far cry ps3 and stuff

[trinibwoy]

Damn the way you people gang bang any positive Nvidia news is hilarious. One guy actually brought up 3Dc :?

 

[Pete]

Thanks for the update, pocketmoon. It all sounds promising. I agree that the skepticism seems undue. Granted, this was at an nV conference, so you'd expect them to play up what they've got. But, last I checked this was the non-denominational 3D Tech and Hardware forum, not the Psssh--Whateva, enVidia *snap* forum.

PM, did they mention if those known few IQ issues with nVidia cards are solved in the next patch, and at what performance hit (if any)?

 

[ChrisRay]

Thanks for the update, pocketmoon. It all sounds promising. I agree that the skepticism seems undue. Granted, this was at an nV conference, so you'd expect them to play up what they've got. But, last I checked this was the non-denominational 3D Tech and Hardware forum, not the Psssh--Whatever forum.

PM, did they mention if those known few IQ issues with nVidia cards are solved in the next patch, and at what performance hit (if any)?

Well two specific scenes dont seem to show a performance hit using the r300 mode in FP32 format.

Though I never got a chance to test many other scenes, (I play the game in r300 mode with FP32 Shaders)

But moving from NV3x to FP32 mode showed a similar framerate in that floor scene with the banded lighting on the floor. I assume from my experiences in r300 room they can remove those Image Quality issues w/o any substancial performance hit.

Chris

 

[Humus]

humus when are we going to see that in games

Depends on how hard we work on getting it out there I guess. It will be included in the next SDK though if nothing else.

 

[pat777]

humus when are we going to see that in games

Depends on how hard we work on getting it out there I guess. It will be included in the next SDK though if nothing else.

Can you show me how this drop in works?

 

[jvd]

Thanks for the update, pocketmoon. It all sounds promising. I agree that the skepticism seems undue. Granted, this was at an nV conference, so you'd expect them to play up what they've got. But, last I checked this was the non-denominational 3D Tech and Hardware forum, not the Psssh--Whatever forum.

PM, did they mention if those known few IQ issues with nVidia cards are solved in the next patch, and at what performance hit (if any)?

Well two specific scenes dont seem to show a performance hit using the r300 mode in FP32 format.

Though I never got a chance to test many other scenes, (I play the game in r300 mode with FP32 Shaders)

But moving from NV3x to FP32 mode showed a similar framerate in that floor scene with the banded lighting on the floor. I assume from my experiences in r300 room they can remove those Image Quality issues w/o any substancial performance hit.

Chris

chris so u disagree with the findings of the website posted later in the thread (and i posted in this )

 

[Humus]

I'll believe it when I see it. BTW, I like the "at work" part. It's still very interesting.

What about the other PS 3.0 effects?

What's wrong with "at work"? That's where I spend 8 hours a day working on things like that. I don't have much time to do things like this at home anymore. But I guess I could make a personal demo of it too. The technique is actually very simple. It took about one day of work to get it up and running.

 

[digitalwanderer]

Unfortunately for nVidia, I have developed a technique at work that will work as a drop-in replacement of dynamic branching that works in the most important situations where you'd otherwise need or prefer ps3.0. You'll probably have to resort to very esoteric effects to really require ps3.0. The effect you mention sounds like a perfect example where this technique would work equally well (or maybe even better) than using dynamic branching, depending on how much the cost of dynamic branches are.
Another typical example is the "early-out" kind of optimization where you for instance detect that the pixel is backfacing the light or is outside the light radius and just return zero instead of going through the all the lighting computations. I have implemented this for a typical "Humus-demo" scene with four lights. With early-out enabled I get 45fps. Without it's 14fps. That's more than 3x performance improvement. If you prefer to dwell in the darkness, the difference is 136fps vs. 25fps. 8) I'm not even sure nVidia's will be able to match that performance increase with ps3.0.

Hey Humus, have I told you how darn much I love you lately?

 

[pat777]

I'll believe it when I see it. BTW, I like the "at work" part. It's still very interesting.

What about the other PS 3.0 effects?

What's wrong with "at work"? That's where I spend 8 hours a day working on things like that. I don't have much time to do things like this at home anymore. But I guess I could make a personal demo of it too. The technique is actually very simple. It took about one day of work to get it up and running.

Ok, can you respond to the question in my previous post? Couldn't nVIDIA just use this technique as well to assist dynamic branching?

 

[Doomtrooper]

At work is just fine

Now a long time ago a certain company failed to offer a certain individual a job, tough luck on their part.

 

[ChrisRay]

Thanks for the update, pocketmoon. It all sounds promising. I agree that the skepticism seems undue. Granted, this was at an nV conference, so you'd expect them to play up what they've got. But, last I checked this was the non-denominational 3D Tech and Hardware forum, not the Psssh--Whatever forum.

PM, did they mention if those known few IQ issues with nVidia cards are solved in the next patch, and at what performance hit (if any)?

Well two specific scenes dont seem to show a performance hit using the r300 mode in FP32 format.

Though I never got a chance to test many other scenes, (I play the game in r300 mode with FP32 Shaders)

But moving from NV3x to FP32 mode showed a similar framerate in that floor scene with the banded lighting on the floor. I assume from my experiences in r300 room they can remove those Image Quality issues w/o any substancial performance hit.

Chris

chris so u disagree with the findings of the website posted later in the thread (and i posted in this )

I havent said that, I said in the two scenes I tested, The performance difference between the Nv3x mode and the r300FP32 was marginal. I am referring to the thread that NV500 guy linked earlier. In those two cases, there was no difference in performance. I am not making any conclusion based off anything other than what I tested.

 

[Humus]

Can you show me how this drop in works?

Sure. For the case of "early-out" with lighting:

1) Render depth-only pass, include ambient in this pass if you want it. You'll probably want a depth-only pass anyway, so this is for free.

For each light {
2) Draw "tagging" pass that tags pixels that needs to be lit by this light. The shader outputs an alpha > 0 if the pixel is supposed to be lit. Alpha test kills fragments that are unlit. The surviving fragment sets stencil to 1. No depth or color writes are needed, so this pass is very cheap.

3) Draw your lighting as usual. Stencil test kills all unlit fragments early, saving an assload of shading power. For all surviving fragments, set stencil back to zero.
}

 

[Humus]

Hey Humus, have I told you how darn much I love you lately?

No, please tell.

 

[Humus]

Ok, can you respond to the question in my previous post? Couldn't nVIDIA just use this technique as well to assist dynamic branching?

It doesn't assist dynamic branching, it just solves the same problem that dynamic branching does in many situations, making them more or less unneccesary in those situations.

 

[Humus]

At work is just fine

Now a long time ago a certain company failed to offer a certain individual a job, tough luck on their part.

That makes it even more fun to rain on their parade.

 

[OpenGL guy]

Sure. For the case of "early-out" with lighting:

1) Render depth-only pass, include ambient in this pass if you want it. You'll probably want a depth-only pass anyway, so this is for free.

For each light {
2) Draw "tagging" pass that tags pixels that needs to be lit by this light. The shader outputs an alpha > 0 if the pixel is supposed to be lit. Alpha test kills fragments that are unlit. The surviving fragment sets stencil to 1. No depth or color writes are needed, so this pass is very cheap.

Except that writing stencil is the same as writing depth if your depth and stencil buffers are combined (as most platforms use). If you have separate depth and stencil buffers, then this is a bigger win.

 

[pat777]

Sure. For the case of "early-out" with lighting:

1) Render depth-only pass, include ambient in this pass if you want it. You'll probably want a depth-only pass anyway, so this is for free.

For each light {
2) Draw "tagging" pass that tags pixels that needs to be lit by this light. The shader outputs an alpha > 0 if the pixel is supposed to be lit. Alpha test kills fragments that are unlit. The surviving fragment sets stencil to 1. No depth or color writes are needed, so this pass is very cheap.

Except that writing stencil is the same as writing depth if your depth and stencil buffers are combined (as most platforms use). If you have separate depth and stencil buffers, then this is a bigger win.

Interesting, seems like common sense if you ask me.

I guess there's no real advantages of dynamic branching unless an nVIDIA guy comes up with a technique that's much better with PS 3.0.

 

[jimmyjames123]

That makes it even more fun to rain on their parade.

So this implies that you have a personal axe to grind with NVIDIA...

Why would anyone want to "rain on" either ATI or NV's parade? Sounds counterintuitive to a positive environment on the forums. The only reason I can think of for such behavior is if one works for either ATI or NVIDIA...hmmm...

but then, of course you do work for ATI!

Humus, if I may dare ask, how much experience do you have working with and coding for a card that fully supports SM 3.0?

 

[Pete]

Thanks, Chris. I was hoping to hear something direct from the devs, too, but they may just let the next patch speak for them.

 

[overclocked]

Very good info pocketmoon66, thank´s for the summary!
One question though about this thing with AA not working with HDR, whst´s the technical reason for this?
I have the same question about Halo , why can´t AA work, or is this temporal until a later patch/driver´s etz..?