• Some users have recently had their accounts hijacked. It seems that the now defunct EVGA forums might have compromised your password there and seems many are using the same PW here. We would suggest you UPDATE YOUR PASSWORD and TURN ON 2FA for your account here to further secure it. None of the compromised accounts had 2FA turned on.
    Once you have enabled 2FA, your account will be updated soon to show a badge, letting other members know that you use 2FA to protect your account. This should be beneficial for everyone that uses FSFT.

x800 filtering artifacts in motion (Far Cry)

T

tranCendenZ

2[H]4U
Joined
Jun 6, 2004
Messages
3,844
A fellow from Beyond3d captured these videos at 800x600 to demonstrate a "shimmering" effect he noticed on his x800PRO that he did not notice on his 9700PRO. When watching the videos, examine the left wall. In the x800PRO video, the wall appears to "shimmer" or jump around. In the 9700PRO video, this effect is not pronounced. Check it out:

X800PRO (Max filtering quality/Brilinear)
http://www.reflectonreality.com/images/r420/FarCryx800ProHiRes_Divx.avi

9700PRO (Max filtering quality/Full Trilinear)
http://www.reflectonreality.com/images/r420/FarCry9700ProHiRes_Divx.avi

P.S. - Any chance of revisiting the X800 vs. 6800 shootout? Since there are visible differences between the filters as seen above, since it has been shown Nvidia's Brilinear and ATI's Brilinear do the same amount of Bilinear filtering, and since it has also been proven that both Nvidia and ATI cards can get a 10-20% performance boost from using Brilinear filtering, would it not be most in accordance with HardOCP's benchmark policies to benchmark the 6800 with Full Trilinear disabled? I think that would certainly be the most fair given this turn of events, and would give a better picture of what resolutions and AF levels each particular card could handle when doing a similar filtering workload with similar output results. Only problem is based on the number of tests you did, that looks like it would be a beast to do over. :/

Anyway, enjoy the movies. Personally I don't think its anything to be excited about, but there is a visible difference.
 
based on Brent's post i doubt we will see new reviews being made. Maybe we will get some new reviews once nvidia gets some WHQL'ed drivers and some retail boards kickin.
 
Bad_Boy said:
based on Brent's post i doubt we will see new reviews being made. Maybe we will get some new reviews once nvidia gets some WHQL'ed drivers and some retail boards kickin.
Click to expand...

Yeh, I figure that would make the most sense. Would be a waste to do the shootout all over just to do it again in 2 weeks when the boards are out I assume. OTOH, a lot of people put faith and reference HardOCP's scores, so they might want to put a disclaimer up on those scores or something.
 
tranCendenZ said:
OTOH, a lot of people put faith and reference HardOCP's scores, so they might want to put a disclaimer up on those scores or something.
Click to expand...

dunno what OTOH means, but i agree with the disclaimer part.
 
Brent_Justice said:
a disclaimer about what
Click to expand...

Well a few things. First this on Page 2:
http://www.hardocp.com/article.html?art=NjExLDI=
"The filtering between the X800 series and the 9800XT look identical. Keep in mind these are the default settings in place via the driver, so when you install your video card, install the drivers, and play your games this is the level of Trilinear filtering you will have by default. The X800 series and Radeon 9800XT provide a "truer" default Trilinear image quality."

Obviously this is not the case, and should be corrected IMO. By default both Nvidia and ATI use brilinear filtering. You can't get full trilinear in any version of ATI's x800 drivers.

I think the most glaring part is the end:
http://www.hardocp.com/article.html?art=NjExLDEw

You list the highest resolution playable with variable levels of AA/AF. I think it would make sense to add a disclaimer such as:

"UPDATE: It has been discovered that ATI X800 cards use a "brilinear" filter instead of full trilinear. This may skew Ansiotropic Filtering performance for ATI cards by up to 20%. We will investigate this issue further in future reviews/"
 
i think what tranCendenZ is saying is that from the videos there is an obvious difference from the r350, hence the increased shimmering. and that it should be noted.
 
Brent_Justice said:
WRT AF ATI has used an adaptive algorithm since the 8500 refer to here: http://www.hardocp.com/article.html?art=MzQ0LDQ= first paragraph

and was improved in the 9700Pro, this is common public knowledge
Click to expand...

First off, as I'm sure you know, the brilinear filter introduced with the 9600 and now X800 has not been present in any other ATI card known to date AFAIK. It is a new, wholly different addition to what ATI has done in the past in terms of filtering.

That being said, I think what I'm getting to is that you make it seems like Nvidia's brilinear is worse than ATI's x800 best filtering quality, while the x800 is actually doing the same amount of bilinear filtering as the 6800's brilinear mode - and that the x800's best filtering is inferior to the NV40's and R350's best filtering. One of the german articles, I think the computerbase.de one, was read by one of the B3D members and found out that ATI's new Brilinear mode in the x800/9600 works like an on/off switch. The author of the article continued to create mips step by step approaching what would be a standard colored mip test, and he found a single mipmap threshold where the _entire scene_ switched from brilinear to full trilinear. This is not "adaptive" in any manner except disabling the new brilinear optimization for synthetic mipmap testing like an on/off switch. That info, combined with IXBTLABS article that shows the amount of bilinear filtering done between NV40 brilinear and ATI Brilinear, shows that ATI's best filter is not significantly better than Nvidia's brilinear filter on the NV40 - plus technically ATI's filter could cause artifacting due to underfiltering when compared to full trilinear on the NV40. Add on top of that movies such as the above which show filtering quality visibly inferior to that of the 9700PRO, and you think that would be something you want to address in your article and the performance results you obtained from the x800PRO/XT. As it stands, adaptive or not, you make it seem like the x800PRO is doing "true trilinear" by default when its really just doing brilinear during games, and trilinear during mipmap testing on every driver set available, with "true trilinear" being nowhere in sight. On the other hand, with the 60.72 drivers you benchmarked the NV40 with you actually can get full trilinear quality out of the NV40, but at a reduced performance of course - or you can use a similar quality brilinear filter and get the 10-20% boost that x800 cards are getting.

In other words, use a disclaimer to show that the x800pro is doing something different than the nv40 and past ati cards in the driver sets you benchmarked, namely forcing brilinear filtering and getting a significant performance boost because of this similar to the geforce fx series. And, based on statements from both companies, it appears Nvidia will provide Full Trilinear in their official drivers, while ATI seems opposed to the idea unless it can be "proven to be useful" by the consumer.

Or you could have just summed it up in one sentence as the post above ;)

I keep hearing this and I think it makes sense. If you can play 1600x1200 with 4xAA and 8xAF at 60fps+ in many games, then ATI users should have the option of upping the filtering quality another notch to full trilinear. However, if review sites focus on speed instead of driver options and image quality, ATI and Nvidia users probably won't have the option in the future. I mean, a lot of reviews, when I say a lot I mean most, put Nvidia's full trilinear filter up against ATI's Brilinear filter. Since both looked OK in the mips, most assumed both were full trilinear. Therefore Nvidia ended up looking slower in Ansiotropic filtering when it was in reality using a higher quality filter that was requested by gamers and review sites numerous times in the past. Reviewers could have used Nvidia's brilinear filter and gotten similar IQ and performance to ATI, but failed to do so thinking both were full trilinear based on mipmap examinations. So where does that leave Nvidia and ATI? The company that developed the optimization that was hidden best from reviewers ended up looking better in the benchmarks. Since Nvidia did not hide its Brilinear mode as well as ATI, it simply looked slower while it was in fact providing a full trilinear filter as opposed to a brilinear filter. My qualm is that if reviewers do not call companies on these practices, we will have these super high end cards that do insane resolutions but are lower quality IQ than previous generations simply so they can get the highest bar in performance.

And the person that loses out in that case is the end user, who has the fastest card on the planet that gets bested by a GeForce4 MX in filtering quality.
 
tranCendenZ said:
That being said, I think what I'm getting to is that you make it seems like Nvidia's brilinear is worse than ATI's x800 best filtering quality, while the x800 is actually doing the same amount of bilinear filtering as the 6800's brilinear mode
Click to expand...

tranCendenZ, it's not. What we got out of the ixbt investigation was only one test case. If ATI's method does what they say it does, then at times it could apply even less filtering than brilinear and at other times a lot more. It depends on how harsh the drivers detect the transition is, thus the adaptive nature they claim.
 
tranCendenZ said:
Anyway, enjoy the movies. Personally I don't think its anything to be excited about, but there is a visible difference.
Click to expand...

If it's nothing to get excited over why are you insisting on revisiting the review and applying disclaimers? :rolleyes:

I read the thread over at beyond3d and at this point it is speculation on what EXACTLY is occuring.
sugggestions have been made for both testing to occur on other games to attempt to recreate the issue. It is also noted by several within the thread this also occurs on both legacy ATI and nVidia cards. So again....if its not something to get excited over why the insistance on a redo? IMO if another review is to occur it needs to occur when nVidia has a final release card with drivers to review.....at this point its all built on limited information.
 
Blackwind said:
If it's nothing to get excited over why are you insisting on revisiting the review and applying disclaimers? :rolleyes:
Click to expand...

i dont think that claims as excitement on his part. excitement would be saying "YOU MUST DO COMPLETE REVIEWS OVER BECAUSE OF WHAT THESE VIDEOS SHOW" ....hes obviously not doing this. he just saying it should be noted in some sort of disclaimer. and i agree.

i think a disclaimer would be necessary, atleast until we get retail boards from nvidia with new retail WHQLed drivers. (instead doing reviews completely over, on information that is limited like you say)
 
Some people just won't stop. :rolleyes:

Brent the reviews are great. Evidently you can't please all of the people all of the time. Ignore these folk. They are just trying to get under your skin.
 
Bad_Boy said:
yes that is exactly the case :rolleyes:
Click to expand...

well then why do YOU want him to change it. Are you afraid that you will forget what you know? You may need it for reference later? Why? How will it help you?
 
I think he changed the LOD. It's the same problem that occurs when you use starstorm drivers or any other driver that alters the LOD, you end up with shimmering and more noticeable banding, heck you could have full triliner filtering, turn up the LOD into the negatives and reproduce the same thing. Let's also remember that the source is some guy on a forum, which in terms of credibility rates pretty low. I would certainly trust Brent over some guy on a forum.

As for those calling for rebenches, HardOCP benched the 6800 with brilinear optimizations enabled, so in HardOCP's review both cards used trilinear optimizations, If you are going to complain you should complain to a site that didn't do this. But of course other sites benched with 61.11 which according to hanners can't disable brilinear so those sites are also optimized vs optimized.
 
Particleman said:
If you read his post he admits he changed the LOD, it's the LOD that causes the shimmering in those videos. It's the same problem that occurs when you use starstorm drivers or any other driver that alters the LOD, you end up with shimmering and more noticeable banding, heck you could have full triliner filtering, turn up the LOD into the negatives and reproduce the same thing.
Click to expand...

You are misunderstanding his post. What he is saying is that ATI's Brilinear filtering affects the LOD, not that he changed the LOD. He is saying that Brilinear filtering doing so can add to shimmering artifacts. You can read more about how brilinear filtering alters the LOD (with pictorial examples) here:

http://www.3dcenter.de/artikel/2003/10-26_b_english.php

(this article was written last year about the NV3x vs the R3xx and brilinear vs trilinear filtering)
 
creedAMD said:
well then why do YOU want him to change it. Are you afraid that you will forget what you know? You may need it for reference later? Why? How will it help you?
Click to expand...

lol who said i was afraid of anything? this just changes what i thought i knew. thats the whole point. and as he said...

a lot of people put faith and reference HardOCP's scores, so they might want to put a disclaimer up on those scores or something.
Click to expand...

i dont know how you confuse trying to get under someones skin with fairness.

creedAMD said:
Brent the reviews are great.
Click to expand...

and i really dont see, where you see that i said they were not great, ive said i liked their review even on other forums. if you have a problem with me, just take it to the pm. no need to go crazy over somthing so simple.
 
tranCendenZ said:
First off, as I'm sure you know, the brilinear filter introduced with the 9600 and now X800 has not been present in any other ATI card known to date AFAIK. It is a new, wholly different addition to what ATI has done in the past in terms of filtering.

That being said, I think what I'm getting to is that you make it seems like Nvidia's brilinear is worse than ATI's x800 best filtering quality, while the x800 is actually doing the same amount of bilinear filtering as the 6800's brilinear mode - and that the x800's best filtering is inferior to the NV40's and R350's best filtering. One of the german articles, I think the computerbase.de one, was read by one of the B3D members and found out that ATI's new Brilinear mode in the x800/9600 works like an on/off switch. The author of the article continued to create mips step by step approaching what would be a standard colored mip test, and he found a single mipmap threshold where the _entire scene_ switched from brilinear to full trilinear. This is not "adaptive" in any manner except disabling the new brilinear optimization for synthetic mipmap testing like an on/off switch. That info, combined with IXBTLABS article that shows the amount of bilinear filtering done between NV40 brilinear and ATI Brilinear, shows that ATI's best filter is not significantly better than Nvidia's brilinear filter on the NV40 - plus technically ATI's filter could cause artifacting due to underfiltering when compared to full trilinear on the NV40. Add on top of that movies such as the above which show filtering quality visibly inferior to that of the 9700PRO, and you think that would be something you want to address in your article and the performance results you obtained from the x800PRO/XT. As it stands, adaptive or not, you make it seem like the x800PRO is doing "true trilinear" by default when its really just doing brilinear during games, and trilinear during mipmap testing on every driver set available, with "true trilinear" being nowhere in sight. On the other hand, with the 60.72 drivers you benchmarked the NV40 with you actually can get full trilinear quality out of the NV40, but at a reduced performance of course - or you can use a similar quality brilinear filter and get the 10-20% boost that x800 cards are getting.

In other words, use a disclaimer to show that the x800pro is doing something different than the nv40 and past ati cards in the driver sets you benchmarked, namely forcing brilinear filtering and getting a significant performance boost because of this similar to the geforce fx series. And, based on statements from both companies, it appears Nvidia will provide Full Trilinear in their official drivers, while ATI seems opposed to the idea unless it can be "proven to be useful" by the consumer.

Or you could have just summed it up in one sentence as the post above ;)

I keep hearing this and I think it makes sense. If you can play 1600x1200 with 4xAA and 8xAF at 60fps+ in many games, then ATI users should have the option of upping the filtering quality another notch to full trilinear. However, if review sites focus on speed instead of driver options and image quality, ATI and Nvidia users probably won't have the option in the future. I mean, a lot of reviews, when I say a lot I mean most, put Nvidia's full trilinear filter up against ATI's Brilinear filter. Since both looked OK in the mips, most assumed both were full trilinear. Therefore Nvidia ended up looking slower in Ansiotropic filtering when it was in reality using a higher quality filter that was requested by gamers and review sites numerous times in the past. Reviewers could have used Nvidia's brilinear filter and gotten similar IQ and performance to ATI, but failed to do so thinking both were full trilinear based on mipmap examinations. So where does that leave Nvidia and ATI? The company that developed the optimization that was hidden best from reviewers ended up looking better in the benchmarks. Since Nvidia did not hide its Brilinear mode as well as ATI, it simply looked slower while it was in fact providing a full trilinear filter as opposed to a brilinear filter. My qualm is that if reviewers do not call companies on these practices, we will have these super high end cards that do insane resolutions but are lower quality IQ than previous generations simply so they can get the highest bar in performance.

And the person that loses out in that case is the end user, who has the fastest card on the planet that gets bested by a GeForce4 MX in filtering quality.
Click to expand...

I found this post, at best, a joke.

ATi did in fact use an adaptive algorithm since the 8500, and only started to use a *new* adaptive algorithm with 9600 series. Don't mock others if you don't even have your facts straight.

Why do you need a disclaimer? It is obvious if you can't tell the difference between Nvidia's trilinear or Ati's new adaptive algorithm, I see no wrong, nor a reason to complain to Ati/Nvidia. If Ati or Nvidia can improve performance while still maintaining the same in game image quality, more power to them.

I don't understand you people, Ati is giving you a performance increase and without sacrificing any image quality, instead of complaining about it, you should be thanking them. Maybe it's me, but I like increased performance...
 
Moofasa~ said:
I found this post, at best, a joke.

ATi did in fact use an adaptive algorithm since the 8500, and only started to use a *new* adaptive algorithm with 9600 series. Don't mock others if you don't even have your facts straight.

Why do you need a disclaimer? It is obvious if you can't tell the difference between Nvidia's trilinear or Ati's new adaptive algorithm, I see no wrong, nor a reason to complain to Ati/Nvidia. If Ati or Nvidia can improve performance while still maintaining the same in game image quality, more power to them.

I don't understand you people, Ati is giving you a performance increase and not scarifying any image quality, instead of complaining about it, you should be thanking them. Maybe it's me, but I like increased performance...
Click to expand...


I applaud ATi and nVidia for any increase in speed, without or with very little loss of IQ. But just to make me fell better, it would be nice to have a little button which gives me a choice between "adaptive" and "constant" filtering. Then I could see for myself if the performance gain is worth the IQ decrease (if any). I'm not sure if this is feasable, however.
 
he increased the gamma very high and turned on the flashlight to make it visible

in my playing the game at normal brightness i did not see shimmering
 
Bad_Boy said:
lol who said i was afraid of anything? this just changes what i thought i knew. thats the whole point. and as he said...



i dont know how you confuse trying to get under someones skin with fairness.



and i really dont see, where you see that i said they were not great, ive said i liked their review even on other forums. if you have a problem with me, just take it to the pm. no need to go crazy over somthing so simple.
Click to expand...

I don't have a problem with you, but let this go. Both companies are doing it, it's not able to be seen in the gameplay. That's what Hardocp reviews, the gameplay. Go complain to toms are anandtech who do the regular drone benchmarks.
 
Brent_Justice said:
he increased the gamma very high and turned on the flashlight to make it visible

in my playing the game at normal brightness i did not see shimmering
Click to expand...

On the mission "Fort" (I hope the US version uses the same mission title)the third-to-last save point is the exact spot. Enhance the gamma value since the scene is quite dark (the videos themselves have been created with the standard gamma setting). Also turn on the flashlights, otherwise you won't see anything.
Click to expand...

i think hes just saying that the videos are stock. but because the level is so dark you might need to enhance the gamma and maybe the flashlights. correct me if im wrong on this.

edit:
creedAMD said:
]
I don't have a problem with you, but let this go. Both companies are doing it, it's not able to be seen in the gameplay. That's what Hardocp reviews, the gameplay. Go complain to toms are anandtech who do the regular drone benchmarks.
Click to expand...

i think your missing the point of the videos, it isnt about the nvidia vs ati. its about the old ati vs new. hes seeing a negative difference between the 9700 he had and the x800 he has with the same settings. and this just reflects on the new ati vs nvidia in benchmarks. i just like fairness, is that wrong?

but ill let this go if you so see it fit.
 
Bad_Boy said:
i think hes just saying that the videos are stock. but because the level is so dark you might need to enhance the gamma and maybe the flashlights. correct me if im wrong on this.
Click to expand...

No, I got the impression he turned the gamma up to max, something no one would do during a real game, and turned on the flashlight. It doesn't mean you can't turn on your flash light, it simply means, if your looking for trouble, you will probably find it. Of course that is the impression I got...
 
Brent_Justice said:
he increased the gamma very high and turned on the flashlight to make it visible

in my playing the game at normal brightness i did not see shimmering
Click to expand...

That set of videos I posted here are at normal gamma levels, and actually show the artifacting better than the ones at high gamma levels he originally posted

He posted these first:

X800PRO (Gamma increased, 400x300)
http://www.reflectonreality.com/images/r420/FarCryx800ProAA_Divx.avi

R9700PRO (Gamma increased, 400x300)
http://www.reflectonreality.com/images/r420/FarCry9700Pro_Divx.avi

***
After he received complaints for turning up the gamma and halving the resolution, he reposted the movies with the normal gamma level and at full resolution, which are the original videos I posted and the ones I will post again below:

X800PRO (No gamma increase, 800x600)
http://www.reflectonreality.com/images/r420/FarCryx800ProHiRes_Divx.avi

R9700PRO (No gamma increase, 800x600)
http://www.reflectonreality.com/images/r420/FarCry9700ProHiRes_Divx.avi

So the videos I originally started this thread with did not, in fact, have increased gamma.
 
so if the videos are at stock gamma settings, same scene, but are different with the video cards, and still one has more shimmering than the other. does that not mean anything? ....you got me confused Moofasa~ lol
edit:

lol i thought this quote from beyond3d was worth posting.
ot: i know how fox mulder must have been feeling. The truth is out there but nobody will believe you
Click to expand...
 
also, increasing LOD, i.e. making it go move into the negative, Decreases Performance, at the same time it makes things more "Shimmery"

it seems unlikely ATI would increase LOD to decrease performance AND make things more shimmery

if they were gonna increase performance it would be to reduce LOD, to increase performance, which would make things more blurry, not shimmery
 
Brent_Justice said:
also, increasing LOD, i.e. making it go move into the negative, Decreases Performance, at the same time it makes things more "Shimmery"

it seems unlikely ATI would increase LOD to decrease performance AND make things more shimmery

if they were gonna increase performance it would be to reduce LOD, to increase performance, which would make things more blurry, not shimmery
Click to expand...

3dcenter.de has pictorial examples that show that Brilinear filtering can have an impact on LOD
http://www.3dcenter.de/artikel/2003/10-26_b_english.php
 
Brent_Justice said:
also, increasing LOD, i.e. making it go move into the negative, Decreases Performance, at the same time it makes things more "Shimmery"

it seems unlikely ATI would increase LOD to decrease performance AND make things more shimmery

if they were gonna increase performance it would be to reduce LOD, to increase performance, which would make things more blurry, not shimmery
Click to expand...

Quick Brent! Some guy on a forum has uncovered a scandal! Stop the presses!! It's back to the drawing board!!

Really guys. Complain to Ati if you think their quality isn't up to snuff. Most people that buy these cards don't even know what the hell trilinear or AF is. To get into such detail is nitpicking. Like I said, take it up with Ati. Start a petition.
 
tranCendenZ said:
That set of videos I posted here are at normal gamma levels, and actually show the artifacting better than the ones at high gamma levels he originally posted

He posted these first:

X800PRO (Gamma increased, 400x300)
http://www.reflectonreality.com/images/r420/FarCryx800ProAA_Divx.avi

R9700PRO (Gamma increased, 400x300)
http://www.reflectonreality.com/images/r420/FarCry9700Pro_Divx.avi

***
After he received complaints for turning up the gamma and halving the resolution, he reposted the movies with the normal gamma level and at full resolution, which are the original videos I posted and the ones I will post again below:

X800PRO (No gamma increase, 800x600)
http://www.reflectonreality.com/images/r420/FarCryx800ProHiRes_Divx.avi

R9700PRO (No gamma increase, 800x600)
http://www.reflectonreality.com/images/r420/FarCry9700ProHiRes_Divx.avi

So the videos I originally started this thread with did not, in fact, have increased gamma.
Click to expand...

Yea but why is no one else seeing this? I think I'll stick with Brent on this one. The truth is out there? Yea he is right, except the truth is that nothing has changed, image quality wise. I'll use my quote from before, if one goes out and looks for trouble, one find trouble (if someone spends all day trying to prove that Ati's new adaptive algorithm doesn't have the same in game image quality as the old adaptive algorithm has, he will find a way). Heck, I am sure I can find an example (a game) of my geforce 4 is better than my 9800 pro. But the fact is, Ati's new adaptive algorithm looks exactly the same as the old adaptive algorithm in *actual* gameplay. It is as simple as that.
 
I must be an idiot...because I prefer the "shimmering" effect in the X800 over the slighty more blurry 9700

Hey...did anyone ask this guy why he's playing at 800X600 with an X800?? hello?
 
TheGameguru said:
Hey...did anyone ask this guy why he's playing at 800X600 with an X800?? hello?
Click to expand...

Probably because that was the only res he could get a R9700PRO to have a decent framerate at to have an equal res comparison with the x800.
 
tranCendenZ said:
Probably because that was the only res he could get a R9700PRO to have a decent framerate at to have an equal res comparison with the x800.
Click to expand...

Doubtful.

Oh yes....let me compare my ti500 to oh..a 6800...I'm SURE I can find a difference....maybe? Yes? No?


:D

Until I see PROOF of something amiss.....there really was no real point to the thread to begin with.
 
Blackwind said:
Doubtful.

Oh yes....let me compare my ti500 to oh..a 6800...I'm SURE I can find a difference....maybe? Yes? No?
Click to expand...

Heh the sad thing is the ti500 arguably has better trilinear filtering than the 6800, assuming Nvidia makes angle-dependent filtering the only option on the 6800 :)

As for the 800x600 resolution, we can already see the 9700PRO is just starting to struggle in the video he posted in terms of framerate. If it went much higher it probably would have been a chopfest and useless in terms of comparison.

Until I see PROOF.....there really was no real point to the thread to begin with.
Click to expand...

What more proof do you want than a real in-game example that demonstrates proposed theory on the possible impact of brilinear filtering in some scenarios? That was the point of this thread.

Again, I don't think its a big deal and I personally wouldn't notice it in game, but I'm not going to pretend the difference doesn't exist either, nor am I going to say its the same as full trilinear. The option for full trilinear in ATI cards should be there for those who want it, especially considering how fast the card is.
 
You must log in or register to reply here.
Back
Top