till now?

Luxology beats NewTek for marketing, I'll give them that. :ohmy:

jay3d
03-07-2012, 11:51 PM
Luxology beats NewTek for marketing, I'll give them that.

I don't think it's just marketing, look at the quality of the program features.
LW 11 just loses to a simple instancing test.

geo_n
03-07-2012, 11:56 PM
This is definitely impressive. Might save some people from buying ozone.
http://www.youtube.com/watch?v=-MXug_ZGUMY&feature=youtu.be

I'd say 2 years more and modo will be overlapping lightwave in most aspects at the pace these two are moving.

prometheus
03-08-2012, 12:58 AM
In fact LightWave itself is under developed, from 2002 till now Luxology managed to build an App from scratch that surpassed LW in almost every way, look at the progress of LW from 2002 till now?

In some areas ..yes, but in others Modo doesnīt match up, and as we all perhaps know, old team and their thoughts of shutting down comletly and build from scratch would have ment no Lightwave and missing a lot of stuff meanwhile.
Besides Modo seems to suffer from a couple of things still before it can replace Lightwave as a full app.

Personally after trying Modo demos, even thou Itīs grandios subdivision modeler, fancy interface and all other stuff, i still feel more comfortable with Lightwave due to that it feels faster and more easier to find itīs way through cameras and lights and surfacing.

but I donīt wanīt to go in to that discussion really, narrowing it down to Hypervoxels and volumetric objects in this thread is easier to bring forward constructive criticism. the areas mentioned of what is lacking is what Newtek can overcome and improve on once they listen and put this in to their schedule of whatīs priority to work on, Im sure they would like to improve hypervoxels, but if they have too much other stuff on the agenda or is missing skilled resources, well..then Itīs gonna take a long..long..long time to get anything done as the Ents say:)

Michael

prometheus
03-08-2012, 01:06 AM
Volumedic CE costs $299 (http://www.sharbor.com/products/MSGN0280015.html). That's less than half the cost of a $795 modo crossgrade.



Luxology beats NewTek for marketing, I'll give them that. :ohmy:

Aha..I missed that volumedic has a lite version, and that it has the feature of creating similar volumetrics on meshes?

anyway, it is still too much for a single feature, what do you get for the purchase of modo vs volumedic, you would end up using a tool mainly aimed for special medical & sience visualisation in order to get this object volumetric feature, and would you really need the other stuff, now modo would + that with a whole lot of other tools more full in features for the whole 3d range.

But neither volumedic or modo is what we should cry out loud for, it is Native lightwave and Itīs volumetric handler that needs attention.

Michael

jay3d
03-08-2012, 01:11 AM
Constructive criticism u say? hahaha, please give me a break!

Constructive criticism in NT world means just that "Oh, nice work guys, u r awesome, brilliant, .. etc.", how many bugs and features got ignored by some arrogant developers in their team through the years, what they did is to add half-assed features that is useless, instead of fixing up the architecture of LW.

Now some will say : "oooo, fixing architecture needs a lot of time, blardy, blardy, baaaarrr ... etc.", well they've been from 2008 till now, and HV, FFX .. not useable and lame ...

prometheus
03-08-2012, 01:11 AM
I can see one interesting bundle package thou if they might collaborate and can bundle with lightwave upgrade sales or similar.

3d coat and volumedic in a bundle, but I am thinking of them two working together, so you can paint or sculpt in 3d coat your mesh and use volumedic to render and slice the mesh etc.

Michael

bazsa73
03-08-2012, 01:11 AM
Let's hate non-LW users!!! :D

prometheus
03-08-2012, 01:13 AM
Constructive criticism u say? hahaha, please give me a break!

Constructive criticism in NT world means just that "Oh, nice work guys, u r awesome, brilliant, .. etc.", how many bugs and features got ignored by some arrogant developers in their team through the years, what they did is to add half-assed features that is useless, instead of fixing up the architecture of LW.

Now some will say : "oooo, fixing architecture needs a lot of time, blardy, blardy, baaaarrr ... etc.", well they've been from 2008 till now, and HV, FFX .. not useable and lame ...

That comment is very constructive indeed...emotional and with hard facts.., donīt think you would get any point through except venting your frustration.

realgray
03-08-2012, 01:27 AM
I love LW. But I will say it will be interesting to see what LW 12 will offer at 695 for most users vs. Modo 701 at 495. All I can say is I hope Newtek pulls off something great.

DigitalSorcery8
03-08-2012, 01:44 AM
That comment is very constructive indeed...emotional and with hard facts.., donīt think you would get any point through except venting your frustration.

But unfortunately not too far off the mark. This is why T4D left LW for SI - MANY items in LW not addressed for years and years. Many requests essentially ignored to add features never completely integrated - such as FFX. We'll just have to wait and see now if the LW team can provide the service pack for LW11 and then tell us what's in store for LW12. We'll see if they begin to fix the underlying problems from years ago. I'm hoping that they do. :thumbsup:

erikals
03-08-2012, 01:59 AM
NT was never focused on improving the VFX part.

will it change? no idea...

geo_n
03-08-2012, 02:07 AM
Constructive criticism in NT world means just that "Oh, nice work guys, u r awesome, brilliant, .. etc.",

Agree. There's too many "Yes men" surrounding NT. All the kissing and hugging isn't helping lw improve. Missing features need to be followed up and a roadmap needs to be presented.

50one
03-08-2012, 02:39 AM
NT was never focused on improving the VFX part.

will it change? no idea...

Well, I think you have answered the question yourself..

Anyway, why re-invent the wheel? There are packages out there that are production proven(Houdini, TP, Krakatoa, FumeFX) and Newtek won't be able to catchup with them anytime soon, so it always gonna stay as a 'niche' CG tool, no hatin' on my part, just being reasonable here.
Unless substantial cash-flow will be injected into development(more devs, new technologies)


I know that lots of folks compare the LW with Modo and are like, "oh look where are they now...", well it's true - they've done a tremendous work on Modo and the speed of adding the features and improving the workflow is great, but, don't forget you've got folks that created the original renderer for LW, the IKBooster, Hypervoxels. So is it really that innovative or rather "Learning on your past mistakes" + adding some new stuff?:)


To be honest I would give the modeler for free in the state it is now and focus on rewriting Layout into Core(which is what they're doing atm perhaps:))

I've tested the volumetrics in Modo and although they're great - my machine is not up-to-date with the hardware so rendering can take a lot of time, even tho the GI / modeling is quite snappy.

I'm going to decide whether I'll go with C4D at the end of the month as the trial is running out and I really want to move into Motion graphics(If not the price - I would do it earlier on), although I don't like the modeling tools in there but that's why I've got the Modo for, anyway would be great to have some of the controls in LW similar to Mograph module/Xpresso / TP / Xref , yeah I know - DPKit - but I'm designer, even tho I've got degree in Software engineering I'm sometimes struggling when operating on vectors / scalars / normals :D

erikals
03-08-2012, 03:11 AM
Well, I think you have answered the question yourself...

i said "was"... :]
hey, who knows, 11 had a bit of it... \:°

prometheus
03-08-2012, 04:14 AM
This is an interesting volumedic video sample, not exactly converting geometry to volume, more loading volume data in to a container I suspect, but cool cloud models might be done this way, that hvīs canīt.
http://www.youtube.com/watch?v=dkxhht8QkaU&feature=endscreen&NR=1

Michael

mav3rick
03-08-2012, 05:20 AM
Agree. There's too many "Yes men" surrounding NT. All the kissing and hugging isn't helping lw improve. Missing features need to be followed up and a roadmap needs to be presented.

i dont see any kissing and hugging regarding NT since Rob, Lino and Matt joined.. in fact quiet opposite. there are few of those that like to stress out on forums quiet often. I would say Dr. House could make a good season about them..
Rob, Lino and Matt are the best that have happen to LW in past decade.... some support could be of use for sure...
If they fail ... lw will fail so lets give some support to those guys and at least write constructive critics and objective posts...


regarding modo.. i am legit upgraded to 601 and from what i checked into it... it is nice eye candy.. but not production ready for anything beyond modeling, painting, sculpting if so (crashing) imho. they are missing some good beta testers (hangs way too much) and shader tree is pain in arse.

prometheus
03-08-2012, 05:28 AM
Ranting about that Newtek not giving a notice to anything isnīt doing any good what so ever, you are entitled to have those feelings sure, but better of keep them to yourself if you are not specific on target..and if it is a lot, make a list of what you feel neglegted and post to the forums or the development team.

Giving critics in a specific area such as why hypervoxels been so neglected and what might be done to improve is way better than ranting, they do not listen or care.

I think Newtek has done a tremendous job for the latest release, not perfect and perhaps not as colorful as recent modo release, and I also miss lotīs of stuff from the core platform along with what is mentioned in this thread post, but perspective of things is important to think about, the lack of hv development and Improvement are something newtek hasnt done well, but other things are done very well....perspective again.

Michael

geo_n
03-08-2012, 05:57 AM
i dont see any kissing and hugging regarding NT since Rob, Lino and Matt joined..

I'm not saying its the nt dev. The yes men are the long time lw users. The people who think lw is ok at this point that it doesn't need anything at all so every lw release is super. Problem is when they realize how behind they are with the app it turns ugly and you have angry ex lwvers going on cgtalk, lux, etc bashing lw for everything.
Lw needs a lot of things and all the requests popping up are people's need, personal or whatever we all have different needs. Just like improved hypervoxels, improved modelling tools and speed, improved ca tools, muscles, renderpass management, integration with existing tech like ikinema, vray, joe hair and shave, cad loaders, etc. Someone not even interested with these features to post and say its not needed at all and lw is ok is just :thumbsup:.

lino.grandi
03-08-2012, 07:53 AM
LW 11 just loses to a simple instancing test.

Can you please explain what you mean?
What's so wrong with LightWave instancing according to you? :question:

Elmar Moelzer
03-08-2012, 08:21 AM
Aha..I missed that volumedic has a lite version
Only "lite" in price, not in features ;)

Can Modo do something like this?
http://www.youtube.com/watch?v=eHAe4DItmrU

This was literally made in a couple of minutes with two imagemaps, a procedural texture and VoluMedic CE.
http://www.youtube.com/watch?v=ghoW68OkfZE
http://www.youtube.com/watch?v=HSG_JKSLjmA

VoluMedic is good for A LOT more than just medical viz.
Btw, you can already convert any polygon object into a volume with VoluMedic and then convert it back to geometry again, if you want to.
You can also directly render any geometry as a volume. It currently an undocumented and somewhat experimental feature, but it works.
We will expand on that in later versions.

I also want to add that all of the screenrecordings were made on a laptop with a Core i7 820QM (1.73 Ghz), which is certainly not a high end machine anymore by todays standards. I bought mine used on Ebay more than 1.5 years ago for just 700 Euros.

Andy Meyer
03-08-2012, 08:30 AM
...and shader tree is pain in arse.
agreed, stupid sucker tree :agree:

Andy Meyer
03-08-2012, 08:42 AM
Elmar, if i dont need VoluMedic for medical stuff then CE version is all i need? or has the full version tools i need for non medical stuff?

lino.grandi
03-08-2012, 09:22 AM
Constructive criticism u say? hahaha, please give me a break!

Constructive criticism in NT world means just that "Oh, nice work guys, u r awesome, brilliant, .. etc.", how many bugs and features got ignored by some arrogant developers in their team through the years, what they did is to add half-assed features that is useless, instead of fixing up the architecture of LW.

I've been working with NewTek for almost 2 years now...and can't see any trace of any "arrogant developer" in the team.




Now some will say : "oooo, fixing architecture needs a lot of time, blardy, blardy, baaaarrr ... etc.", well they've been from 2008 till now, and HV, FFX .. not useable and lame ...

While I can agree that FIberFX needs work, I wouldn't define HV feature as un-usable for sure.
Anything can be improved, of course. Posts like this one are not helping so much anyway. ;)

Celshader
03-08-2012, 09:30 AM
anyway, it is still too much for a single feature, what do you get for the purchase of modo vs volumedic, you would end up using a tool mainly aimed for special medical & sience visualisation in order to get this object volumetric feature, and would you really need the other stuff, now modo would + that with a whole lot of other tools more full in features for the whole 3d range.

True, but I can't count on a studio owning the latest modo when I need it. These days I usually get hired as either a LightWave artist or a RealFlow artist. I've also been hired as a Poser artist (2007) and a messiah:studio rigger (2008), but not modo. The studios out here aren't looking for modo artists.

However, I suspect I will run into Volumedic CE on one of my future LightWave gigs, so I'm tempted to get a head start on learning it.


But neither volumedic or modo is what we should cry out loud for, it is Native lightwave and Itīs volumetric handler that needs attention

Sure. That's why I'm thinking of adding Volumedic CE alongside my Turbulence FD license. Both tools + the native Volumetric Node Editor should tide me over while the dev team sorts things out.

Elmar Moelzer
03-08-2012, 09:33 AM
Elmar, if i dont need VoluMedic for medical stuff then CE version is all i need? or has the full version tools i need for non medical stuff?
Andy, you can use both for both :)
VoluMedic CE lacks features that artists wont miss much, but that are essential for users that need to do repeatitive tasks as well as analysis and measurement, or that require the ultra fast rendering and GPU Raycasting the full version has. CE makes most sense with LW 10 and higher since you will have to more heavily rely on VPR (though we have ideas to give it an OpenGL preview as well).
Either way, you can check out the tutorials that I posted in the VoluMedic CE- thread here. These are all things that VoluMedic CE can do.

omichon
03-08-2012, 09:42 AM
Deleted

Matt
03-08-2012, 09:48 AM
Constructive criticism u say? hahaha, please give me a break!

Constructive criticism in NT world means just that "Oh, nice work guys, u r awesome, brilliant, .. etc.", how many bugs and features got ignored by some arrogant developers in their team through the years, what they did is to add half-assed features that is useless, instead of fixing up the architecture of LW.

Now some will say : "oooo, fixing architecture needs a lot of time, blardy, blardy, baaaarrr ... etc.", well they've been from 2008 till now, and HV, FFX .. not useable and lame ...

Thank you for your thoughts Jameel, we're always interested to hear where we can improve LightWave, I'll make sure I pass your concerns to the development team so we can do even better.

Elmar Moelzer
03-08-2012, 09:53 AM
RFRK mesh importer
What sort of mesh is that? Sorry, I am a bit clueless right now.

omichon
03-08-2012, 10:02 AM
What sort of mesh is that? Sorry, I am a bit clueless right now.

RFRK stands for Realflow RenderKit.
But it was more a joke due to some euphoria because of all the quality tools we got lately.


Edit : Sorry, it could be a little confusing since I deleted the post we are referring to. You are quite fast to reply Elmar ;)

Celshader
03-08-2012, 10:07 AM
Oliver beat me to it, but...


What sort of mesh is that? Sorry, I am a bit clueless right now.

RealFlow RenderKit (http://www.realflow.com/rf_renderkit.php). It builds meshes from RealFlow *.bin particle sequences on rendertime.

The advantage I've seen from RFRK is the massive savings in hard drive space. For my RealFlow->LW (http://vimeo.com/22542405) projects and RealFlow->Maya projects, I used RealFlow to generate mesh sequences from the *.bin particle sequences. The mesh sequences were then rendered in either LightWave, mental ray or Maxwell. It worked fine, but RealFlow-generated object sequences can take up a fair bit of space on the hard drive.

For the one RealFlow->Houdini (http://www.youtube.com/watch?v=TiNdoCEgBOk&hd=1) project I worked on, we used RFRK to generate meshes in the render engine on render time. Considering all the RealFlow *.bin sequences we calculated for that project nearly tipped over the server, RFRK must have spared a ton of hard drive space.

Elmar Moelzer
03-08-2012, 10:20 AM
Hmm, wouldnt it be better if we could somehow render realflow particles directly in VoluMedic? Or any particle system for that matter?
Not that I am announcing anything here, or so.

omichon
03-08-2012, 10:21 AM
Thanks Jen, for the clarification :)
It was probably completely off topic since I don't know the technology under the hood, but I see the possibilities of Volumedic...
Just a thought.

omichon
03-08-2012, 10:23 AM
Hmm, wouldnt it be better if we could somehow render realflow particles directly in VoluMedic? Or any particle system for that matter?
Not that I am announcing anything here, or so.

Well, that was my point, it's in your mind now ;)

prometheus
03-08-2012, 10:34 AM
Thank you for your thoughts Jameel, we're always interested to hear where we can improve LightWave, I'll make sure I pass your concerns to the development team so we can do even better.

you got to be kidding..no need for sarkasm Matt;)



Only "lite" in price, not in features ;)

Can Modo do something like this?
http://www.youtube.com/watch?v=eHAe4DItmrU

This was literally made in a couple of minutes with two imagemaps, a procedural texture and VoluMedic CE.
http://www.youtube.com/watch?v=ghoW68OkfZE
http://www.youtube.com/watch?v=HSG_JKSLjmA

VoluMedic is good for A LOT more than just medical viz.
Btw, you can already convert any polygon object into a volume with VoluMedic and then convert it back to geometry again, if you want to.
You can also directly render any geometry as a volume. It currently an undocumented and somewhat experimental feature, but it works.
We will expand on that in later versions.

I also want to add that all of the screenrecordings were made on a laptop with a Core i7 820QM (1.73 Ghz), which is certainly not a high end machine anymore by todays standards. I bought mine used on Ebay more than 1.5 years ago for just 700 Euros.

Great stuff elmar, do not know what to invest in now..turbulenceFD or volumedic, Yes the complex object with volumes is interesting and sweet.
Probably Turbulence first thou.

I would like to have a feature thou to lower the density on the other edge of a volume object and also introduce a noise or use procedural textures on the same outer rim of the object but try to keep the inside volume more intact.

By the way, I think I got censured here:devil:, this thread post was called Modo 601 new volumetric shader surpassed hypervoxels, and now it is called Request for Hypervoxels with comparitive note.

I guess that is The forum rules and moderators privligie thou, cant say anything about that..except for whining like a kid being moderated
:neener::

cheers
Michael

erikals
03-08-2012, 10:43 AM
Hmm, wouldnt it be better if we could somehow render realflow particles directly in VoluMedic? Or any particle system for that matter?
Not that I am announcing anything here, or so.

i was just thinking about that actually...
a problem though might be the blending mode.

http://forums.newtek.com/attachment.php?attachmentid=99476&d=1321408594

 

Celshader
03-08-2012, 10:48 AM
Hmm, wouldnt it be better if we could somehow render realflow particles directly in VoluMedic? Or any particle system for that matter?
Not that I am announcing anything here, or so.

That's what RFRK does right now for other packages. It would be a lovely addition to VoluMedic if VoluMedic can duplicate the look of RealFlow metaballs.

Right now artists can load RealFlow *.bin particles directly into Layout and render them with HyperVoxels. I've done this for mist and water bubbles.

I've also written tools at different studios to convert *.bin particle sequences into *.lwo partigon clouds with embedded speed endomorphs. 1-point partigons look great for whitewater FX.

However, most of my "water splash" RealFlow work involves RealFlow-generated mesh sequences because I really like the way RealFlow metaballs blend together. If VoluMedic could load RealFlow *.bin particle data and duplicate that look, that would be awesome.

erikals
03-08-2012, 10:55 AM
This is definitely impressive. Might save some people from buying ozone.
http://www.youtube.com/watch?v=-MXug_ZGUMY&feature=youtu.be

I'd say 2 years more and modo will be overlapping lightwave in most aspects at the pace these two are moving.

nice looking indeed, you might be able to do the same thing in Ogo Taiki. (haven't tried)
Taiki will definitely be faster though, and very tweak-friendly using VPR.
http://www.youtube.com/watch?v=asiypf9eIp4
 

Matt
03-08-2012, 11:08 AM
you got to be kidding..no need for sarkasm Matt;)

Actually, not being sarcastic at all, but I would rather be positive about this stuff than be all defeatist, I mean, how should I respond?

mav3rick
03-08-2012, 11:28 AM
Actually, not being sarcastic at all, but I would rather be positive about this stuff than be all defeatist, I mean, how should I respond?

hmm ... you shouldn't... why should you enter in debate on that level with guy that is simply venting on absolute non constructive way.. there is few more patrolling around forum and biaching just about anything that starts with L and end with W..

Celshader
03-08-2012, 11:50 AM
This is definitely impressive. Might save some people from buying ozone.
http://www.youtube.com/watch?v=-MXug_ZGUMY&feature=youtu.be

The camera in this video doesn't actually fly into the cloud at any point. This suggests that modo Clouds may have the same weakness as LightWave HyperVoxels.

Right now there's no render hit to fly into and through a LightWave Volumetric Light, but there is a render hit when flying into and through a LightWave HyperVoxel. I have to put a Distance to Camera gradient on the Dissolve property of HyperVoxels when they get too close to the camera.

prometheus
03-08-2012, 01:02 PM
starting to get of topic for the "improve hypervoxels"
But talking about a tool that is not underdeveloped even thou it isnīt newteks native tool, but something that probably will take some time before modo getīs it..

TurbulencFD.. check jawsets site a new release is up with a fireshader opengl preview.

When talking about clouds and flying through it and nicer cloud volumes, well you can always simulate a smoke cloud with turbulenceFD and use frame step of 0 and use a value of desired frame offset to have a static fluid cloud that you can fly through.

the old zitterhand has some examples of that on youtube.
TurbulenceFD and volumedic CE is two wonderful tools that modo doesnīt have a connection with.

Michael

prometheus
03-08-2012, 01:03 PM
Or use TFD. :D
http://youtu.be/NlnAVpcEXQ4 - no real rendertime hit. Placed the camera right inside a cloud to check exactly that. Well, the advantage of having only one voxel in space, whereas HVs' can overlap and take the rendertime through the roof...

Regarding rendering particles with volumedic, Realflow apparently uses more information than just the position to generate the meshes - otherwise it wouldn't take so long. The 'magic' why it doesn't look like

Haha..that was funny, just when I wrote the post beneath..you Beat me to it, funny we were writing this same topic at almost the same time.

Elmar Moelzer
03-09-2012, 07:42 AM
http://www.youtube.com/watch?v=E7oyHuX4A38&feature=youtu.be

I know it is not paricularily great, but I only had a few minutes to spare on this. It is really just the puffy clouds procedural combined with a gradient texture to fade them off towards the top ( to avoid a sharp terminator at the borders of the container).
I am sure that people could achieve much better results with more time and effort.
Rendertime 150 sec per frame on average (Core i7 820 QM with 1.73 Ghz), no rendertime increase when penetrating the clouds.

omichon
03-09-2012, 07:48 AM
Can't play the video here (The URL contained a malformed video ID).

prometheus
03-09-2012, 07:53 AM
Can't play the video here (The URL contained a malformed video ID).

Same error here.

Michael

prometheus
03-09-2012, 08:35 AM
Should be this link, I assume (from here (http://forums.newtek.com/showpost.php?p=1227242&postcount=21)): http://www.youtube.com/watch?v=E7oyHuX4A38 Interesting. :)

I thought that was turbulence first, but that is in fact volumedic since elmar did it I guess?

By the way oliver, you have done "freezed" cloud frames from turbulenceFD, I noticed that in the last version before the newly V1 release, Youīd have to set frame offset to a - value...like - 60 and no a positive value, as I recall it should be a positive value right?

Have to wait til I get home and check if that was changed with the V1 release.

My wish...well turbulenceFD and volumedic, wonderful, but if one would like to match sky realism we would also need a more physical sky and fog such as the ogo taiki environment could bring us, it wont really elevate to that excellent level of wonderful skies otherwise.
Extremly slow thou..for final renders even thou vpr might be quite faster.
http://www.asahi-net.or.jp/~pq1a-ogs/taiki_e.html

would be wonderful to use turbulenceFD cloud simulated frames or something from volumedic in conjuction with the air properties from ogo taiki, otherwise ogo taikiīs textured based volumetrics gives large scale clouds
that are much more realistic then setting hvs up.

Michael

Celshader
03-09-2012, 08:52 AM
Should be this link, I assume (from here (http://forums.newtek.com/showpost.php?p=1227242&postcount=21)): http://www.youtube.com/watch?v=E7oyHuX4A38 Interesting. :)

Dave's looking over my shoulder, and he thinks it looks cool. He'd love to see it with Denis Pontonnier's Gardner Clouds procedural texture.

He also says that "just for video alone," Volumedic CE is worth the $300.

erikals
03-09-2012, 08:59 AM
Ogo is really fast when using VPR for then to export it. http://www.youtube.com/watch?v=asiypf9eIp4

Volumedic CE is interesting though... as it uses another approach...

Elmar Moelzer
03-09-2012, 09:44 AM
Ok, fixed the link.
Should be working now.

Elmar Moelzer
03-09-2012, 09:51 AM
Thanks for the kind words, Jen.
Everyone, please understand that VoluMedic is not a cloud generator or a fire plugin or a smoke or explosion plugin. For fire and smoke T4D is probably more suitable (without having tried it). That said, I will be trying a few effects shots with VoluMedic over the coming days, if I find the time, besides all my other duties.

omichon
03-09-2012, 10:05 AM
With a more suitable procedural and like you said, more time spent on it, I am confident it could produce some really interesting results.
Volumedic seems to be more versatile that I could have imagine at first glance. Thanks Elmar :)

Elmar Moelzer
03-09-2012, 10:39 AM
I am trying dponts procedural right now and it makes for some realistic, but also very dense cloud covering. One would need to somehow limit it with another procedural texture or something like that. I am going to play a bit with it and see what I can come up with.

prometheus
03-09-2012, 10:50 AM
hereīs some images I did with Hvīs..rendertimes and voxel clouds to compare with..

http://forums.newtek.com/attachment.php?attachmentid=100331&d=1324210201

http://forums.newtek.com/attachment.php?attachmentid=100332&d=1324210212

http://forums.newtek.com/attachment.php?attachmentid=100182&d=1323633549

http://forums.newtek.com/attachment.php?attachmentid=100196&d=1323679963

Earth scene from space with HVvoxels for cloud..not quite realistic, huge rendertimes.
http://forums.newtek.com/attachment.php?attachmentid=100529&d=1324836732

check thread for rendertimes
http://forums.newtek.com/showthread.php?t=124411

Michael

Elmar Moelzer
03-09-2012, 11:18 AM
With VoluMedic it is more like sculpting the clouds out of a single block of voxels. Unlike it is with HVs, where you have some blobs that are then displaced.

prometheus
03-09-2012, 11:41 AM
With VoluMedic it is more like sculpting the clouds out of a single block of voxels. Unlike it is with HVs, where you have some blobs that are then displaced.

Indeed it is like that, and the big issue with HVīs wich caused me to post this very thread based on comparison to Modoīs new volumetric shader that appearently can be applied on full geometry and not only particles and vertices.

Running some test now with the latest TurbulenceFD to check cloud simulations..
Have a busy weekend with tests, got to test Houdini and Itīs volumetrics on geometry, and pyrofx.
Then I think Ill get to install volumedic Demo and test with a Fresh Lightwave 11.

I wonder If one could utilize 3d coat and sculpt voxel cloud geometry and possibilities to convert to voxels in lightwave?

Michael

Elmar Moelzer
03-09-2012, 12:41 PM
I have got to check with Wolfgang whether the VoluMedic CE Demo is fully tested and uploaded yet. VoluMedic 3.0 does not have the "Complex" Object type yet that is needed for clouds and stuff.
So you might have to hold back with that for another couple of days.

prometheus
03-09-2012, 12:48 PM
I have got to check with Wolfgang whether the VoluMedic CE Demo is fully tested and uploaded yet. VoluMedic 3.0 does not have the "Complex" Object type yet that is needed for clouds and stuff.
So you might have to hold back with that for another couple of days.

Great..thanks for the heads up, I got loads of things to do anyway.

Michael

Elmar Moelzer
03-09-2012, 03:18 PM
So the demo is seemingly up now. Why Wolfgang called it VoluMedic 35Cedemo.exe I dont know. It is only the demo for CE. The 3.5 demo will come when version 3.5 is ready next week.

Elmar Moelzer
03-09-2012, 08:05 PM
Here is another test. This is a few more texture layers, mostly of the crumple procedural...
I am not really sure what to do there, just toying arround. At 4am everything starts to look the same anyway ;)

papou
03-10-2012, 06:31 AM
nice thread.
Volumedic is cool. But if will be hard to limit the clouds size with it.
It's like doing clouds with spinning trick.
(i suggest to use dented procedural that can be scaled)
i hope i can see a realistic cloud done with Volumedic. I will buy it immediately.

I really hope you can dig the realflow way too.
That's the first thing i though when i see your Lw logo reconstruction vid.

sami
03-10-2012, 08:24 AM
Only "lite" in price, not in features ;)

Can Modo do something like this?
http://www.youtube.com/watch?v=eHAe4DItmrU

This was literally made in a couple of minutes with two imagemaps, a procedural texture and VoluMedic CE.
http://www.youtube.com/watch?v=ghoW68OkfZE
http://www.youtube.com/watch?v=HSG_JKSLjmA

VoluMedic is good for A LOT more than just medical viz.
Btw, you can already convert any polygon object into a volume with VoluMedic and then convert it back to geometry again, if you want to.
You can also directly render any geometry as a volume. It currently an undocumented and somewhat experimental feature, but it works.
We will expand on that in later versions.

I also want to add that all of the screenrecordings were made on a laptop with a Core i7 820QM (1.73 Ghz), which is certainly not a high end machine anymore by todays standards. I bought mine used on Ebay more than 1.5 years ago for just 700 Euros.

Those videos are amazing! If I had known Volumedic was that flexible and easy to use, I can see lots of uses for it besides medical (which I have yet to have a need for). Does the turbulent logo in this video cast appropriate shadows and reflections? And how does it handle UV maps for other volumes?

Sorry if this is hijacking the thread. Just curious...

Celshader
03-10-2012, 08:39 AM
i hope i can see a realistic cloud done with Volumedic. I will buy it immediately.

Dave Jerrard now owns a copy of Volumedic CE, so it's only a matter of time. ;)

In the past, Dave would create clouds for Dogfights by using Denis Pontonnier's Gardner Clouds texture on a single large Hypervoxel. Sometimes he'd use Beer shading, but usually he used Rayleigh shading. This was fine for flyovers and gliding close to the cloud, as long as the camera did not fly into the HyperVoxel itself. Volumedic CE has him pretty excited because there's no render hit for flying into the volumetric.

Elmar Moelzer
03-10-2012, 09:31 AM
Well, you can always paint a 3d- texture that acts as an alpha map. VoluMedic offers tools for that too.

Elmar Moelzer
03-10-2012, 09:39 AM
Does the turbulent logo in this video cast appropriate shadows and reflections? And how does it handle UV maps for other volumes?
Very good questions!
Yes, it is compatible with all of LWs light types. It also shows up in raytraced reflections and also is capable of reflecting itself and the environment via raytracing (only in solid mode though).
There is currently no UVmapping for Volumetrics that I know about. I do have a couple of ideas for workarrounds though that we might be employing in the future.


Dave Jerrard now owns a copy of Volumedic CE, so it's only a matter of time.
Jen, I sent you an email, did you get that (the email adress that SH specified on the order). You should be able to DL VoluMedic already from the location specified in it.


Sometimes he'd use Beer shading, but usually he used Rayleigh shading.
This is a very curious thing. In VoluMedic we have a parameter "shading softness", that I found can emulate most of the shading of HVs. You can use values above 100% as well. That is what I did for the cloud renderings that I posted here. It gives a really nice "SSS" look for skin, wax and clouds.

prometheus
03-10-2012, 02:42 PM
Dave Jerrard now owns a copy of Volumedic CE, so it's only a matter of time. ;)

In the past, Dave would create clouds for Dogfights by using Denis Pontonnier's Gardner Clouds texture on a single large Hypervoxel. Sometimes he'd use Beer shading, but usually he used Rayleigh shading. This was fine for flyovers and gliding close to the cloud, as long as the camera did not fly into the HyperVoxel itself. Volumedic CE has him pretty excited because there's no render hit for flying into the volumetric.

would like to see some sample of that, I wonder if the technique is the same standard that I used?

Michael

Celshader
03-10-2012, 03:15 PM
would like to see some sample of that, I wonder if the technique is the same standard that I used?

Michael

This shot uses a single HyperVoxel and the Turbulence texture for the clouds:
http://www.youtube.com/watch?v=dGUB9vbiRJw#t=16m16s

You can see the shadow of the farther plane as it enters the clouds.

Dave got better results with Denis Pontonnier's Gardner Clouds texture, however. He's looking for an example of that one right now, since he thinks the Turbulence cloud "kinda sucks." ;D

prometheus
03-10-2012, 03:29 PM
This shot uses a single HyperVoxel and the Turbulence texture for the clouds:
http://www.youtube.com/watch?v=dGUB9vbiRJw#t=16m16s

You can see the shadow of the farther plane as it enters the clouds.

Dave got better results with Denis Pontonnier's Gardner Clouds texture, however. He's looking for an example of that one right now, since he thinks the Turbulence cloud "kinda sucks." ;D


whereīs that in the clip 16m 16 s in or?

to bad the fire and smoke vfx isnīt the best in the show, compromized by the timeline they had to spit it out I guess, thatīs another story thou, the clouds looks a little better the little I can see of it.

Michael

prometheus
03-10-2012, 03:35 PM
here was my samples..posting links again, only stills...some different styles and techniques, some is using subpatch grids to control density of the voxels, and some is just two or three nulls, mixture of dented, fbm an gardner clouds.
beer and rayleigh illumination mostly used...some fog to blend clouds.

Blahh...wrong linking, check previous page and posts from me.


Michael

prometheus
03-10-2012, 03:44 PM
hereīs my sample images..
Even one scene trying to replicate the aeroplane going down from superman returns...have to revisit that one thou:)

Michael

Celshader
03-10-2012, 04:25 PM
whereīs that in the clip 16m 16 s in or?

The shadow of the far plane gets cast upon the HyperVoxel cloud around 16m 22s:
http://www.youtube.com/watch?v=dGUB9vbiRJw#t=16m21s

Celshader
03-10-2012, 04:28 PM
hereīs my sample images..
Even one scene trying to replicate the aeroplane going down from superman returns...have to revisit that one thou:)

Michael

Thank you for sharing these. :)

prometheus
03-10-2012, 04:31 PM
The shadow of the far plane gets cast upon the HyperVoxel cloud around 16m 22s:
http://www.youtube.com/watch?v=dGUB9vbiRJw#t=16m21s

Yeah I finally noticed, Jerrard also just put another clip up on his channel, so I see the gardner clouds better, reminds about what I also did.
Heres Jerrards clip...
http://www.youtube.com/watch?v=_NIOIPzukq4&context=C4d5eeacADvjVQa1PpcFOmHfpMf9WhxnXuSIBnGQtN jFex5cFFVls=


Right now I am Fooling around with TurbulenceFd and still frame simulations for clouds..might post it later on vimeo.
donīt know if to use subgrid detail (slows down render) or use higher resolution in the grid.

another thing except for the full geometry volumetrics like modo now has, I can think of better ways to mix hypertextures depending on location,node connections or geometry shape, so we can have whispy turbulence at the rim of the cloud object but retain a type of billow,perlin gardner cloud puffs inside.
Currently we can only do mixes in the node editor but not the same as I would want.


Michael

prometheus
03-11-2012, 05:49 AM
First tests with clouds from fluids and TurbulenceFD..
http://vimeo.com/38295925

drawback would be that you do not quite know how the simulation will be, and you cant really design the shapes as with hypervoxels or perhaps volumedic.
But it renders pretty fast compared to HV

Using smoke shader and activated the opacity channel beneath that.

some slight subgrid detail in there too.

A little to small cloud fluid emitter plane perhaps, only 300 meters.

Michael