[ShaidarHaran]

Originally Posted by doob

http://ompf.org/forum/viewtopic.php?f=3&p=8319

After reading the coments regarding the lack of a general cache in current gpu's and how it draws back the performance on sparse voxel octree ray casting it crossed my mind, would edram be a good solution here as a general cache?(slapped together as a parent die ala R500/C1)

If so how much would be minimaly usefull? 2, 4, 6MB?

Now there's an interesting idea. Makes me wonder if denser cache technologies such as ZRAM may be well-suited to this task, since EDRAM suffers from relatively poor density compared to say SRAM or ZRAM.

[Forrest]

Thread hijacked

[ShaidarHaran]

LOL, how so? We're still discussing ray-tracing... These sort of discussions have a tendency to branch off

yes, I went there

[Andrew Lauritzen]

Originally Posted by Forrest

Thread hijacked

Only if you admit that GPUs don't suck at raytracing... and you admit that there's nothing too special about it and its unsurprising that AMD could be doing some raytracing on the GPU considering quite a few such demos have already been shown in the past. [Don't get me wrong, it's still cool, just not that surprising.] All of the discussion pertains to these points, so let me know if you're convinced

[Jawed]

From the very sparse comments made during the presentation:

http://forum.beyond3d.com/showpost.p...postcount=4383


http://forum.beyond3d.com/showpost.p...postcount=4386

it sounds like id's technique is similar.

The Q&A section of the presentation:

http://www.bootdaily.com/index.php?o...k=view&id=1126

fills in some odd details. You're looking for the bits where the young Soprano is talking.

Jawed

[Forrest]

Originally Posted by Andrew Lauritzen

Only if you admit that GPUs don't suck at raytracing... and you admit that there's nothing too special about it and its unsurprising that AMD could be doing some raytracing on the GPU considering quite a few such demos have already been shown in the past. [Don't get me wrong, it's still cool, just not that surprising.] All of the discussion pertains to these points, so let me know if you're convinced

Thanks for the info. Hoping to see more cool stuff when larrabee comes out.

[Jawed]

Oh and a bit more:

http://www.youtube.com/watch?v=Bz7AukqqaDQ#

The Lightstage-captured woman is pretty impressive.

Jawed

[Andrew Lauritzen]

Originally Posted by Jawed

http://www.bootdaily.com/index.php?o...k=view&id=1126

Cool, thanks for the link!

But rasterization+raytracing hyrbid rendering is just *crazy-talk*! It'll never work...

[ShaidarHaran]

Originally Posted by Jawed

Oh and a bit more:

http://www.youtube.com/watch?v=Bz7AukqqaDQ#

The Lightstage-captured woman is pretty impressive.

Jawed

That was more than just "impressive", it was downright astounding!

[MfA]

If voxels do get popular people will figure out the same thing they did for them as they did for triangles, raytracing just doesn't make much sense for primary rays or shadow rays. It's easier to code, certainly with voxels ... but it has the same lousy locality of reference it has with triangles.

You can throw rays at the scene, or you can throw the scene at the rays ... triangles, voxels .... it doesn't matter.

[Bouncing Zabaglione Bros.]

Originally Posted by ShaidarHaran

That was more than just "impressive", it was downright astounding!

I was pretty stunned. When that section first came up, I thought it was supposed to be film of the actress doing the actual recording. When he said that this was the computer generated version, it was totally photorealistic (at least to the limits of the online movie). It will be interesting if they get the compression down as far as they hope, then you could really use this technique all over.

[Sxotty]

Originally Posted by nAo

Jon Olick will talk about id's raycasting on sparse octree tech at siggraph.

Awesome! I have been wondering what the deal is with it. Like I said before this then has the potential to make physics in games take a gigantic leap as well.

[catisfit]

Originally Posted by Jawed

http://www.youtube.com/watch?v=Bz7AukqqaDQ#

The Lightstage-captured woman is pretty impressive.

Originally Posted by ShaidarHaran

That was more than just "impressive", it was downright astounding!

Originally Posted by Bouncing Zabaglione Bros.

I was pretty stunned. When that section first came up, I thought it was supposed to be film of the actress doing the actual recording. When he said that this was the computer generated version, it was totally photorealistic (at least to the limits of the online movie).

Is there a decent quality video of that, or even some screenshots? It looks good in that youtube shaky-cam video (at 700Mb per frame, it ought to ) but I'd like to see it properly to judge how good it really looks.

[CNCAddict]

Originally Posted by MfA

raytracing just doesn't make much sense for primary rays or shadow rays

What does this mean wrt games and this new amazing visual technology that OTOY is showing off? From what I gathered, everything in the video is rendered and lit in realtime. However, if what you say has some impact on this technology being applied to interactive dynamic games, then it would be good to shoot my dreams down early before they climb to high

[Simon F]

Originally Posted by MfA

You can throw rays at the scene, or you can throw the scene at the rays ... triangles, voxels .... it doesn't matter.

Nice summary! Reminds me of Sutherland et al's classic paper.

[Andrew Lauritzen]

Originally Posted by MfA

You can throw rays at the scene, or you can throw the scene at the rays ... triangles, voxels .... it doesn't matter.

Yes indeed, which is why I'm confused that they're emphasizing the ray casting side of things rather than LOD, which is the real benefit of a (topologically limited/free) voxel representation.

But whatever, coherence/locality is swiftly becoming the only thing that matters in rendering and parallel computing, so I don't see rasterization-like techniques going away any time soon.

[MfA]

Originally Posted by CNCAddict

From what I gathered, everything in the video is rendered and lit in realtime.

Perhaps, or perhaps they bake in the lighting for the entire animation sequence (not in a pure color sense, but with say spherical harmonics). Can't rightly say at the moment.

[Jawed]

More information:

He summarises how he got to where he is now and focuses on the Transformers real-time trailers:

http://youtube.com/watch?v=ALD4NRR_DwY

This is a demonstration of rendering into a browser window. 3x RV770s doing 2 Transformers in a city scene with dynamic lighting computed in real time.

http://youtube.com/watch?v=iwTcvk5IuB4

I think these are polygon rather than voxel based.

This is the best Lightstage related talk:

http://youtube.com/watch?v=aCgeVDLpV6k

Jawed

[CNCAddict]

Originally Posted by MfA

Perhaps, or perhaps they bake in the lighting for the entire animation sequence (not in a pure color sense, but with say spherical harmonics). Can't rightly say at the moment.

In one of the videos, Jules says they can pop a spot light in at will and that each of the 100 lights in the building is a real light source. I'm just trying to find the truth behind all the glitz. I guess we'll know more when/if they release the new Ruby demo and also when Id talks about the voxel rendering at siggraph.

P.S. Not sure if I missed this, but it seems like good info http://anteru.net/2008/07/25/242/

[MfA]

Originally Posted by CNCAddict

In one of the videos, Jules says they can pop a spot light in at will

At will is fine, but is it interactive and realtime?

Here's a nice thread on the topic at hand.

[CNCAddict]

Originally Posted by MfA

At will is fine, but is it interactive and realtime?

Here's a nice thread on the topic at hand.

http://www.youtube.com/watch?v=iwTcv...eature=related

"no baked lighting, nothing's precomputed" minute 1:14
"100's of lights in the building all being rendered in real time, calculated in real time"

minutes 2:25 he puts a spotlight at the camera location and the characters shadow is on the opposing building


Thanks for the link also, I'm reading it right now