Anandtech to use Emulators for benchmarking?!!

[fivefeet8]

http://www.anandtech.com/video/showdoc.html?i=1986

Wow!!! Emulation benchmarks on a Hardware site. Interesting indeed. Looks like our plugin coding master Pete helped them with it. Cool. :bounce:

Edit: I don't quite agree with their conclusion though. The article could have done better with some Petes OGL(1)/D3d benchmarks too. So we can have some FSAA goodness benchmarks and full hardware framebuffer benches. And maybe throw in a few lewpys glide with a Voodoo 5 5500 for reference.

They should have also used the "Very high x, very high y" settings too.

[n_w95482]

Let's just hope they know how to properly configure the emulators. They could always come here if they needed help .

[Yeloazndevil]

interesting article

[Cylinder]

>And maybe throw in a few lewpys glide with a Voodoo 5 5500 for reference.

Don't take this the wrong way, I kinda see a point in the distance, almost.. but why? For the 50 people that still even use one?

[fivefeet8]

Originally Posted by Cylinder

>And maybe throw in a few lewpys glide with a Voodoo 5 5500 for reference.

Don't take this the wrong way, I kinda see a point in the distance, almost.. but why? For the 50 people that still even use one?

For comparison's sake. And for the readers of anandtech to get more information about emulation.

[mr.saraiva]

Originally Posted by n_w95482

Let's just hope they know how to properly configure the emulators. They could always come here if they needed help .

from the article:

Originally Posted by AnandTech

Aside from writing high quality plugins, Pete is also a very cool guy and he had a lot of useful advice for us when we let him know that we would be tackling the issue of emulator benchmarking.

I guess they know how to configure the plugins...at least gfx.

[AChrisRay]

Originally Posted by fivefeet8

For comparison's sake. And for the readers of anandtech to get more information about emulation.

Would be useless. Your well capable of loading fraps yourself and determining your average frames per second ect,


I have to admit those frame buffer results dont make sense, Doesnt OpenGL 2.0 Plugin use the pbuffer?



Err now that I see they are using the Shader 2.0 arb extension. The results dont surprise me at all. That has more to do with the Shader limitation than framebuffer tho. In my opinion.


But it's interesting none the less.

[fivefeet8]

Originally Posted by AChrisRay

Would be useless. Your well capable of loading fraps yourself and determining your average frames per second ect,


I have to admit those frame buffer results dont make sense, Doesnt OpenGL 2.0 Plugin use the pbuffer?



Err now that I see they are using the Shader 2.0 arb extension. The results dont surprise me at all. That has more to do with the Shader limitation than framebuffer tho. In my opinion.


But it's interesting none the less.

Yeah, I could, but not very many people have a voodoo 5 to compare it with.

I agree about the framebuffer tests. Maybe if they used a heavy FB effects summons. I really wonder if they even did though. Of course they could use MGS to test FB speed, but that's just me.

[blehz0r]

aren't pete's plugins a little bit optimized for nvidia chips? (not talking about the ogl2 plugin)

[AChrisRay]

Originally Posted by fivefeet8

Yeah, I could, but not very many people have a voodoo 5 to compare it with.

I agree about the framebuffer tests. Maybe if they used a heavy FB effects summons. I really wonder if they even did though. Of course they could use MGS to test FB speed, but that's just me.

Ya what they are doing there really isnt a test of frame buffer speed at all. it Creates a a Shader bottleneck, Not a framebuffer bottleneck, That conclusion is currently flawed IMO,


The Only reason battle scenes and 2d change FPS so much is because shaders are being used to filter all of it.


I still wouldnt want to use an ATI Card with OpenGL 2.0 since they still dont support Max X/Y right?

[Razor Blade]

We will take a look at a very early Game Cube emulator running "The Legend of Zelda: The Windwaker"

Sweet. There as exited about Next Gen emus as we are.

[fivefeet8]

Originally Posted by AChrisRay

I still wouldnt want to use an ATI Card with OpenGL 2.0 since they still dont support Max X/Y right?

*me looks at Razor Blade*

Why don't cha try it out and tell use already.

[Razor Blade]

It cannot use Max X but can use Max Y. Hope that answers any questions.

[AChrisRay]

Originally Posted by Razor Blade

It cannot use Max X but can use Max Y. Hope that answers any questions.

Ya What I thought (

[Razor Blade]

Just curious. Since I'll probably never see for myself, but is there that much of a difference between the two?

[AChrisRay]

Originally Posted by Razor Blade

Just curious. Since I'll probably never see for myself, but is there that much of a difference between the two?

Lemme Show you. I have screens of it Right here.




(Normal X/Y With Shader Set to level 3)
Btw heres settings used for this pic.

Plugin: Pete's OpenGL2 Driver 2.3
Author: Pete Bernert
Card vendor: NVIDIA Corporation
GFX card: GeForce FX 5900/AGP/SSE/3DNOW!

Resolution/Color:
- 1024x768 Fullscreen - NO desktop changing
- Internal X resolution: 1
- Internal Y resolution: 1
- Keep psx aspect ratio: off

- Filtering: 6
- Hi-Res textures: 1
- TexWin pixel shader: on
- VRam size: 128 MBytes

Framerate:
- FPS limitation: on
- Frame skipping: on
- FPS limit: Auto

Compatibility:
- Offscreen drawing: 1
- Framebuffer effects: 1
- Framebuffer uploads: 1

Misc:
- Scanlines: off
- Line mode: off
- Screen filtering: on
- Shader effects: 1/3
- Game fixes: off [00000000]






And Here's the Same settings cept Maximum X/Y Resolution is set.

Plugin: Pete's OpenGL2 Driver 2.3
Author: Pete Bernert
Card vendor: NVIDIA Corporation
GFX card: GeForce FX 5900/AGP/SSE/3DNOW!

Resolution/Color:
- 1024x768 Fullscreen - NO desktop changing
- Internal X resolution: 2
- Internal Y resolution: 2
- Keep psx aspect ratio: off

- Filtering: 6
- Hi-Res textures: 1
- TexWin pixel shader: on
- VRam size: 128 MBytes

Framerate:
- FPS limitation: on
- Frame skipping: on
- FPS limit: Auto

Compatibility:
- Offscreen drawing: 1
- Framebuffer effects: 1
- Framebuffer uploads: 1

Misc:
- Scanlines: off
- Line mode: off
- Screen filtering: on
- Shader effects: 1/3
- Game fixes: off [00000000]








As you can see. Max X/Y makes a huge difference in Quality. Almost enables an Anti Aliasing

[Razor Blade]

NOOO! Damn, Too bad for me. Well thanks for the quick reply and in depth may I add. Looks like I'll just use good old OpenGL. I like FSAA better anyways.

[AChrisRay]

Originally Posted by Razor Blade

NOOO! Damn, Too bad for me. Well thanks for the quick reply and in depth may I add. Looks like I'll just use good old OpenGL. I like FSAA better anyways.

I know what you mean. I prefer the old Plugin Most of the time too. However in games which have nasty framebuffer effects ect, I use 2.0



Just for comparison sake. Heres 1.75 With 8x AA and my settings


Plugin: Pete's OpenGL Driver 1.1.75
Author: Pete Bernert
Card vendor: NVIDIA Corporation
GFX card: GeForce FX 5900/AGP/SSE/3DNOW!

Resolution/Color:
- 1024x768 Fullscreen - NO desktop changing [32 Bit]
- Keep psx aspect ratio: off

Textures:
- R8G8B8A8
- Filtering: 6
- Hi-Res textures: 1
- VRam size: 128 MBytes

Framerate:
- FPS limitation: on
- Frame skipping: on
- FPS limit: Auto

Compatibility:
- Offscreen drawing: 2
- Framebuffer texture: 2
- Framebuffer access: 0
- Alpha multipass: on
- Mask bit: off
- Advanced blending: off

Misc:
- Scanlines: off
- Line mode: off
- Unfiltered FB: off
- 15 bit FB: off
- Dithering: off
- Screen smoothing: off
- Screen cushion: off
- Game fixes: off [00000000]






Which In my opinion looks the best and doesnt even need shaders to filter.

[Razor Blade]

One thing. Do you think that your 8xfsaa is that much different than 4xfsaa? Cuz on my card I have 6x but never really use it because I cannot see a big improvement? 4x for me makes everything smooth as butta.

[AChrisRay]

Originally Posted by Razor Blade

One thing. Do you think that your 8xfsaa is that much different than 4xfsaa? Cuz on my card I have 6x but never really use it because I cannot see a big improvement? 4x for me makes everything smooth as butta.

Eh 4x on Nvidia cards isnt that great, Unlike ATI which has a great 4x pattern. Nvidia requires 4xS to make the difference.


I just throw 8x On because it looks impressive and works in OpenGL. But IMO 4xS is Nvidias best Quality/Performance along with 2x.


6x didnt offered much improvemebnt over 4x for me with my 9500 Pro either.