PDA

View Full Version : LightWave 9.0 Feature info please...



spirit_of_stars
10-31-2005, 09:03 AM
Anyone have precision on this LightWave 9.0 Feature Preview?

New Volumetric Shader :
Allows for fast rendering of large numbers (order of 1 million) of simple volumetric particles.

What is this?

(check at http://www.newtek.com/lightwave/lw9featureprelim.php)

Dodgy
10-31-2005, 10:08 AM
I speculated that it could be used, along with faster more accurate particle collisions (also mentioned) to implement a fluid dynamics system... but that's just speculation on my part :)

spirit_of_stars
10-31-2005, 10:26 AM
Possible... thank you

Only someone from newtek could awnser that...

Captain Obvious
10-31-2005, 12:03 PM
My guess is that it's a way of rendering volumetrics using vast amounts of tiny polygon-based particles. Since each individual particle is exactly the same mesh, it doesn't use an awful lot of memory. Lightwave only needs to consider the meshes' individual positions, and I'm sure LW handles vast amounts of nulls (for example) without any real problems.

RedBull
10-31-2005, 12:23 PM
This is one of the many plugins bought from TLU and Jarno...
I believe...

It's a volumetric particle shading system, like HV only faster..... :)
Together with the Waterpool plugin, HV should be much cooler in LW9..
And fluid simulations should be much easier in LW9 with those...

3dworks
11-01-2005, 03:11 AM
Anyone have precision on this LightWave 9.0 Feature Preview?

New Volumetric Shader :
Allows for fast rendering of large numbers (order of 1 million) of simple volumetric particles.

What is this?

(check at http://www.newtek.com/lightwave/lw9featureprelim.php)

could be a new instancing system... they also announced advanced integration with the eco system of vue - so maybe it's related to this.

Red_Oddity
11-01-2005, 05:15 AM
Or it might be something like Maya's HW renderer, but for particles...

lardbros
11-01-2005, 12:17 PM
Well, since there is a limit of around 100,000 or something (cant remember the figure) hypervoxel particles, then i would assume that this means LW will be able to render many many more, which can only be a good thing.

Hopefully nice water will be possible with many more particles and faster calculations?

DMarkwick
11-02-2005, 07:12 AM
Its probably going to have many uses, like HV do now. So maybe we can look forward to better, more detailed, animatable & faster clouds etc.

loki74
11-02-2005, 07:01 PM
Actually, CFD clouds is very do-able... I mean, I've heard of creating a N-S equation solver for 2D gaseous flows in little over 100 lines of C code. Granted, it is a very simple solver, expanding it to interact with other objects and forces and such, as well as making it 3D would be (I imagine) quite simple for NewTek. Also, making it work with liquids and not just gases shouldn't be overly complex either. Combine that with HV's, and I think we'll have some pretty nice SPH going on! ...that would be SO sweet! (btw, all this stuff is what I read in a paper by Jos Stam, who works for Alias and apparently had a hand in developing Maya's FluidEffects)

lardbros--nice HV water is actually possible now... well, somewhat nice. I've gotten it to be a smooth pour and not that clumpy look one expects with HVs. My main thing is the way it collides with stuff--not very much like liquds.

RedBull
11-02-2005, 07:31 PM
Actually, CFD clouds is very do-able... I mean, I've heard of -

SNIP -

(I imagine) quite simple for NewTek. Also, making it work with liquids and not just gases shouldn't be overly complex either.



You've heard of, and you imagine, that writing CFD into LW is easy after reading a 5 page .pdf....... LoL Sorry to take the piss, but that's funny.... :D

100 codes of C..... :)
Just an example the "Hello World.c" sample is 104 lines (with comments)
I have read Joe's papers, and followed his work, and he's a genius....
thats why it took Alias a **** load of money, a huge physics team, many years to develop and intergrate to Maya FluidFX...

And it would be literally Tens of thousands lines of code.....
And it would be very hard to do accurate realtime versions in 3D....

So don't go holding your breath, based on Iv'e heard, and I imagine.....:)
You will end up very breathless.....

I will say, thanks to Jarno and TLU..... LW9 will offer a slew of cool new features, and stuff like this shader and HV deformer like waterpool,
indeed this will allow far more believable HV fx than LW8's does.

Having said that.... I'm aware of a FluidDynamics plugin, being developed
by a very smart LW programmer, and an another well known LWer......
All i will say is, you will of modeled with at least one of his tools...
And i have no idea, how far along and when it's due.... (So don't ask! ;)
It's going to be closer to GLU3D, than MayaFLuids...

So really life in this area, is looking better than it has in sometime.. :)

Jarno
11-02-2005, 10:15 PM
Sure, a grid-based 2D Navier Stokes solver isn't that difficult and can be done in a few hundred lines. But...

It won't do any interaction with other scene elements. It will only be 2D, not 3D. It will be limited to a rectangular grid. It will be low resolution or slow. It will look okay, but not great. It won't do splashing or breaking waves. And extending it to 3D will be a no-go for memory and time reasons.

There is a reason that products like RealFlow cost quite a bit of money, and big studios don't go giving away their internally produced CFD tools.

---JvdL---

loki74
11-02-2005, 11:19 PM
note: I never said easy. I said do-able. Unless you mean to tell me that it is literally inconcievable for a CFD to be written for LightWave... frankly, we are not in disagreement. If a good implementation were to happen, obviously it would be a massive amount of coding. I mean, lets look at the facts--thats 100 lines of code for a 2d solver, sans any interaction with anything other than the grid boundaries. So really by the time you factor in ALL of the possible things that would happen, and extend it to 3D, we're looking at a huge amount of code. There is no debate about that.

I suppose my point was, up until I read that paper, I had the notion that even a simple 2D sover with no interaction with other objects would itself be a massive amount of code. So in relative terms, you see why this would be exciting, or at the very least, revealing to me.

======

I was unaware of a CFD plugin being written for LW. (I'm assuming this isnt the one that showed up on SpinQuad a whiles back, because from what I've read, the chances of the LW community or NewTek getting that product are highly unlikely. I believe this developers SQ nick is Mikado or something similar) Would you mind posting up some details?

lardbros
11-03-2005, 03:41 AM
lardbros--nice HV water is actually possible now... well, somewhat nice. I've gotten it to be a smooth pour and not that clumpy look one expects with HVs. My main thing is the way it collides with stuff--not very much like liquds.

What i mean is that you need sh*t loads of particles to get nice flowing water... this means that the HV render limit won't be low anymore, so more particles will be able to render. Before i've always had to increase the size of the HV because my LW won't render them when you need more because they are small.

I agree, it's possible definitely, just looks nicer with MANY more and smaller particles. :D It's true, the collision calculations are what ruin LW's chances at more extreme water effects, hopefully this is being seen to as we speak?

RedBull
11-03-2005, 12:29 PM
I was unaware of a CFD plugin being written for LW. (I'm assuming this isnt the one that showed up on SpinQuad a whiles back, because from what I've read, the chances of the LW community or NewTek getting that product are highly unlikely. I believe this developers SQ nick is Mikado or something similar) Would you mind posting up some details?

It's okay Loki74, i didn't mean any offence...... It was just kind of funny,
i also used to hope and imagine about 3D CFD, (than i started programming) :)

I can't mention any details, as i'm really not privy to many of them,
only to say, that this programmer sold, some plugin(s) to Newtek for LW8.
And is now producing a plugin, based on CFD for Lightwave, with a well known and respected LW Artist..... I'm not sure of the progress. or when it's due for completion...... I get to thinking that it will be similar in performance to Glu3D.

Besides it's a bit rude of me, to announce other peoples plugins, so i will let them do it, when it's a little closer to release... I'm sure it will make a big splash, i respect this programmers maths skills very much.....

erikals
11-03-2005, 12:45 PM
Wow, now that would be cool.. :jam:

loki74
11-03-2005, 06:43 PM
It's okay Loki74, i didn't mean any offence...... It was just kind of funny,
i also used to hope and imagine about 3D CFD, (than i started programming) :)

I can't mention any details, as i'm really not privy to many of them,
only to say, that this programmer sold, some plugin(s) to Newtek for LW8.
And is now producing a plugin, based on CFD for Lightwave, with a well known and respected LW Artist..... I'm not sure of the progress. or when it's due for completion...... I get to thinking that it will be similar in performance to Glu3D.

Besides it's a bit rude of me, to announce other peoples plugins, so i will let them do it, when it's a little closer to release... I'm sure it will make a big splash, i respect this programmers maths skills very much.....

haha dont worry, no offense taken. I just wanted to make sure nobody was of the impression that I was some unnecessarily and unrealisticly ambitious kid who had dreams of glory and no knowledge or means of attaining that glory... thats always an embarassing place to be in, y'know?

That said, I will probably fiddle around a bit with the code Jos provided in that PDF, for the sake of becoming a better programmer if nothing else.

About this CFD plugin you speak of.. damm, now im all anxious to hear more!! lol. I hope it will be Mac-compatable! :thumbsup:

===

lardbros--yeah, that is most definately true. I hit the particle limit trying to get the pour look even vaguely laminar. Between this particle limit and the aforementioned CFD plugin, the future is looking very bright. I just wish the future would hurry up and get here!! :D

RedBull
11-03-2005, 07:50 PM
Yeah i know that place well..... I'm there often in fact! :)

Joe's code is fairly simple, and is fast and compiles without problems...
I had a look about a year back.... And it reminds me of the basics for particle ilussion, from wondertouch.... There is a fair bit of code on the net anyway..

Forgot to mention, if your looking for more code to look at with CFD
Blender 2.40 added a fluids option, similar to what i imagine the LW plugin,
i mentioned will implement.... Seems Blender favoured using the LBM methods.
This code should be far closer to what you would see in LW (be it in C++)
than what you will find publically by Joe...

http://www.blender3d.org/cms/Fluids_simulation.675.0.html
http://www.ntoken.com/p_fluid.html
White papers and source....

colkai
11-04-2005, 02:44 AM
Besides it's a bit rude of me, to announce other peoples plugins, so i will let them do it, when it's a little closer to release... I'm sure it will make a big splash, i respect this programmers maths skills very much.....
His alias doesn't start with an 'A' does it? I know of a certain math genius who is working on "something" at the moment. :p ;)

erikals
04-17-2009, 01:46 PM
I chat with Aurora a couple of days ago,
he says he is looking into making some CFD simulator, however, this thread was posted 4 years ago...

Redbull, do you know anything new?