New D3D FSAA Viewer

[Joe DeFuria]

Add "TemporalAAFrameThreshold" to the reg with no value or 0 it should solve the problem you have

Yup...did the trick. Thanks!

 

[Joe DeFuria]

Not for us poor LCD users.

Well, it's looking good for this LCD user!

Haven't tried games yet, but FSAA Viewer, 1600x1200@60 Hz, 4X+2T looking solid! Haven't tried any other modes yet.

What ATI needs to do (if possible) is offer a hot-key to switch between T modes...

 

[Moose]

Some checkbox controls of any type would be nice.

BTW I find that if I change "t" settings all I have to do is also change another D3d setting in the control panel (like AA or AF) and then select apply and the new mode sticks.

 

[dr3amz]

fixed mine

anyway - just ran farcry at very high settings / 3xt / 2aa getting 50-100fps and almost 6x aa quality

has anyone said it yet?

6800 - what ya say to that!

 

[digitalwanderer]

FarCry looks insanely good for me at 4xt2 and runs well, I'm happy.

Off to lose meself testing Painkiller some more.... (And for some weird reason I'm also maxing out the AA & AF and playing the original SoF and it is just freaking gorgeous to me! )

 

[dr3amz]

FarCry looks insanely good for me at 4xt2 and runs well, I'm happy.

Off to lose meself testing Painkiller some more.... (And for some weird reason I'm also maxing out the AA & AF and playing the original SoF and it is just freaking gorgeous to me! )

lol - amen to that me? i'm gonna go enjoy the missus

nn hf

 

[digitalwanderer]

One thing is bugging me though....

Before when I enabled AA with v-sync on it would NEVER run at my refresh rate, always like at a fraction of it depending on what level of AA I ran.....this AA doesn't seem to have that problem.

It's like triple-buffering is on in D3D all the time or something....am I way off base or is this different?

 

[jimmyjames123]

If there was one area where the NV40 seemed vulnerable from the initial reviews, it was with respect to AA. For one, the R3xx was still given a slight edge in AA quality. Also, the NV40 was crippled at 8xAA (of course, this is due to 4xSS + 2xMS mix, and the inherent inefficiencies). If ATI can effectively implement this new temporal AA technique on the R420, the NV40 is going to have a very tough time hanging with certain games and certain AA settings unless they try a new approach.

Does anyone think that we could see a temporal AA technique on the NV cards?

 

[Tim Murray]

If there was one area where the NV40 seemed vulnerable from the initial reviews, it was with respect to AA. For one, the R3xx was still given a slight edge in AA quality. Also, the NV40 was crippled at 8xAA (of course, this is due to 4xSS + 2xMS mix, and the inherent inefficiencies). If ATI can effectively implement this new temporal AA technique on the R420, the NV40 is going to have a very tough time hanging with certain games and certain AA settings unless they try a new approach.

Does anyone think that we could see a temporal AA technique on the NV cards?

Not a real one, no. NV40 doesn't support programmable sample patterns.

 

[digitalwanderer]

If ATI can effectively implement this new temporal AA technique on the R420

Considering that I've gotten it to work pretty damned effectively just mucking about with some registry entries on drivers that before today we didn't know supported it on my R300 than I don't think ATi is going to have too many problems getting it implemented on the R420.

But that's just a guess, what do I know.

 

[jimmyjames123]

If ATI can pull it off, brilliant move on their part. Will make many people quickly forget about SM 3.0, although developers will still be interested!

 

[Moose]

heh,

Temporal AA
3dc
12-16 pipes
clock speed bump
mem speed bump
low heat
low noise
low power
fairly mature drivers


Next week is going to be fun!!!

well for some anyway...

 

[PaulS]

 

[Geo]

So, given this thing works best at lower res and stable frame rates, would this be Xbox2 tech getting a field test? ATI figuring they already made the investment so might as well leverage into the PC space for those who would benefit and get some broad real-world experience at the same time?

 

[Joe DeFuria]

Is there any reason why you wouldn't want the "combined" sample pattern to be "sparse"?

In other words, why isn't "4X Effective Temporal AA" the same pattern as the normal 4X mode?

(Would that make it more prone to flickering?)

 

[Geo]

So, given this thing works best at lower res and stable frame rates, would this be Xbox2 tech getting a field test? ATI figuring they already made the investment so might as well leverage into the PC space for those who would benefit and get some broad real-world experience at the same time?

 

[MuFu]

Is there any reason why you wouldn't want the "combined" sample pattern to be "sparse"?

In other words, why isn't "4X Effective Temporal AA" the same pattern as the normal 4X mode?

(Would that make it more prone to flickering?)

This is exactly what I've been wondering. Plus the individual patterns are sub-optimal so they're actually going to look worse on moving edges than the basic mode.

 

[Joe DeFuria]

So, given this thing works best at lower res and stable frame rates, would this be Xbox2 tech getting a field test? ATI figuring they already made the investment so might as well leverage into the PC space for those who would benefit and get some broad real-world experience at the same time?

I personally don't think this has anything to do with x-box...and I'd guess it works better with higher res than lower res, given the same refresh rate.

 

[Joe DeFuria]

This is exactly what I've been wondering. Plus the individual patterns are sub-optimal so they're actually going to look worse on moving edges than the basic mode.

Ah...

I think you actually answered your (and my) own question.

The patterns they selected are probably based on a comprimise between getting the best quality on moving edges, and the best quality on non-moving edges....

 

[Geo]

. . . given the same refresh rate.

Yes, well, that's the rub when discussing low-res vs high-res, isn't it? Particularly when xbox1 did not exactly ship with a top-end cpu, even for the time. Still, between the two of us I know who is more likely to be right.