"Trilinear" Filtering in Motion

[MikeC]

I've posted two 5MB video clips.

One is from a GeForce FX 5900 Ultra using the 44.03 Detonator drivers:

The other is from a Radeon 9800 Pro using the 3.6 Catalyst drivers:

Video was taken from UT2003 at 800x600 using FRAPS and compressed in Microsoft's MPEG4 format.

Download the Codec here for Windows Media Player -
http://www.cabcconnection.com/support/CodecDownLoad.asp


http://www.nvnews.net/temp/1.avi

http://www.nvnews.net/temp/2.avi


UseTrilinear=True and LevelOfAnisotropy=1 in UT2003.ini file.

MoveForward speed in the User.ini file was changed from the default of +300 to +1 when capturing the video.

I also have video that was taken with 8X anisotropic filtering enabled on both graphics cards.

 

[Nick]

Damn, trilinear filtering certainly raised your frags a lot!

 

[Slides]

2nd one is ATI?

 

[Pete]

Are we talking driver set to App Pref and trilinear forced via UT2K3 ini file? Both had fairly obvious MIP-maps, but the second had a very harsh MIP-map transition between the first and second MIP-maps.

As an aside, I honestly don't think I've noticed MIP-map transitions with my 9100. I have 16xAF on all the time, and it probably pushes the MIP-maps back so far that it's not easily noticable in-game (mainly Wolf:ET). I do notice texture shimmer in some games (the Tiger Woods 2003 demo), though I don't know if that's ATi's AF's fault, or LOD bias, or the game itself.

 

[MikeC]

Are we talking driver set to App Pref and trilinear forced via UT2K3 ini file?

Yes, I just completed a major re-write of my post

Glad you were able to spot the differences in the videos!

 

[nate240]

The first video produces a smooth transition from sharp to blurry, while the second has two noticable transitions from sharp to blurry. If I had to guess, I'd say the first is the radeon, the second the FX. Interesting to say the least if both were tested using the "same" settings.....

 

[K.I.L.E.R]

I can only see a pixelated mess.

 

[swanlee]

I'll have to disagree
The first seems to transition much smoother from clear to blurry while the second one has almost a straight line of blur across the screen around the cross hair.

I think the first one is radeon and the second is an FX. If I'm wrong more props to Nvidia as that straight line of blur across the cross hair is quite annoying

 

[Reverend]

Um... what's next? The "noticeability of high-rez textures vs low-rez textures when playing games"?

Talk about trying to prove the obvious.

No, wait, "obvious" is a subjective word apparently! ......

 

[Slides]

Watched it again, and yeah first one does look more smoother, so it must be ATI.

 

[MikeC]

Talk about trying to prove the obvious.

It's not obvious to some people until you explain to them what to look for.

 

[StealthHawk]

Um... what's next? The "noticeability of high-rez textures vs low-rez textures when playing games"?

Talk about trying to prove the obvious.

No, wait, "obvious" is a subjective word apparently! ......

It's not obvious because some people don't want to see the truth. Like our friends over at [H] who have their readers convinced that there is no quality reduction. In fact, they say there is no quality reduction in motion either! And people believe them.

Which from the look of these videos, is utterly ridiculous. Sure, they were grainy, but video 2 looks like bilinear to me.

 

[madshi]

Hey Mike,

that's a good one, thanks for doing this! I'm impressed these days about how you guys at nvnews handle the situation. Most of you see the truth and say it. My hat's off to you... Btw, did you post this at HardOCP?

I see no mipmap lines in the first video, but I do see at least one crystal clear mipmap line in the second one, which I find very annoying.

Brent, are you listening? Now I'd *really* like to hear your opinion about this.

 

[K.I.L.E.R]

Um... what's next? The "noticeability of high-rez textures vs low-rez textures when playing games"?

Talk about trying to prove the obvious.

No, wait, "obvious" is a subjective word apparently! ......

Dude, is this an indirect attack?
Your use of sarcasm is mildly offensive.

MikeC has done nothing to deserve that. Even if he has done something to deserve it, who are you to judge that?

 

[Reverend]

Oh no, don't get me wrong! I think what Mike's doing is worth his time. I never mentioned "Mike" in my post, dude (... I hate people calling me "dude" when they know my name).

I'm just amazed about this situation, that's all. I never imagined it would be hard to differentiate trilinear from non-trilinear.

 

[K.I.L.E.R]

Reverend(aka Dude ), it's easy to tell them a part if you have a good quality video or actually try both trilinear/bilinear in game.

That video looks like a pixelated mess, I can barely tell the floor from the weapon.

 

[MikeC]

That video looks like a pixelated mess, I can barely tell the floor from the weapon.

The video was taken at a resolution of 800x600, so be sure your desktop resolution is at 1024x768 or higher during playback. If it's played back at a resolution lower than 800x600, you'll probably get distortion.

Here's a screenshot I took during playback. Keep in mind the screenshot is a low quality JPG.

 

[Doomtrooper]

Mike thanks for pointing out the obvious, since I got banned on [H] for trying to explain the differences in motion it is good too see you exposing the truth to the public.

And supporting the bandwidth for these movies too, you should look for a new domain name and get into a general 3Dgraphics site like here.
Your review quality and methods are well done.

 

[K.I.L.E.R]

Mike, I just need to download some new video codecs.

I discussed it all here:
http://www.beyond3d.com/forum/viewtopic.php?t=7069

I discussed the differences in motion here. I also stated that 1 stage trilinear filtering makes a very large difference over pure bilinear.

 

[nelg]

Mike, I guess there is an interesting lesson here for Kyle. If you are going to argue with someone, it's best not to do it with one who has a website.