Nvidia Altered its 8xS Sample Pattern with 61.11 drivers

I guess you folks haven't seen any alphas on fences so far or haven't noticed that render2texture shadows do not get in fact antialiased with any form of MSAA so far. There's nothing wrong with the screenshots from what I can tell.

3DCenter is keeping a window of doubt open for it, yet it looks at least to me like the 8x (reg18) setting is truly a 4xRGMS+2xOGSS combination and has thus a 4*4 EER:

http://www.3dcenter.org/artikel/nv40_benchmarks/index6.php

Performance numbers for each game/application under "Setting3", obviously the higher numbers.
 
Re: Nvidia Altered its 8xS Sample Pattern with 61.11 driver

ChrisRay said:
Since I cant seem to upload it here, I will link to where I could. It appears to be 4 samples of Multi sampling and 2 samples of Super Sampling.. I'm gonna do some quality tests in a bit, Personally I dunno if its going to look better or worse,

But its definately less brute force than before. (But its all ordered)

Since there's no performance difference I don't see much reason for not in fact picking the far more effective 4xRGMS mode for the combination.

Less "brute force" is understandable; there's a huge difference between sacrificing 1/4th of your fill-rate (4xSSAA) and only 1/2 (2xSSAA).
 
Re: Nvidia Altered its 8xS Sample Pattern with 61.11 driver

Ailuros said:
ChrisRay said:
Since I cant seem to upload it here, I will link to where I could. It appears to be 4 samples of Multi sampling and 2 samples of Super Sampling.. I'm gonna do some quality tests in a bit, Personally I dunno if its going to look better or worse,

But its definately less brute force than before. (But its all ordered)

Since there's no performance difference I don't see much reason for not in fact picking the far more effective 4xRGMS mode for the combination.

Less "brute force" is understandable; there's a huge difference between sacrificing 1/4th of your fill-rate (4xSSAA) and only 1/2 (2xSSAA).


Well I wasnt referring to the 6800 Series, I was referring to the Geforce FX series, This new AA is crap btw, It produces terribly Quality IMO, And it just seems broken for the NV40 cards,

4xS and 6x look better than both the new 8x Modes on the FX line. But I hope they fix this. So Far I have been exceedingly disapointed with the 6x.xx series of drivers.

its always some new thing that ticks me off about them.
 
trinibwoy said:
DemoCoder said:
Pottle. Kettle. Black.

WTF is a Pottle??!! ROFLMAO!!!

Well actually, out here in New Zealand it is what we call a particular design of plastic ice-cream container for around two litre volumes.

Perhaps DemoCoder has been globe trotting?
 
I personally think DC just misspelled it, meaning not pottle, but pittle......as in to piss on...... kind of like so many here are doing to each other ATM..... This place needs some serious housebreaking... ;)
 
The AA testers pics inthe nvnews thread seem to show 4x supersampling except the geometry samples aren't in the same position as the texture ones. The only other time I've seen this is on an 8500 (and its varients) :?

thats for the 8xS of rivatuner.


The 8xs of the drivers seems odd. Any chance you could get a shot of it in 'both' mode rather than 'three square'?.
 
Ailuros said:
I guess you folks haven't seen any alphas on fences so far or haven't noticed that render2texture shadows do not get in fact antialiased with any form of MSAA so far. There's nothing wrong with the screenshots from what I can tell.

The cars are suffering from some pretty bad aliasing too, as are parts of the track. Are there alphas on pretty much everything in that game?
 
Back
Top