Cowboy X
06-03-04, 11:50 PM
I had some views and questions on ATI's new adaptive filtering method ( now called "trylinear" )and initially started this thread : http://www.ocforums.com/showthread.php?t=297327&highlight=brilinear . It later became clear to me this was a good thing and I contentedly withdrew from wasting time in debates over it . In the last week several hardware sites have decided to 'educate' the masses on the issue . Some do a good job while the others are just so blinkered that it just isn't funny . So I decided to post my views on the matter after looking through the now wealth of reviews, editorials and tech papers on filtering . Is this a cheat , or a good optimisation ?
Firstly, unlike the stuff posted at some sites , there is nothing such as a defined standard way for performing trilinear filtering . A trilinearly filtered 3d scene is one in which the scene's mip map boundaries are not visible either statically or in motion . It should also remove the wave/moving line effect of bilinear filtering . The way that this additional filtering is added to bilinear filtering is as far as I see it , fairly standard .But what is important to note is that trilinear filtering is an effect or end state or result , not a method ( in my opinion , furthermore there are few definitions to be found which tie one into a given method eg full scene or not . ) Classically display drivers would filter the entire scene , not just the relevant mip maps boundaries and incur a performance hit ( bear in mind that trilinear filtering is already several times more intennsive than bilinear filtering ) . Thus the desire /need to optimise drivers . Many people have placed Nvidia's much maligned adaptive method in the same boat as ATI's but I think they are quite distinct and will deal with those differences now .
Nvidia's method , called "brilinear" filtering by many, has several problems . While offering a good performance boost and decent enough image quality for most gamers/reviewers , its main flaw stems from it not producing the trilinear effect ( emphasis on effect ). That is , brilinear in many cases didn't properly blend or filter the mip map boundaries , meaning that there are several games/situations where there is obvious decreased image quality . In some cases this decreased quality varies from barely/not noticeable to downright distracting in actual gameplay ( it shows up best in motion not static screenshots ). Visible mip map boundaries or transitions in gameplay = No trilinear filtering . That is why myself and others have no problem in refuting Nvidia's initial statements that it was true trilinear ( remember trilinear is an effect/result/endpoint ) . Nvidia first implemented this feature in UT2003 (yes, just as ATI is currently accused, Nvidia back then didn't inform a single reviewer ) and then put it in all D3D titles ( not sure about Open Gl ) . Is brilinear a bad thing ??? No ! At least not intrinsically .
This could be a fantastic speed optimisation for several gamers out there :
1/ Those who just want maximun framerate no matter what .
2/ Those who cant' see the differences in brilinear or consider brilinear's quality good enough for them .
3/ Those with a game/computer or both that cannot run at a preferred resoloution or setting with 'full' trilinear enabled at their preferred framerate . This is currently very important in the low-end and will be important in the future as today's FX cards move from highend to midrange to the low-end . It therefore can add prolonged life to such cards in the future at a quality above plain bilinear filtering .
But there are a few major problems :
1/ There was no 'turn it off 'feature in the drivers . Thus true mip map boundary blending was not available ( trilinear ) .
2/ Nvidia's drivers happily told you that you or your game selected trilinear , but in fact it wasn't turned on at all .
3/ Nvidia and some reviewers initially denied that the quality was less .
4/ Eventually Nvidia admitted that the quality was lower and promised us via Kyle at [H] that true trilinear would be enabled soon . The truth is that up to now all geforce FX users are stuck with this reduced quality . Even when card's can handle it , trilinear is not an option ! I am certain that there are many games which a 5800 to 5950u level card could happily handle with trilinear on , but the option is not there . You must then ask , is this there to benefit the gamer , or benchmarks and the company's bottom line ?
The difference with ATI's adaptive method now called "trylinear" is pretty simple . Common sense dictates that the only thing which will make the boundary between mip map A + B visible is if there are differences or gaps between the two . What ATI claims to do is to calculate those differences and blend/filter the boundaries based on how different they are from each other . So if A + B are grossly different and will be visible to the gamer then maximum filtering is necessary and will be applied to prevent the boundary from being apparent . But in the same scene maps D + E are quite similar at their boundary therefore less filtering is needed to hide the edges . This method is intelligent , interesting and efficient if not controversial to some . So far it has been shown to result in properly blended mip map boundaries in gameplay . No visible mip map boundaries or transitions in gameplay = Trilinear Filtering I see it as that simple , but maybe that's just me .
I want to take this paragraph to now comment on the issue of coloured mip maps . Some titles like COD , Quake3 and UT2003show X800 performance losses with coloured mip map boundaries . The colouring of mip maps is important to developers and recently gamers when it helped prove the brilinear nature of Nvidia's filtering last year . For those who don't know the colouring of mip maps looks like this : http://hardocp.com/images/articles/10817474486qLMOmeutS_3_6.jpg http://www.tomshardware.com/graphic/20040603/images/pic12.jpg . A properly filtered border ( trilinear ) will merge with the next making it very difficult to see where one colour or mip map ends or begins relative to the other . Many surmised that ATI's performance loss was due to ATI's driver detecting that the gamer/reviewer had turned on coloured mip maps to check filtering quality . They then accused the driver of turning off the optimisation to render the older 'full scene' trilinear to fool the public and thus they called it a cheat .Wht is really happening here is that the sharply and distinctly coloured mip maps will present to the ATI driver as a situation where the boundaries have the maximum possible differences . As a result maximum filtering will be needed on every boundary not just some . Therefore performance must logically drop . I don't see the big conspiracy here , although several more learned and experienced people than I are reading alot more into it . One site (THG ) even quotes John Carmack's early view on the matter :
" This is indeed a "cheat" that both major vendors now do. Instead of always sampling the two adjacent mip map levels and doing a full blend between them, they have plateaus where only a single mip level is sampled, reducing the average samples from 8 to about 6.
It is actually a pretty sensible performance enhancement, with minimal visual issues. However, having drivers analyze mip map uploads to hide the cheat is an unfortunate consequence.
The colored mip map option in Q3 should have absolutely zero performance impact in the absence of performance options like this.
John Carmack "
But sadly this quote shows that Mr.Carmack was at the time not informmed about the situation with ATI's driver and thus it was a statement made in a vaccuum . Yet it has still been posted after some of those issues were answered ............... why ? Others have been keen to point out where ATI advised the use of coloured mip maps and the turning off of Nvidia's brilinear as further examples of conspiracy to unfairly use ATI's brand of brilinear ( trylinear) versus the 6800u's 'full' trilinear in reviews . This is partly based on the assumption that ATI is specifically detecting coloured mip maps which as I said earlier sounds pretty unlikely . But more importantly these conspiracy theorists have yet to point out any place where gameplay visuals are negatively affected . There are also yet to disprove ATI's assertion of achieving the trilinear effect with their new method . So like a a game of Jenga that argumment falls .
ATI however does not come out of this smelling like a rose . Just as with the Quack issue back in the day , ATI has in my opinion failed to handle the situation well , although they may have been in one of those damned if you do and damned if you don't siuations . Communication is key :
1/ They didn't properly explain to reviewers or even developers the differences in the x800's ( and 9600's ) filtering method . In fact you would have to dig very deep to find some of the statements by ATI emmployees stating that the filtering method would be changing with this new generation .
2/ the argument that they are awaiting a patent is plausible but I will have to wait and see . Either way too little info was given .
3/ Some of their subsequent statements on the issue have not been clear enough .
Finally , in my opinion , ATI's new adaptive filtering method is a good thing . Until such time as someone shows instances where the algorithm visibly fails , I cannot wait to use such tech this year ( if the price is right ) . If it does fail in certain software/situations , then as with nvidia's brilinear I will demand like others that ATI provide a 'turn it off ' option . This would allow me to chose the older 'full scene' method in instances where it looks better and/or performance loss isn't an issue . I believe that is reasonable for all gamers like me . But for others .................... they just seem to have too mch time on their hands ( or axes to grind ) . I'm going now to put on my asbestos suit and await replies .
/end rant/article .
Cowboy X
This article is my opinion so I expect people to differ , argue and even flame , in all areas especial technical ones I welcome any criticism or better yet education . This is not necessarily an official view of Overclockers.com :)
Firstly, unlike the stuff posted at some sites , there is nothing such as a defined standard way for performing trilinear filtering . A trilinearly filtered 3d scene is one in which the scene's mip map boundaries are not visible either statically or in motion . It should also remove the wave/moving line effect of bilinear filtering . The way that this additional filtering is added to bilinear filtering is as far as I see it , fairly standard .But what is important to note is that trilinear filtering is an effect or end state or result , not a method ( in my opinion , furthermore there are few definitions to be found which tie one into a given method eg full scene or not . ) Classically display drivers would filter the entire scene , not just the relevant mip maps boundaries and incur a performance hit ( bear in mind that trilinear filtering is already several times more intennsive than bilinear filtering ) . Thus the desire /need to optimise drivers . Many people have placed Nvidia's much maligned adaptive method in the same boat as ATI's but I think they are quite distinct and will deal with those differences now .
Nvidia's method , called "brilinear" filtering by many, has several problems . While offering a good performance boost and decent enough image quality for most gamers/reviewers , its main flaw stems from it not producing the trilinear effect ( emphasis on effect ). That is , brilinear in many cases didn't properly blend or filter the mip map boundaries , meaning that there are several games/situations where there is obvious decreased image quality . In some cases this decreased quality varies from barely/not noticeable to downright distracting in actual gameplay ( it shows up best in motion not static screenshots ). Visible mip map boundaries or transitions in gameplay = No trilinear filtering . That is why myself and others have no problem in refuting Nvidia's initial statements that it was true trilinear ( remember trilinear is an effect/result/endpoint ) . Nvidia first implemented this feature in UT2003 (yes, just as ATI is currently accused, Nvidia back then didn't inform a single reviewer ) and then put it in all D3D titles ( not sure about Open Gl ) . Is brilinear a bad thing ??? No ! At least not intrinsically .
This could be a fantastic speed optimisation for several gamers out there :
1/ Those who just want maximun framerate no matter what .
2/ Those who cant' see the differences in brilinear or consider brilinear's quality good enough for them .
3/ Those with a game/computer or both that cannot run at a preferred resoloution or setting with 'full' trilinear enabled at their preferred framerate . This is currently very important in the low-end and will be important in the future as today's FX cards move from highend to midrange to the low-end . It therefore can add prolonged life to such cards in the future at a quality above plain bilinear filtering .
But there are a few major problems :
1/ There was no 'turn it off 'feature in the drivers . Thus true mip map boundary blending was not available ( trilinear ) .
2/ Nvidia's drivers happily told you that you or your game selected trilinear , but in fact it wasn't turned on at all .
3/ Nvidia and some reviewers initially denied that the quality was less .
4/ Eventually Nvidia admitted that the quality was lower and promised us via Kyle at [H] that true trilinear would be enabled soon . The truth is that up to now all geforce FX users are stuck with this reduced quality . Even when card's can handle it , trilinear is not an option ! I am certain that there are many games which a 5800 to 5950u level card could happily handle with trilinear on , but the option is not there . You must then ask , is this there to benefit the gamer , or benchmarks and the company's bottom line ?
The difference with ATI's adaptive method now called "trylinear" is pretty simple . Common sense dictates that the only thing which will make the boundary between mip map A + B visible is if there are differences or gaps between the two . What ATI claims to do is to calculate those differences and blend/filter the boundaries based on how different they are from each other . So if A + B are grossly different and will be visible to the gamer then maximum filtering is necessary and will be applied to prevent the boundary from being apparent . But in the same scene maps D + E are quite similar at their boundary therefore less filtering is needed to hide the edges . This method is intelligent , interesting and efficient if not controversial to some . So far it has been shown to result in properly blended mip map boundaries in gameplay . No visible mip map boundaries or transitions in gameplay = Trilinear Filtering I see it as that simple , but maybe that's just me .
I want to take this paragraph to now comment on the issue of coloured mip maps . Some titles like COD , Quake3 and UT2003show X800 performance losses with coloured mip map boundaries . The colouring of mip maps is important to developers and recently gamers when it helped prove the brilinear nature of Nvidia's filtering last year . For those who don't know the colouring of mip maps looks like this : http://hardocp.com/images/articles/10817474486qLMOmeutS_3_6.jpg http://www.tomshardware.com/graphic/20040603/images/pic12.jpg . A properly filtered border ( trilinear ) will merge with the next making it very difficult to see where one colour or mip map ends or begins relative to the other . Many surmised that ATI's performance loss was due to ATI's driver detecting that the gamer/reviewer had turned on coloured mip maps to check filtering quality . They then accused the driver of turning off the optimisation to render the older 'full scene' trilinear to fool the public and thus they called it a cheat .Wht is really happening here is that the sharply and distinctly coloured mip maps will present to the ATI driver as a situation where the boundaries have the maximum possible differences . As a result maximum filtering will be needed on every boundary not just some . Therefore performance must logically drop . I don't see the big conspiracy here , although several more learned and experienced people than I are reading alot more into it . One site (THG ) even quotes John Carmack's early view on the matter :
" This is indeed a "cheat" that both major vendors now do. Instead of always sampling the two adjacent mip map levels and doing a full blend between them, they have plateaus where only a single mip level is sampled, reducing the average samples from 8 to about 6.
It is actually a pretty sensible performance enhancement, with minimal visual issues. However, having drivers analyze mip map uploads to hide the cheat is an unfortunate consequence.
The colored mip map option in Q3 should have absolutely zero performance impact in the absence of performance options like this.
John Carmack "
But sadly this quote shows that Mr.Carmack was at the time not informmed about the situation with ATI's driver and thus it was a statement made in a vaccuum . Yet it has still been posted after some of those issues were answered ............... why ? Others have been keen to point out where ATI advised the use of coloured mip maps and the turning off of Nvidia's brilinear as further examples of conspiracy to unfairly use ATI's brand of brilinear ( trylinear) versus the 6800u's 'full' trilinear in reviews . This is partly based on the assumption that ATI is specifically detecting coloured mip maps which as I said earlier sounds pretty unlikely . But more importantly these conspiracy theorists have yet to point out any place where gameplay visuals are negatively affected . There are also yet to disprove ATI's assertion of achieving the trilinear effect with their new method . So like a a game of Jenga that argumment falls .
ATI however does not come out of this smelling like a rose . Just as with the Quack issue back in the day , ATI has in my opinion failed to handle the situation well , although they may have been in one of those damned if you do and damned if you don't siuations . Communication is key :
1/ They didn't properly explain to reviewers or even developers the differences in the x800's ( and 9600's ) filtering method . In fact you would have to dig very deep to find some of the statements by ATI emmployees stating that the filtering method would be changing with this new generation .
2/ the argument that they are awaiting a patent is plausible but I will have to wait and see . Either way too little info was given .
3/ Some of their subsequent statements on the issue have not been clear enough .
Finally , in my opinion , ATI's new adaptive filtering method is a good thing . Until such time as someone shows instances where the algorithm visibly fails , I cannot wait to use such tech this year ( if the price is right ) . If it does fail in certain software/situations , then as with nvidia's brilinear I will demand like others that ATI provide a 'turn it off ' option . This would allow me to chose the older 'full scene' method in instances where it looks better and/or performance loss isn't an issue . I believe that is reasonable for all gamers like me . But for others .................... they just seem to have too mch time on their hands ( or axes to grind ) . I'm going now to put on my asbestos suit and await replies .
/end rant/article .
Cowboy X
This article is my opinion so I expect people to differ , argue and even flame , in all areas especial technical ones I welcome any criticism or better yet education . This is not necessarily an official view of Overclockers.com :)