ATi is ch**t**g in Filtering

It's called TWIMTBP.

 

Tiger Woods 2004 is a TWIMTBP game and the 9800P/9800XT run absolute circles around the 5950FX, not to mention many EA Sports games run better on R300/360 hardware than FX hardware so that theory sounds okay, but in reality is not a factor in some cases....

 

..but it's always a good excuse for some if it is not that way

 

Sad but oh so true....

 

I am confused as to one issue (and very late in asking), is the 5950 running trilinear as stated or is it running brilinear. If anyone wants to see whats really going on with these images, in photoshop select the whole image then use the equalize command (equalize entire image based on selected area). It makes it much easier to spot the filtering.

 

In this case, trilinear means "whatever the card uses when trilinear is selected."

 

I guess what I am asking is are Nvidia's trilinear optimizations disabled?

Becase the x800 appears to be doing slighty more filtering than the 5950, but noticibly less than the 9800.

If the 5950 is considered to be doing full trilinear in the images then I think there is no issue here, if it is doing brilinear then the x800 should be considered brilinear too.

 

There is no way to disable the "trilinear optimizations" on NV3x. It is All Brilinear, All The Time.

 

Thx, got to say I am disappointed.

 

Ha!

ATi's new stuff is always trilyinicious. 8)

 

Dig, don't be The F-Word. It makes me angry, and I don't think you'd like me when I'm angry... Yes, I become Howard Dean.

It's not trilinear. It may look close, but it's not trilinear. I'm not about to go and say, "Well, ATI, you may not be doing trilinear when I want you to and specifically request it, but GOOD ENOUGH! YOU'RE OFF THE HOOK!" after my brilinear tirades of the past year.

 

Let's try this one more time, shall we? I'll even help ya out Baron and add some extry emphasis to the bits that should have clued you in that I was bashing ATi with me post.

Ahem...

"Ha!

ATi's new stuff is always trilyinicious. 8) "

HINT: It ain't a typo or accidental mis-spelling...

 

42 pages?
Must be alot to talk about.
It will be a eye opener to see what the next drivers do.
I hope they put full tri bake in on all 7 stages and still have the "try" as a choise.

 

Is that mean it is really all the time or only with aniso filtering?
And what about ati's new filtering solution?

 

No. I'm not saying that at all. I'm saying that it will happen by accident. In other words, I don't be believe that ATI's solution could possibly handle all corner cases.

 

Of course you don't, I wouldn't expect anything less. I like your philosophy: Guilty until proven innocent. There's no indication of any problems with what ATI is doing, yet you're convinced there is because we haven't seen such an example.

-FUDie

 

yes, whereas if you recall the (nvidia) brilinear days, what was his story then?

 

Would this be like the problem you say we're likely to see wrt to textures and ATI's 24 bit precision?

...

Anyways, you did catch the comment from ATI that they wanted feedback to improve their algorithm (which afaik is software based)?
Thus even if what you're saying really does happen ATI can fix it with a driver-update.

So I don't really see what your problem with this is.

 

That it can't be disabled. That nobody was told about the optimization until it was discovered. That ATI has yet to make any statement that actually sheds any light on exactly what is done.

 

Bah, yes I can see that problem, but that was not what I was talking about...

You do not adress that problem in these posts, which are the ones I quoted and thus replied to.

But I'll rephrase my previous reply:

Would this be like the problem you say we're likely to see wrt to textures and ATI's 24 bit precision?

...

Anyways, you did catch the comment from ATI that they wanted feedback to improve their algorithm (which afaik is software based)?
Thus even if what you're saying really does happen ATI can fix it with a driver-update.

So why should that (a potential new texture going against what the current algorithm is checking for) be such a big problem as you make it out ot be, since it (the algorithm) can be modified?