When is 4X FSAA Not 4X FSAA?

whql

Regular
When it runs on an NVIDIA GeForce FX 5200 evidently...

http://www.xbitlabs.com/articles/video/display/nv31-nv34_8.html

Having analyzed the image quality, we can conclude that in OpenGL with 4s FSAA enabled in Quality and Performance modes NVIDIA GeForce FX 5200/5200 Ultra uses only two subpixels instead of four to calculate the pixel color. Moreover, in case of an even frame two upper subpixels are taken from the 2x2 block, while in case of odd frames – two lower subpixels. This is exactly where this unpleasant lines trembling effect comes from. However in high resolutions with high refresh rate or in case of an LCD monitor with high inertia, this trembling is hardly noticeable. So, even though 2s anti-aliasing is involved, you get the impression that there is fully-fledged anti-aliasing based on 4 subpixels.

Seems like it wasn't just for the texture modes NVIDIA took a leaf out of SiS's book. :rolleyes:
 
It's particularly interesting that after their discource of the relative merits of AF & AA techniques, etc, they proceed to bench the various products with control panel settings in their "default" positions.
 
They also say the FX 5200 brings "true multisampling"* to nV's low-end part; I was under the impression the GF4MX also used "true multisampling." I also thought nV 2x was RGMS, not OGMS.

Their games selection was a tad limited, too. Couldn't they find something other than an FPS to benchmark? Warcraft 3 or Commanche 4, maybe?


* If by "true multisampling" they mean "fake multisampling," then I agree. :p
 
Good thing SimonF is on holidays, I am sure he'd have something to say about this... :LOL:

I guess this would result in crawling/jittering edges (due to changing sample points per frame) even when everything is static - not exactly what you desire when enabling AA IMHO.

K-
 
Eh, at least nVidia is the proud inventor of "Dynamic Placement AntiAliasing" :p

This is really evyl IMO...
Hmm, it would be nice if nVidia did separate sliders for AA quality, AF quality, ... - as well as different Desktop / OpenGL / Direct3D settings, like 3dfx did.


Uttar
 
1: You guys dogging it wouldn't buy the card anyway
2: Its an option for those who do. They can choose not to use that mode

Me personally, I'd rather have the option that I never used, rather than not have the option.
 
Me personally, I'd rather have the option that I never used, rather than not have the option.

Seems that they are giving you an option that just relates to something entirely different, so you don't have that particular option.
 
RussSchultz said:
Me personally, I'd rather have the option that I never used, rather than not have the option.

The problem is not "having the option", rather, it's missrepresenting it as 4X FSAA....
 
Mmmmm. I was talking with Andrew Humber the other day about a 5200 reference sample and he mentioned that they are currently waiting on a newer driver revision to alleviate some FSAA screen capture issues - I wonder if this is what he was talking about.
 
DaveBaumann said:
Mmmmm. I was talking with Andrew Humber the other day about a 5200 reference sample and he mentioned that they are currently waiting on a newer driver revision to alleviate some FSAA screen capture issues - I wonder if this is what he was talking about.

I spoke to his (Andrews) collegue Adam Foat a few days ago and he mentioned the same issue (ie capturing AA shots with FX) but concerning the 5800 Ultra.
He also mentioned that that a new driver will fix up 3DMark03 issues and texture filtering issues, but I got sort of a feeling that we misunderstood eachother on these two latter points.

So what would such a driver do? Blend two frame buffer dumps? ;)
 
Ante P said:
I spoke to his (Andrews) collegue Adam Foat a few days ago and he mentioned the same issue (ie capturing AA shots with FX) but concerning the 5800 Ultra.
He also mentioned that that a new driver will fix up 3DMark03 issues and texture filtering issues, but I got sort of a feeling that we misunderstood eachother on these two latter points.

So what would such a driver do? Blend two frame buffer dumps? ;)

No comment.

(There you go, Russ ;-))
 
I do seem to remember a general level of surprise over the 5200's AA performance considering it lacks compression. I wonder if colourless could put some artificial stress into his FSAA sample position app to slow it down to a visible frames per sec and see what its doing then.
 
martrox said:
The problem is not "having the option", rather, it's missrepresenting it as 4X FSAA....

It's not the first time it happens.
Remember '4X, performance' FSAA?
 
DaveBaumann said:
Mmmmm. I was talking with Andrew Humber the other day about a 5200 reference sample and he mentioned that they are currently waiting on a newer driver revision to alleviate some FSAA screen capture issues - I wonder if this is what he was talking about.
This doesn't seem to be a sceen capture issue in the sense that you need a special filter in HyperSnap to capture the post-DAC image; xbit said they saw the fault while playing, not while taking screenshots.

Anyone want to clear up what they meant by proper multisampling?
 
Back
Top