Examples of ATI filtering failing?

[tEd]

i can understand that i've waited with judgment too until i've seen it today. I have to say that i'm very sensitive to such things. Others may not notice the difference but i do

Yeah, that's the suckbutt bit about this...it really is something that I think it's going to come down to a personal opinion on that you'll actually need to sit and play with the hardware to find out.

Could you describe a few games and the specific things you're noticing? A subjective description is probably the best we'll be able to get without the hardware...I don't see anyway a screenshot could really get it and I still don't like the quality of game movies.

I need ATi & nVidia to send me cards, THAT'S what I need.... 8)

MP2 = increase in moire. Actually i was a little shocked how it looked. It was pretty bad. Probably worst case of the specific level i've tested(e1 level7)

Halo = less smooth transition of the mipmaps.

Nfsu = less smooth tranistion of the mipmaps. It was very good visible here because of the middle line

I need to do some more testing. While i definately see the differnce it's not really a problem in fast paced games like ut or quake3. In some other games like halo,MP2 or nfsu(because of the middle line) i just want to use trilinear as the smoother mipmap transition makes a difference for me and i think i can expect to get the full potential of the card i bought which was not cheap btw.

 

[Kombatant]

i can understand that i've waited with judgment too until i've seen it today. I have to say that i'm very sensitive to such things. Others may not notice the difference but i do

Yeah, that's the suckbutt bit about this...it really is something that I think it's going to come down to a personal opinion on that you'll actually need to sit and play with the hardware to find out.

Could you describe a few games and the specific things you're noticing? A subjective description is probably the best we'll be able to get without the hardware...I don't see anyway a screenshot could really get it and I still don't like the quality of game movies.

I need ATi & nVidia to send me cards, THAT'S what I need.... 8)

MP2 = increase in moire. Actually i was a little shocked how it looked. It was pretty bad. Probably worst case of the specific level i've tested(e1 level7)

Halo = less smooth transition of the mipmaps.

Nfsu = less smooth tranistion of the mipmaps. It was very good visible here because of the middle line

I need to do some more testing. While i definately see the differnce it's not really a problem in fast paced games like ut or quake3. In some other games like halo,MP2 or nfsu(because of the middle line) i just want to use trilinear as the smoother mipmap transition makes a difference for me and i think i can expect to get the full potential of the card i bought which was not cheap btw.

As the saying goes...

 

[Hanners]

I would put money on rTool fixing the Halo issues - Possibly the other two as well? Give it a try and see what happens.

 

[digitalwanderer]

As the saying goes...

Hey, not cool.

This thing IS going to be subjective Komb, so having people just tell us what they see/think is going to be critical in analysing this try/bry impact.

While I agree some visual aids would be nice, the fellows opinion shouldn't be discounted just because he don't have any up. Not everyone has an easy time getting/uploading them, and mipmap lines can be a bitch to show in screenshots. (Don't get me started with the problems with capturing ingame movies!)

Please, please, PLEASE let's try and not be confrontational about this and be more open minded and just examine the evidence as it's presented?

 

[jvd]

dw if he wont post screen shots or a video why should i believe him. I have an x800pro here and halo too and i don't see a problem. If he is having these problems why wont he post proof

 

[tEd]

I would put money on rTool fixing the Halo issues - Possibly the other two as well? Give it a try and see what happens.

i't s not that believe me. I'm just seeing the difference , it's not totaly huge like going from triliner to bilinaer , but i see that the mipmap transition is not as smooth as it was with my r9800pro.

For me personally it's a visible difference for other it isn't. That's why i think a disable option is the best solution for everyone. There are games brilinear is fine with me and there are games which it isn't. Give me the choice that's all.

 

[Hanners]

I would put money on rTool fixing the Halo issues - Possibly the other two as well? Give it a try and see what happens.

i't s not that believe me. I'm just seeing the difference , it's not totaly huge like going from triliner to bilinaer , but i see that the mipmap transition is not as smooth as it was with my r9800pro.

Have you tried rTool regardless though? So far everyone I've asked that question has refused to for whatever reason, so until someone can confirm (with comparison screenshots if possible) that it really is something other than texture stage optimisations then I'm afraid I'll have to remain skeptiocal.

Alternatively, you're more than welcome to mail me a save game from a point in Halo or Max Payne 2 where the transitions are obvious, and I can take look on my 9600 if that helps.

 

[digitalwanderer]

dw if he wont post screen shots or a video why should i believe him. I have an x800pro here and halo too and i don't see a problem. If he is having these problems why wont he post proof

Oh, well that's different...I didnt' know you were speaking from personal experience.

Sorry, please ignore me...I think I'll go clean my living room or set up a pool or something, it's the kids first day of spring break and I'm a bit frazzled.

 

[Socos]

It has already failed.

Their are two was you can talk about its failure.

1. Is mathmatically which I'm sure anyone who understands what trilinear is knows it has already failed this.
2. IQ I think the poll clear shows that more poeple think that trilinear looks better then trylinear.

Now you could try a third option but ATI has specifically said stuff making this third way not useful

3. The IQ degrading is small enough for the corresponding Preformance boost.

But ATI have stated explictly their is no IQ degradation at all they claim it lokos better in some case but never worse.


Truely if ATI's implementation was dynamic and didn't have any problems then ATI wouldn't have to disable ( which they do ) when colour mipmaps are used. Since it isn't able to handle these situation I once again give this as a reason for its failure.

Bob,

Where are those screenshots/video's showing this failure????

Too me where ATI failed is when they tried to act like they were not doing something similar to what Nivida was doing, and then had the review sites put the two cards side by side with Nivida's optimizations turned off and no way to turn it off on the ATI cards. That was shady too me, but the fact that no one, and I mean no one has yet to point out a difference in output tells me it must be a pretty good alogrithm. Your point of it failing because it turns itself off when color'd mip maps are used doesn't hunt. It is doing exactly what it is supposed to do because if it did not turn itself off it would be lowering IQ.

There are plenty of people out there with two systems, probably one with a 9600 the other with a 9800 or 9700 or GF4, or something that should be able to say hey when I play this game I can see shimmering on this card that I don't see on this card.

No one has pointed this out yet, and I am sure their are plenty of sites out there looking.

 

[digitalwanderer]

There are plenty of people out there with two systems, probably one with a 9600 the other with a 9800 or 9700 or GF4, or something that should be able to say hey when I play this game I can see shimmering on this card that I don't see on this card.

No one has pointed this out yet, and I am sure their are plenty of sites out there looking.

Again I'll say it, I've been trying to see the difference on my 9600/9700 and just can't no matter how hard I try.

That still doesn't mean there isn't a difference, but it definately says to me that it's very marginal at best and it is currently beyond my ability to notice any image degradation.

 

[jvd]

There are plenty of people out there with two systems, probably one with a 9600 the other with a 9800 or 9700 or GF4, or something that should be able to say hey when I play this game I can see shimmering on this card that I don't see on this card.

No one has pointed this out yet, and I am sure their are plenty of sites out there looking.

Again I'll say it, I've been trying to see the difference on my 9600/9700 and just can't no matter how hard I try.

That still doesn't mean there isn't a difference, but it definately says to me that it's very marginal at best and it is currently beyond my ability to notice any image degradation.

dw .

I havee borrowed games . Downloaded demos trying to see everything.


I own

tnt

3dfx voodoo1 , 2 , 3 ,5

geforce sdr , ddr

geforce 2 , geforce 2 ultra

geforce 3 , 500 ti , 4600ti

geforce fx 5800ultra

rage maxx

rage 3d

radeon vivo 64meg

radeon 8500pro 64 meg

radeon 9600pro (or xt dunno off the top of my head my sister bought it )

9700po

9800 xt aiw

x800pro

in a few weeks x800xt and geforce 6800gt

Now i don't own these all. My father , 3 sisters and brother in law own them along with me. But since my brother in law and older sister lives 10 mins ago (geforce 4 ti 4600)

So trust me I can say there is no diffrence in any of the games i've tried and i've tried alot .

If i found something i'd throw up a video of it. But i haven't .

 

[digitalwanderer]

EDITED BITS: Sorry Dave

 

[jvd]

which is exactly why i'm sick of people saying that they see it but wont take 10 mins to make a video. Or to take pictures showing it. Because unless i'm blind as hell i can't see anything .

 

[digitalwanderer]

which is exactly why i'm sick of people saying that they see it but wont take 10 mins to make a video. Or to take pictures showing it. Because unless i'm blind as hell i can't see anything .

I can see that, but I figure arguing with them is kind of like arguing with the weather.

The burden of proof lies with those claiming there is a discrepency, until they can show that I remain pretty confident from my own investigation that it's a moot point. 8)

 

[jvd]

which is exactly why i'm sick of people saying that they see it but wont take 10 mins to make a video. Or to take pictures showing it. Because unless i'm blind as hell i can't see anything .

I can see that, but I figure arguing with them is kind of like arguing with the weather.

The burden of proof lies with those claiming there is a discrepency, until they can show that I remain pretty confident from my own investigation that it's a moot point. 8)

right and thats why i ask and even offer to host pics and movies and you notice how this guy in this very thread ignores me ?

That goes to tell you alot about what he is really seeing .

 

[Charmaka]

Serious Sam SE technology test level on a 9600 Pro AIW, Cat 4.3 (98SE), control panel to "balanced", game options: Trilinear, 16x AF, everything set as much to "normal" as I can discern. Steep sets of steps in-game always seem to induce moire patterns; it's much more obvious when you start setting options to "sharpen" but it's still there with things scaled back. Not presenting this as conclusive proof one way or the other and I don't know if this is confined to the 9600 or is a problem with SS:SE in general, but it's definitely there. It's easier to see on the level with the three temples and the jaguar statues as there's less of a glare problem (position for the above screenshot is fairly fixed as the sun needs to be hidden behind the archway to be able to see anything) and the steps are longer and more accessible. So yeah.

 

[tEd]

which is exactly why i'm sick of people saying that they see it but wont take 10 mins to make a video. Or to take pictures showing it. Because unless i'm blind as hell i can't see anything .

1. what you guys have with the pictures. it won't show you anything at all.

2. i made one movie so far but frankly it's not really a good representation from what i see on my screen and they are stuttery which doesn't help. I need some more time and eventually you get your movies. Maybe next week or so

I wanna put back my r9800 and do some more testing but it takes time. Anyway i see a difference here and there that's my personal opinion on the matter so far. I don't see why this is such a big deal for others. If you don't see a difference that's fine with me i'm just giving my opinion on the matter from my own experince that's all.

 

[jvd]

Serious Sam SE technology test level on a 9600 Pro AIW, Cat 4.3 (98SE), control panel to "balanced", game options: Trilinear, 16x AF, everything set as much to "normal" as I can discern. Steep sets of steps in-game always seem to induce moire patterns; it's much more obvious when you start setting options to "sharpen" but it's still there with things scaled back. Not presenting this as conclusive proof one way or the other and I don't know if this is confined to the 9600 or is a problem with SS:SE in general, but it's definitely there. It's easier to see on the level with the three temples and the jaguar statues as there's less of a glare problem (position for the above screenshot is fairly fixed as the sun needs to be hidden behind the archway to be able to see anything) and the steps are longer and more accessible. So yeah.

great job on the balanced buddy you should put the the texture slider and other sliders all the way up as ati already said they get more aggressive as you lower the bar . That is ofcourse why you can lower the bar

 

[jvd]

which is exactly why i'm sick of people saying that they see it but wont take 10 mins to make a video. Or to take pictures showing it. Because unless i'm blind as hell i can't see anything .

1. what you guys have with the pictures. it won't show you anything at all.

2. i made one movie so far but frankly it's not really a good representation from what i see on my screen and they are stuttery which doesn't help. I need some more time and eventually you get your movies. Maybe next week or so

I wanna put back my r9800 and do some more testing but it takes time. Anyway i see a difference here and there that's my personal opinion on the matter so far. I don't see why this is such a big deal for others. If you don't see a difference that's fine with me i'm just giving my opinion on the matter from my own experince that's all.

I want videos . I already told u i would host .

Before you said it was apparent as soon as you played a game . Now you say you see it here and there ? Which one is it ?

What is wrong with the movie that you can't send it to me ?

 

[Charmaka]

OpenGL slider at "balanced" results in:

AA: app pref
Max res: app pref
AF: app pref
AF method: app pref
Texture preference: high quality (i.e. max)
Mipmap detail level: high quality (ditto)
Vsync: default off
Truform: off

Explain where I'm going wrong here please. My understanding is that AA is irrelevant (as it's MS and thus not affecting textures), AF and AF method should be app controlled for it to work (ie not defaulting to bilinear for second and subsequent stages etc), vsync shouldn't matter, truform is a non-issue and the other two are on max anyway. Please enlighten me.