Hellbinder
03-10-03, 05:42 PM
This was PM'd to me by CCCmaurader. So all thanks to him. Its a reply he recieved from the technical editor of MaxPC from a Question sent in about the off numbers in their review of the 9800pro.
Anisotropic filtering…
nVidia has three different performance settings, while ATI has two. ATI’s default is quality vs. performance, which is a compatibility thing for older cards (like the Radeon 8500) that couldn’t use both trilinear filtering and aniso at the same time. It uses the same algorithm for both if the card supports it. nVidia’s uses three different paths through the pipeline depending on the setting in the drivers. It’s default setting is “balanced”. In the future, I’ll probably run the quality nVidia setting, but I didn’t know about it until after that issue went to press. The difference should be less than 5% though.
///Will
--
Technical Editor
Maximum PC
Thus MxPC's FX/9800 comparison was indeed run in ballanced mode for the FX and Quality mode for the 9800pro. Which is undoubtedly being repreated in several other reviews where the numbers seem off or outright Ridiculous.
It Further illustrates that these guys have not one CLUE about the hardware they are testing and Further supports some of DoomTroopers comments about Journalistic integrity.
1. The FX supposedly uses "3 different paths through the pipeline" depending on what mode is selected??
I mean.. what a bunch of total nonsense.. its trying to make it sound like its some special and technical feat to use Bilinear filtering instead of TRilinear.. :rolleyes:
2. ATI’s default is quality vs. performance, which is a compatibility thing for older cards (like the Radeon 8500) that couldn’t use both trilinear filtering and aniso at the same time. It uses the same algorithm for both if the card supports it.
Give me a break.. First of all its not the same algorithm at all. This has been shown by review sites countless times. The R300's Algorithm is able to handle 45' angles just fine. Second its not a "compatability" thing.
Most importantly the idiot does not even understand that the Nvidia Ballanced is nearly the EXACT SAME THING as ATi's performance. Both use Bilinear sampled AF.
3. In the future, I’ll probably run the quality nVidia setting, but I didn’t know about it until after that issue went to press. The difference should be less than 5% though
What??? 5% difference? :rolleyes:
Just another example of how generally speaking Reviewers and editors are biased and uneducated morons. Who form their opinions on whatever Nvidias little PR documents tell them to. None of them even begin to have the technical knowledge to post a real representative review. For which there is no excuse.
Two years ago I had a couple ideas about 3dgraphics, had some opinions etc.. And if you search out my oooold posts you will see that i was not as knowledgable or informed as I am now. What happened?? I went and Read countless articles and White papers, I made Contacts with people in the industry who work for various companies, i even started to do some 3d programming. In other words I Educated, and better Equiped myself for understanding and debating 3D hardware and Graphics. I still have a long way to go, but I am FAR more Equiped than I was.
This is no different than what these reviewers and editors SHOULD be doing at these Review sites and Magazines. Especially the ones that have been there for 2-4 years. Its completely Ridiculous that after a couple years these guys STILL have no clue, and post whatever is spoon Fed them.
Anisotropic filtering…
nVidia has three different performance settings, while ATI has two. ATI’s default is quality vs. performance, which is a compatibility thing for older cards (like the Radeon 8500) that couldn’t use both trilinear filtering and aniso at the same time. It uses the same algorithm for both if the card supports it. nVidia’s uses three different paths through the pipeline depending on the setting in the drivers. It’s default setting is “balanced”. In the future, I’ll probably run the quality nVidia setting, but I didn’t know about it until after that issue went to press. The difference should be less than 5% though.
///Will
--
Technical Editor
Maximum PC
Thus MxPC's FX/9800 comparison was indeed run in ballanced mode for the FX and Quality mode for the 9800pro. Which is undoubtedly being repreated in several other reviews where the numbers seem off or outright Ridiculous.
It Further illustrates that these guys have not one CLUE about the hardware they are testing and Further supports some of DoomTroopers comments about Journalistic integrity.
1. The FX supposedly uses "3 different paths through the pipeline" depending on what mode is selected??
I mean.. what a bunch of total nonsense.. its trying to make it sound like its some special and technical feat to use Bilinear filtering instead of TRilinear.. :rolleyes:
2. ATI’s default is quality vs. performance, which is a compatibility thing for older cards (like the Radeon 8500) that couldn’t use both trilinear filtering and aniso at the same time. It uses the same algorithm for both if the card supports it.
Give me a break.. First of all its not the same algorithm at all. This has been shown by review sites countless times. The R300's Algorithm is able to handle 45' angles just fine. Second its not a "compatability" thing.
Most importantly the idiot does not even understand that the Nvidia Ballanced is nearly the EXACT SAME THING as ATi's performance. Both use Bilinear sampled AF.
3. In the future, I’ll probably run the quality nVidia setting, but I didn’t know about it until after that issue went to press. The difference should be less than 5% though
What??? 5% difference? :rolleyes:
Just another example of how generally speaking Reviewers and editors are biased and uneducated morons. Who form their opinions on whatever Nvidias little PR documents tell them to. None of them even begin to have the technical knowledge to post a real representative review. For which there is no excuse.
Two years ago I had a couple ideas about 3dgraphics, had some opinions etc.. And if you search out my oooold posts you will see that i was not as knowledgable or informed as I am now. What happened?? I went and Read countless articles and White papers, I made Contacts with people in the industry who work for various companies, i even started to do some 3d programming. In other words I Educated, and better Equiped myself for understanding and debating 3D hardware and Graphics. I still have a long way to go, but I am FAR more Equiped than I was.
This is no different than what these reviewers and editors SHOULD be doing at these Review sites and Magazines. Especially the ones that have been there for 2-4 years. Its completely Ridiculous that after a couple years these guys STILL have no clue, and post whatever is spoon Fed them.