Just to make sure there is no confusion. What's shown on the Dynamite website are two different technologies.
The first few snapshots/animation relate to "grid-based" fluid simulation. No particles are involved, the simulation is applied on a 3D voxelized space, each voxel having it's density of smoke or other fluid. The advantage of such a solution is that it gives quite good results when simulating realistic fluid motion. The disadvantages are that it slows down to a crawl once you bring up the resolution to production quality. It also has much less control than particle-based methods and the simulation is confined to a small bounding-box space.
The last four smoke renders (after the head renders) are particles rendered with a volumic shader. Particles are simulated using a standard particle system, so it's much faster. Other advantages are more control over the simulation making achieve a wider range of effects including non physically realistic effects. Also, the simulation is not bound to a volume. However, although a rendered still image looks very nice, the fluid movement is not as fine as with grid-based simulations but still looks quite nice when tweaked right. This is the solution you find in Lightwave and plug-ins such as Binary Alchemy.
Both can be used depending on the effect seeked, the volumic shader approach is most popular partly because it is computes in reasonable time. There has been a lot of research on "grid-based" fluid simulations in the last few years. A good part of it is focused on making the simulation faster. So there is hope for the first approach to be more widespread.
.. Ronald
-----Original Message-----
From: owner-xsi(at)Softimage.COM [mailto:owner-xsi(at)Softimage.COM]On Behalf
Of Marc-Andre Carbonneau
Posted At: Tuesday, June 13, 2006 11:54 AM
Posted To: xsi
Conversation: Re: Re: [OT] dynamite v1.1
Subject: RE: Re: Re: [OT] dynamite v1.1
We use Lightwave once in a while.
The fireballs in Prince of Persia 3 opening sequence are done with
Lightwave.
Very nice voxels. You can do so much with basically just one shader.
-----Original Message-----
From: owner-xsi(at)Softimage.COM [mailto:owner-xsi(at)Softimage.COM] On Behalf
Of Kris Rivel
Sent: June 13, 2006 11:24 AM
To: XSI(at)Softimage.COM
Subject: RE: Re: Re: [OT] dynamite v1.1
I find it even stranger that this kind of stuff is made for Lightwave
initially. Are there that many high-end places using Lightwave heavily
these days? For the past year or two I get the feeling that most places
are using Maya, XSI or Max. Since Maya and Max has such a huge market
share, you would think the first gen versions of these plug-ins would
come out for those apps first. I guess whoever wrote this suite of
plugins was probably just heavily invested in Lightwave.
Kris
Is there something in LW that makes it suitable for voxels? Seems
they've always had that sort of stuff while almost no other apps have
something similar, even 3rd-party plugins!
On 6/13/06, Marc-Andre Carbonneau <marc-andre.carbonneau(at)ubisoft.com>
wrote:
Forget it. Get Fedkiw to do something wecan use.
He just posted new stuff for this year'sSiggraph
http://graphics.stanford.edu/~fedkiw/
From:owner-xsi(at)Softimage.COM [mailto:owner-xsi(at)Softimage.COM] On Behalf
Of Vince Fortin
Sent: June 13, 2006 9:23 AM
To: XSI(at)Softimage.COM
Subject: Re: Re: [OT] dynamitev1.1
Then again, Realflow ismore for liquids. Sure it has limitations but
it's still the only one of itscategory available on the market.
About grid-based fluids, I don't believe they are the holy grail of fx
but,without adding more fuel to the fire, you got to start somewhere to
improve thesystem. I know some of us on this list had them for XSI. We
got oursimplemented in literally no time with some features even
surpassing the Mayaones.
The point is, it's no big deal now to program them (I find it harder to
justifythe need to have them in XSI vs. buying a Maya license).
The good news is there's more and more alternatives around and
inspiration forsomeone who'd like to take the challenge to program them.
I don't expect Softimage to become a leader in the world of simulation
(eventhough I seriously wish they were) and I hope they come up with a
nicerparticle system way before a CFD one!
Vincent
On 6/13/06, kimaldis <kim(at)aldis.org.uk>wrote:
From: owner-xsi(at)Softimage.COM[mailto:owner-xsi(at)Softimage.COM] On Behalf
Of Votch
Sent: 13 June 2006 08:25
To: XSI(at)Softimage.COM
Subject: [Norton AntiSpam] Re:[OT] dynamite v1.1
Realflow may be great for viscous fluids like water
[kim aldis]
I wouldn'texactly call it great. You can force it but there's limits to
what it'll do andif you push it more than a bit it starts to creak at
the edges. It's also verylimited in terms of texturing. And you can't
get it to foam which is quite alimitation.
but it's lacking when it comes to gases, fireand clouds. With a bit of
effort XSI's particle system can produce nice workbut it's completely
lack of a voxel based CFD solver pushes it way behind thecurve. Just
about every competing package has a CFD solver either natively orvia
plugins. I'm really hoping Flowline or thinking particles makes it's way
toXSI sometime soon.
--
stockholm.postproduction | www.stopost.se
---
Unsubscribe? Mail Majordomo(at)Softimage.COM with the following text in
body:
unsubscribe xsi
---
Unsubscribe? Mail Majordomo(at)Softimage.COM with the following text in body:
unsubscribe xsi
---
Unsubscribe? Mail Majordomo(at)Softimage.COM with the following text in body:
unsubscribe xsi
