HL2 AA issue a thing of the past?

[andypski]

I don't want you to think that I'm disagreeing with the basic statement, because I'm not. I was just stating that the example provided of Max Payne wasn't something in which I could see such artifacts. I have no reason to doubt Gabe or you...

That's ok - the artifacts in Max Payne were more subtle than I remembered them being. Like I said, it's probably dependent on the exact part of the game, and I may have seen a particularly good (or bad on this case) example of the problem. If you look at the 3DMark03 example that I gave you should see the problem much more clearly.

I'm still not clear on centroid texturing as defined by the API. I've read various reports that it isn't supported under DX9, and others that it is. Of course, if it isn't supported under DX9, my next question is obvious as to its inclusion in certain DX8-9 "compliant" software. What's the story?

Centroid sampling is not defined in the specification for Pixel shader 2.0. It is defined in the specification for Pixel shader 3.0 as an instruction modifier, so in pixel shader 3.0 you can declare which attributes are to be sampled at the centroid of the current fragment, and which ones at the center of the pixel. In order to expose centroid sampling through DX9 you have to have a Pixel Shader 3.0 part.

 

[WaltC]

Centroid sampling is not defined in the specification for Pixel shader 2.0. It is defined in the specification for Pixel shader 3.0 as an instruction modifier, so in pixel shader 3.0 you can declare which attributes are to be sampled at the centroid of the current fragment, and which ones at the center of the pixel. In order to expose centroid sampling through DX9 you have to have a Pixel Shader 3.0 part.

OK, thanks much--that puts it nice and succintly and tells me some things about Valve's longevity expectations for the software and about the current "DX9" chips being marketed. Thanks again.

 

[saf1]

I swear I see these artifacts daily when I run DaOC's expansion pack Shrouded Isle...

 

[Myrmecophagavir]

I'm surprised you can't see the artifacts Walt, I had a look and they're very noticeable even at 1024x768. Start a new game with 2x AA and don't move, just leave Max standing on the station platform. As he moves his head about you should see lots of pink and white pixels appearing aroundt the edge of his head and also on the join where the side of his head becomes the top.

Here's a picture (164 KB .bmp); I think it's more noticeable than this in game because you see them sparkle on and off, and I didn't manage to get a particularly artifact-heavy frame.

 

[Squidlor]

HL2 Team Chat Log

http://www.hl2db.com/forums/viewtopic.php?p=186#186



[18:16] <valve|gary> "What about FSAA?"
[18:16] <valve|gary> Anti-aliased has been fixed on all cards.

[18:15] <valve|gary> "What's the strategy for scalability?"
[18:15] <valve|gary> All models are LOD's discretely. You can make lower-end versions of the models as necessary. Displacement maps are simpler on the lowend. You can also tag entities as only showing up on specific dx-levels in the maps. On the lowend, you get less lightmap resolution. As for shaders, shaders typically fallback from complex normal-mapped, specular shaders to single pass simple shaders on the lowend.

[18:16] <valve|gary> "Do you use HLSL or Cg for your dx9 shaders?"
[18:16] <valve|gary> HLSL.

[18:23] <valve|gary> "Will there be a benchmarking routine in HL2?"
[18:23] <valve|gary> Yes. There is a detailed benchmark released before the game ships where you can run a particular card through any of the DirectX levels where appropriate.

[18:23] <valve|david> Sound system: Sounds are all driven by a script file, and sounds are played by label, not by WAV filename, so once you add the hooks into the game code to play the sound, exactly which WAV gets played, over what channel, and with what probability is all controlled by the script. This is great because the sound guy can iterate on sounds without the help of a programmer. We also doppler-shifted bullet sounds, which is just cool!

[18:40] <valve|gary> "What's the largest possible texture."
[18:40] <valve|gary> 2048x2048. If there is a need for larger textures, we can up this limit.

[18:45] <valve|gary> "Is there going to be a version of XSI distributed with the HL2 sdk?"
[18:45] <valve|gary> Yes. We are working very closely with SoftImage to make the integration as tight as possible. This includes distribution of a free version of XSI. More details on that will be released over the coming weeks.

 

[swaaye]

I bet Nvidia poo'ed itself when it saw Valve announce that the FX cards didn't cut it at all. They prolly dropped a bomb on their driver programmers. "Get this working or get out." Man PR like that isn't good for a company relying on gamers to buy their product.

I still can't get over how well designed overall the R3x0 seems to be. It's such an amazing contrast to the R200, chips that always seemed to have a showstopper bug for every special feature.

 

[Bjorn]

I bet Nvidia poo'ed itself when it saw Valve announce that the FX cards didn't cut it at all. They prolly dropped a bomb on their driver programmers. "Get this working or get out."

They said "Anti-aliased has been fixed on all cards" which i guess is more then just a driver update by Nvidia. Though i guess that that doesn't necessarily mean that Nvidia didn't have anything to do with it

 

[Bouncing Zabaglione Bros.]

I still can't get over how well designed overall the R3x0 seems to be. It's such an amazing contrast to the R200, chips that always seemed to have a showstopper bug for every special feature.

It's the power of ArtX. At the time it looked like a quiet little aquisition, but looking back we can see that Dave Orton and ArtX have totally revitalised and turned the staid old ATI into a serious contender. They've gone for engineering excellence.

When you think about it, it is a real watershed, a real turnaround. The fact that everyone was so surprised by the quality of R3x0 (including Nvidia) shows just how much of a jump ATI has made with thier first Dave Orton driven products.

 

[Brimstone]

I still can't get over how well designed overall the R3x0 seems to be. It's such an amazing contrast to the R200, chips that always seemed to have a showstopper bug for every special feature.

It's the power of ArtX. At the time it looked like a quiet little aquisition, but looking back we can see that Dave Orton and ArtX have totally revitalised and turned the staid old ATI into a serious contender. They've gone for engineering excellence.

When you think about it, it is a real watershed, a real turnaround. The fact that everyone was so surprised by the quality of R3x0 (including Nvidia) shows just how much of a jump ATI has made with thier first Dave Orton driven products.

Most of the ArtX workers are former Silicon Graphics people. I remember reading one interview how someone from SGI was praising Tim Van Hooks engineering skills by calling him a wizard. AFAIK, Tim Van Hook is one of the top people at ArtX.

 

[WaltC]

HL2 Team Chat Log

http://www.hl2db.com/forums/viewtopic.php?p=186#186

Beat me to it... Anyway, nice to know they've fixed the AA problem.

 

[Pr( )ZaC]

Sorry to bring this topic up but here's what Gabe has to say about the AA "issue":

We aren't using the same technique on ATI and NVIDIA hardware to avoid the long standing artifacts occuring when you anti-alias games that use lightmaps.

question:
Are you using the "via Pixel Shaders" method with nVidia hardware, to fix the AA issues?
Could you confirm that ATi have fixed the AA using Centroid sampling?

Gabe's answer:
That's correct.

So, it looks like it'll be PS2.0 AA for nVidia and Centroid AA for ATi

 

[Malo]

question:
Are you using the "via Pixel Shaders" method with nVidia hardware, to fix the AA issues?
Could you confirm that ATi have fixed the AA using Centroid sampling?

Gabe's answer:
That's correct.

well the next thing to see is which one has the bigger hit on performance? somebody mentioned that the pixel shader method for nV would mean a considerable it, and with already poor shader performance, would this seriously hurt nV's AA potential with HL2?

 

[tEd]

question:
Are you using the "via Pixel Shaders" method with nVidia hardware, to fix the AA issues?
Could you confirm that ATi have fixed the AA using Centroid sampling?

Gabe's answer:
That's correct.

well the next thing to see is which one has the bigger hit on performance? somebody mentioned that the pixel shader method for nV would mean a considerable it, and with already poor shader performance, would this seriously hurt nV's AA potential with HL2?

that will depend on what nvidia will do with the hl2 shaders