Volume 5 - VDB Creation - LW 2019

[gdkeast]

FWIW, Adam Gibson over at Liberty 3D has released Vol 5, VDB Creation, in his learning LightWave 2019 series. When I was getting back into LW, I found his videos helpful. I try to support those who are putting out LW educational content as much as I can.

[Markc]

Just saw the email for this.
Will probably pick this up, looks similar in output to TFD (fire example).

[Qexit]

Looks interesting and for a good price for over 3 hours of tutorials, it even includes a free Dilophosuarus Dinosaur Model. This link will take you to his page on Liberty 3D:

https://www.liberty3d.com/citizens/adam-gibson/

[Markc]

Hopefully Adam’s videos will explain things......
Tried to open the vdb explosion scenes included in lw content, keep getting error ‘load .vdb scene’, select it and same message. Both in 2018 and 2019.

[prometheus]

Just saw the email for this.
Will probably pick this up, looks similar in output to TFD (fire example).

You can do a lot of similar stuff, so in output yes...but VDB simulation is in my experience too slow compared to turbulenceFD or just doing a sim in blender and import as VDB.
VDB and gas solver is interesting...but I think they need to do something about the calculation time for it to really kick off.

That said...it has probably little to do with the tutorial from Adam...unless he has some demonstration on how to simulate it faster but still remaining decent voxel representation, without sacrificing the quality too much.

[gar26lw]

is there one for the blender/lw combo you mention?

[prometheus]

is there one for the blender/lw combo you mention?

Is there what?
tutorial for blender fire and smoke fluids to Lightwave VDB? in such case I don´t know if Adam covers that?

It´s not that hard to get blender fire smoke in to Lightwave, you´ll just load the vdb sequences created in blender (if cache is set to vdb that is) and fiddle with it, the difficulties is mostly
about getting good quality, which often means initial high quality simulations in blender I think, there is only three modes of interpolation in Lightwave where box and quadratic yields the best result...but I am not sure
any of the new vdb filter tools can do any magic, I think I tried by trial and error and it just crashed a lot...which could be a cause of me not understanding how to use them properly, If they by any chance is supposed to work
to smooth vdb files at all.

Setting up the fire, flame and density channels requires a bit of work, but not as hard as getting good quality I think.

Houdini has a smoothing vdb option node that you put in the output to make it more smoother..but that´s houdini.

Sample first try of making the bullet fracture and collision explosion In lightwave, sending to blender as mdd files and simulating the fire and smoke on to those pieces..then back again to lightwave as vdb and render in Lightwave.

I used weight paint in blender around the breaking aream, and used that as fluid fire and smoke emission density...It´s not possible with Lightwave and turbulenceFD.

Lightwave needs a native Layout weight paint function...and TFD needs a way to implement weight emission.


https://vimeo.com/302915512

[stevecullum]

...
I used weight paint in blender around the breaking aream, and used that as fluid fire and smoke emission density...It´s not possible with Lightwave and turbulenceFD...

https://vimeo.com/302915512

Although turbulence can't use weight maps, it can generate smoke from particles. You could try to use some of the new mesh info nodes to calculate the distance from the base points to the deforming points of a fracturing object as the 'T' value of a particle emitter. Where its greater than 0, start emitting particles. Just a thought based experiment, so would need to try it out to see if it would work.

[prometheus]

Although turbulence can't use weight maps, it can generate smoke from particles. You could try to use some of the new mesh info nodes to calculate the distance from the base points to the deforming points of a fracturing object as the 'T' value of a particle emitter. Where its greater than 0, start emitting particles. Just a thought based experiment, so would need to try it out to see if it would work.

I haven´t even installed turbulenceFD in 2019, not sure I am up to it since I think I prefer blender fluid and smoke..and possibly using vdb for rendering..If I would like to render in Lightwave that is.
As for using particles...interesting advice..which I do know of, but I fear that wouldn´t yield the same fluid emission as a weight map, it will have to be adapted to the particle size, velocity etc...and it is ovecomplicated to go with that workflow in my opinion, you have to wrestle with so much more parameters including particle dynamics.

Bryphi77 used some nodal tricks to have particles emitt from weigth maps in Lightwave..I haven´t been looking in to that though.

But..it ultimately comes down to workflow ease...just being able to paint where you want your fire and smoke to emitt from, in blender it´s a breeze.


I did a test with clothfx in lightwave on a curtain a while ago, simulating it and save out mdd, send to blender and paint on the cloth.then used fluid weight paint emission...and the fire and smoke follows the deformation nicely.


This is particle emission from weight paint, but you just do the same with fluids and select the vertex map in the fluid emitter settings instead to get fire and smoke from the weight map.
I am waiting eagerly to have this function of weight paint "natively" in layout...and emission options for both particles and fluids.

[sadkkf]

Just wondering if anyone knows what kind of quality we can expect from VDB fire and smoke. The screenshots on the tutorial page don't look nearly as photorealistic as what TFD can produce, but then again, the Liberty site is kind of rubbish and doesn't really showcase their products as well as it could.

[prometheus]

Just wondering if anyone knows what kind of quality we can expect from VDB fire and smoke. The screenshots on the tutorial page don't look nearly as photorealistic as what TFD can produce, but then again, the Liberty site is kind of rubbish and doesn't really showcase their products as well as it could.

First..the liberty site is not kind of rubbish, it fills an important role, I do agree however...they should showcase their tutorials with better "trailers showcasing the end results" and previews of various stages in the tutorials..that would be a great way to sell it better.
It requires additional amount of effort to include that for every tutorioal and a firm believe that would actually help sell their product better...how they see it I do not know.

As far as quality in VDB, mostly a resolution thing I suppose, I do not thing there´s nothing that prevents a VDB file to have good enough resolution for final end results.
Then the shading, while the TFD curve controlls offer a unique way to style the shading to your hearts desire, the volumetric shading isn´t PBR...and I believe that in VDB format and With pbr volumetrics..the result should be able to reach a higher level of realism, it´s just that it takes more work to tweak the vdb shading than in the TFD interface.

Edit...
The screenshots they put up on the site does have very low resolution..it doesn´t look good to present something at that level of comression, they should also consider to change images to png instead when it comes to screenshots of any software UI with text and values...that it isn´t may account for some part of it not looking good, but the actual simulation and vpr fire isn´t that much different in look from TFD, except for that it hasn´t got any curve shading going on.

[sadkkf]

First..the liberty site is not kind of rubbish, it fills an important role, I do agree however...they should showcase their tutorials with better "trailers showcasing the end results" and previews of various stages in the tutorials..that would be a great way to sell it better.
It requires additional amount of effort to include that for every tutorioal and a firm believe that would actually help sell their product better...how they see it I do not know.

As far as quality in VDB, mostly a resolution thing I suppose, I do not thing there´s nothing that prevents a VDB file to have good enough resolution for final end results.
Then the shading, while the TFD curve controlls offer a unique way to style the shading to your hearts desire, the volumetric shading isn´t PBR...and I believe that in VDB format and With pbr volumetrics..the result should be able to reach a higher level of realism, it´s just that it takes more work to tweak the vdb shading than in the TFD interface.

Edit...
The screenshots they put up on the site does have very low resolution..it doesn´t look good to present something at that level of comression, they should also consider to change images to png instead when it comes to screenshots of any software UI with text and values...that it isn´t may account for some part of it not looking good, but the actual simulation and vpr fire isn´t that much different in look from TFD, except for that it hasn´t got any curve shading going on.

Please don't misunderstand me. I'm only saying the *design* is rubbish. The content and purpose it serves is not in question. The quality of their tutorials have always been superb in my experience and I never intended to imply anything else.

If their primary role is to sell video tutorials, I would expect higher quality screenshots at the very least. Video clips of their tutorials and/or of the end results would be very helpful. All the images should be National Geographic quality, IMO, if that's what they're selling.

I apologize for not being clearer.

If VDB can produce results similar to TFD, that would be great. I will have to add this video to my wishlist.

[prometheus]

sample of gas solver for cloud rise evolution, simulation driven by particle velocity together with gas solver, yielding soft whispy edges in a way that is truly hard to achieve if not almost impossible with fractal noise only.
Could have icreased quality to avoid noise.

Now imagine more advanced shading with included blackbody for fire and smoke and you will get an idea of what level it could reach.

[prometheus]

Please don't misunderstand me. I'm only saying the *design* is rubbish. The content and purpose it serves is not in question. The quality of their tutorials have always been superb in my experience and I never intended to imply anything else.

If their primary role is to sell video tutorials, I would expect higher quality screenshots at the very least. Video clips of their tutorials and/or of the end results would be very helpful. All the images should be National Geographic quality, IMO, if that's what they're selling.

I apologize for not being clearer.

If VDB can produce results similar to TFD, that would be great. I will have to add this video to my wishlist.

Yes...I think I understood your real meaning of rubbish, but you didn´t say the "design" is rubbish..you actually said the site, which could be interpreted as including the content, but I suspected you didnt mean that.
I agree..they need to present screenshots better, and include video previews.

Great you apologize for not being clearer.

the problem with the VDB as I feel right now, is simulation speed and setting things up, as well as VPR speed when navigating within a vdb file that has gas solvers on..very slow, it´s much faster to handle by saving the gas solver and vdb file out, and strip it down so it only contains the end result and load that to a new vdb file null container, much faster to deal with in VPR and rendering...but you loose all interactivity once that is done.

TFD is much faster on simulating and has much more features I think suited for fluid fire and smoke, easier to set up, but misses PBR volumetrics.

[sadkkf]

Yes...I think I understood your real meaning of rubbish, but you didn´t say the "design" is rubbish..you actually said the site, which could be interpreted as including the content, but I suspected you didnt mean that.
I agree..they need to present screenshots better, and include video previews.

Great you apologize for not being clearer.

the problem with the VDB as I feel right now, is simulation speed and setting things up, as well as VPR speed when navigating within a vdb file that has gas solvers on..very slow, it´s much faster to handle by saving the gas solver and vdb file out, and strip it down so it only contains the end result and load that to a new vdb file null container, much faster to deal with in VPR and rendering...but you loose all interactivity once that is done.

TFD is much faster on simulating and has much more features I think suited for fluid fire and smoke, easier to set up, but misses PBR volumetrics.

I know the Liberty 3D people work hard and I also know maintaining a site with this content is not easy so I don't want to disparage their efforts. I'm also a little more critical than most because my day job is designing and building websites and applications.

I'm glad you brought this up, too, because I don't want anyone else misunderstanding either.

I'm really starting to wonder about VDB vs TFD. I've gotten used to TFD and can make some nice sims with it, but also wonder about its future with LightWave. Is it worth digging into VDB after investing the time learning it and the money spent on tutorials?