ATi is ch**t**g in Filtering

[Lezmaka]

if i steal something from you and give you back, for free, an as much worthing thing, looking the same, what did i take from you?

if i take you a one dollar note, and give you another one, does it hurt you?

there is nothing stolen. only replaced. by an equal performing thing

I don't see how any of that really relates to the issue here, but here's soemthing to think about if you believe that.

A lot of job applications for retail store sover here in the US ask questions like that. Some of them are like this:

An employee who has worked for the company for 10 years, takes $300 from the stores safe, at the end of each month to pay rent, but later replaces it. The manager sees her putting the money back one day and fires her. Do you think the manager took the correct action?

The correct answer (according to how the business views the question) is yes, the manager did the correct thing by firing the emoployee. You would probably say that the employee wasn't hurting the business because the money was replaced, but that doesn't matter. I believe that taking something that doesn't belong to you without the other persons consent, even if you replace it, is not right.

But again, I don't see how your argument applies the this issue.

 

[Lezmaka]

Unlike having something real stolen from you.

People please, stop with the world ending dramatics.

Hey if you pay for somethign and it doesn't do what they advertise it does I'd call that stealing.

No, that's not stealing. That would be false advertising, there's a difference.

Theft (a synonym of theft is steal)

A criminal taking of the property or services of another without consent

False advertising

Definition
: the crime or tort of publishing, broadcasting, or otherwise publicly distributing an advertisement that contains an untrue, misleading, or deceptive representation or statement which was made knowingly or recklessly and with the intent to promote the sale of property, goods, or services to the public

 

[davepermen]

I don't see how any of that really relates to the issue here, but here's soemthing to think about if you believe that.

A lot of job applications for retail store sover here in the US ask questions like that. Some of them are like this:

An employee who has worked for the company for 10 years, takes $300 from the stores safe, at the end of each month to pay rent, but later replaces it. The manager sees her putting the money back one day and fires her. Do you think the manager took the correct action?

The correct answer (according to how the business views the question) is yes, the manager did the correct thing by firing the emoployee. You would probably say that the employee wasn't hurting the business because the money was replaced, but that doesn't matter. I believe that taking something that doesn't belong to you without the other persons consent, even if you replace it, is not right.

But again, I don't see how your argument applies the this issue.

this is too offtopic. all i was replying to, is, that stealing is the wrong word. they changed something, and thats all. and its something to not bother about. but some people will not understand this, till pr tells them its SMARTFILTER..

 

[weeds]

www.ixbt.com/video2/nv40-rx800-3.shtml

It's in russian, but screenshots are universal and utilitie itself is in English.

The main idea is: rv3x0/r420 are using full trilinear for static and dynamic DX textures, optimized trilinear is used for managed textures. And there certainly IS a difference in quality.

(The article comparing trilinear in real world games on r360 and r420 should be ready soon too.)

Thanx for the link , So IQ ISN'T the same as full Trilinear.. Interesting..

 

[christoph]

but some people will not understand this, till pr tells them its SMARTFILTER..

yep, its seems to be smart enough to properly distinguish wether the mipmaps are auto generated or not

 

[Suspicious]

nothing is being stolen - if you can't see it!

So if you are blind and I take the money from your table it is ok because you didn't see me? I guess you would also approve rendering in grayscale for color-blind folks? After all, it would make terrific improvements in rendering speed -- as much as 4 times the FPS.

if i take you a one dollar note, and give you another one, does it hurt you?

Stealing perhaps wasn't the right word (my native language is not English). I was referring to the fact that they removed full trilinear filtering behind our back and replaced it with fake, so the answer to your quetion may be: yes, it hurts me, because in this case you give me back forged one dollar bill.

prove to me it doesn't do what they advertise it does?

OK I will, it advertises FULL TRILINEAR ALL THE TIME and IT DOES NOT DO IT ALL THE TIME. Who gives the $hite if it is even visible or not?

Of course, now that you have seen it for yourself my arguments are irrelevant. ATI <c h e a t e d>, I am disappointed and I am sure many others are too. Hopefully, lessons are learned -- no company should be trusted and praised, case closed.

 

[AlphaWolf]

Who gives the $hite if it is even visible or not?

I have news for you, it's all an optical illusion. Your display is only capable displaying 2d images. You have been cheated all along.

 

[Suspicious]

I have news for you, it's all an optical illusion. Your display is only capable displaying 2d images. You have been cheated all along.

True. And Carlos Castaneda would say that "optical" itself is an illusion because it is only a filtered perception of real world energy.

 

[bloodbob]

If anyone hasn't checked this out yet ( posted on the previous page ) please check out this link http://www.ixbt.com/video2/nv40-rx800-3.shtml it has some Quantitative measurements of the optimisation for both nvidia and ATI. Because it is russian you may wish to use a web transaltor like http://www.freetranslation.com/web.asp .

 

[EricVonZipper]

Fits this thread perfectly

...lifted from http://www.ataricommunity.com/forum...528861dee36857e6274bef92f&threadid=382345


this also works......

How many Beyond3d forum members does it take to change a light bulb....

1 to change the light bulb and to post that the light bulb has been changed.

14 to share similar experiences of changing light bulbs and how the light bulb could have been changed differently.

7 to caution about the dangers of changing light bulbs.

1 to move it to the Lighting section.

2 to argue then move it to the Electricals section.

7 to point out spelling/grammar errors in posts about changing light bulbs

5 to flame the spell checkers.

3 to correct spelling/grammar flames.

6 to argue over whether it's "lightbulb" or "light bulb" ... another 6 to condemn those 6 as stupid

2 industry professionals to inform the group that the proper term is "lamp"

15 know-it-alls who claim they were in the industry, and that "light bulb" is perfectly correct.

19 to post that this forum is not about light bulbs and to please take this discussion to a lightbulb forum.

11 to defend the posting to this forum saying that we all use light bulbs and therefore the posts are relevant to this forum.

36 to debate which method of changing light bulbs is superior, where to buy the best light bulbs, what brand of light bulbs work best for this technique and what brands are faulty.

7 to post URLs where one can see examples of different light bulbs.

4 to post that the URL's were posted incorrectly and then post the corrected URLs.

3 to post about links they found from the URLs that are relevant to this group which makes light bulbs relevant to this group.

13 to link all posts to date, quote them in their entirety including all headers and signatures, and add "Me too".

5 to post to the group that they will no longer post because they cannot handle the light bulb controversy.

2 to roll their eyes.

1 moderator to lock the thread.

3 to tell him that this is not acceptable and 1 moderator to open it.

4 to say "didn't we go through this already a short time ago?"

13 to say "do a Google search on light bulbs before posting questions about light bulbs".

1 forum lurker to respond to the original post 6 months from now and start it all over again.

 

[cthellis42]

Lightbulb, of course. It's a perfectly valid word, so why complicate it?

 

[bloodbob]

Lightbulb, of course. It's a perfectly valid word, so why complicate it?

My word grammar checker says that "Lightbulb, of course." is a fragemented sentence. You probably should have said "The correct answer is lightbulb of course."

LOL

Getting back to reality I would like to see some FPS graphs like the ones clootie did but for a bunch of different textures of the same size. This would tell us if ATI's method really is adaptive for the "contents" of the texture rather then only on the methods of the mipmap generation.

 

[davepermen]

Stealing perhaps wasn't the right word (my native language is not English). I was referring to the fact that they removed full trilinear filtering behind our back and replaced it with fake, so the answer to your quetion may be: yes, it hurts me, because in this case you give me back forged one dollar bill.

i'm not speaking english at my living place as well. they don't replace it with fake. they replace it with an OPTIMIZED version. you know programming? then you should not that optimized does not mean non-equal. it just means bether designed for a certain case.

but it looks like you don't want to get this. if it doesn't hurt, why are you crying?

 

[Suspicious]

they don't replace it with fake. they replace it with an OPTIMIZED version.

I disagree. Read on.

you know programming? then you should not that optimized does not mean non-equal. it just means bether designed for a certain case.

As the matter of fact I do know programming and I am into optimizations and assembler so let me tell you something:

In many cases optimization is acceptable because it introduces only slight precision loss (one bit error) like when you replace a / const with a * (1 / const) because float division on P4 takes 23 cycles while multiplication takes only 6.

There is also algorhitmic optimization. Take for example Super PI program which many review sites still use for benchmarking CPUs. It is an old program written in 1995 so it is not very optimized. It uses Gauss-Legendre algorithm for PI calculation and it takes 47 seconds on my 2.8 GHz Prescott to calculate 1,000,000 decimal digits. BUT, there is another program PiFast43 using different algorhitm (Chudnovsky), and highly optimized FFT for big multiplications and it takes only 3.23 seconds on my 2.8 GHz Prescott for the same job. That is also acceptable optimization since there is no difference in output between two programs mentioned.

However, what ATI did is different from both examples:

1. They haven't used numerical optimizations that give ~3.8x speedup and introduce random one bit errors that could also be deemed acceptable.

2. They haven't invented new algorhitm for full trilinear filtering that works ~14.5x faster and gives the same result as in Chudnovsky + FFT .vs. Gauss-Legendre case I mentioned.

Instead they "invented" algorhitm which reduces precision of full trilinear filtering or even falls back to bilinear in some cases which is clearly NOT what I have requested through the game or control panel settings. And I don't like decisions removed from me and I guess neither do you? In case you really don't care then let me decide what you should think about it -- even if what they did was not bad, they did it in a bad way. 8)

 

[defiant_]

If anyone hasn't checked this out yet ( posted on the previous page ) please check out this link http://www.ixbt.com/video2/nv40-rx800-3.shtml it has some Quantitative measurements of the optimisation for both nvidia and ATI. Because it is russian you may wish to use a web transaltor like http://www.freetranslation.com/web.asp .

"the images received on chips ATI, will sometimes look more precise, and the images received on chips NVIDIA, - more dim, but at movement of the chamber of video, received on chips ATI, will be more inclined to occurrence of moire and "песочка" (dithering), arising because of an insufficient filtration (under-filtering)."

There goes the "IQ is the same" theory.

 

[jvd]

If anyone hasn't checked this out yet ( posted on the previous page ) please check out this link http://www.ixbt.com/video2/nv40-rx800-3.shtml it has some Quantitative measurements of the optimisation for both nvidia and ATI. Because it is russian you may wish to use a web transaltor like http://www.freetranslation.com/web.asp .

"the images received on chips ATI, will sometimes look more precise, and the images received on chips NVIDIA, - more dim, but at movement of the chamber of video, received on chips ATI, will be more inclined to occurrence of moire and "песочка" (dithering), arising because of an insufficient filtration (under-filtering)."

There goes the "IQ is the same" theory.

sigh. I don't see what they are talking about . Are they sure they are putting af on . There is no reason not to have 16 tap af on with those cards .

 

[Evildeus]

We are talking of trilinear and the optimisation. Then why not bilinear and 16 AF.

 

[Demirug]

sigh. I don't see what they are talking about . Are they sure they are putting af on . There is no reason not to have 16 tap af on with those cards .

they talk about that the optimisation did not end with the brilinear filter. They R420 have an other surprise inside. It is an AF-optimisation that even works if you only use the bilinnear filter. It use the same method to check if it can use it or not as the allready known optimisation.

And dont believe it is a hightech method. It is simple, stupid simple.

 

[jvd]

Well I can't read russian and I don't care to know about 3d cards not being able to take care of thier kids. But this is what i gather .

Use optimized трилинейной filtrations strongly enough lifts speed of performance of a filtration;
The prize at use optimized трилинейной filtrations is approximately identical at both manufacturers;
On screenshots the images received on chips ATI, will be the images received on chips NVIDIA, - more dim will sometimes look more precise, and. But at movement of the chamber of video, received on chips ATI, will be more inclined to occurrence of moire and "песочка" (dithering), arising because of an insufficient filtration (under-filtering);
NVIDIA on the current drivers for NV40 allows the user to include or disconnect optimization globally;
ATI does not allow the user of an opportunity of a choice, the driver itself decides to include whether or not optimization. For the certain classes of structures the current version of the driver does not use optimization never, for the others optimization is included, when the driver considers, that less detailed MIP levels are the reduced copy previous MIP levels.

So they say at times it looks more percise but at movement of chamber of video they wil lbe more inclined to occurrence of moire.

Basicly what I take from that is sometimes they are more percise but they are also inclined to show more moire.

Now inclined does not = that it shows moire.

As i have a card here in my house and it doesn't show any more moire than my 9700pro.

So what is this new af tweak ?

 

[Demirug]

Well I can't read russian and I don't care to know about 3d cards not being able to take care of thier kids. But this is what i gather .

So they say at times it looks more percise but at movement of chamber of video they wil lbe more inclined to occurrence of moire.

Basicly what I take from that is sometimes they are more percise but they are also inclined to show more moire.

Now inclined does not = that it shows moire.

As i have a card here in my house and it doesn't show any more moire than my 9700pro.

So what is this new af tweak ?

It is some kind of pixel flicker on objects that are far away from you.

R420 plays with the LOD-formula. You can see this in fullcolor with an AF-tester that works around ATIs detection method. Or you can use some special benchmark programms that show you what is going on.

Already say too much