How do you feel about ATI right now?

Discussion in 'AMD Graphics Cards' started by HardwareHeaven, May 21, 2004.

  1. HardwareHeaven

    HardwareHeaven Administrator Staff Member

    Joined:
    May 6, 2002
    Messages:
    32,274
    Likes Received:
    163
    Trophy Points:
    88
    With all the recent talk about optimisations and filtering techniques, what do all our ATI users feel about this? do you feel ATI have cheated the end user or do you feel what they are doing with filtering techniques is acceptable?

    Most of you have read ATIs public statement and read the chat log from the ATI CHAT a few nights ago...... but it seems like some sites such as techreport still feel the issue is not resolved and ati have been perhaps less than honest

    Whatever the merits of ATI's adaptive trilinear filtering algorithm, ATI appears to have intentionally deceived members of the press, and by extension, the public, by claiming to use "full" trilinear filtering "all of the time" and recommending the use of colored mip map tools in order to verify this claim. Encouraging reviewers to make comparisons to NVIDIA products with NVIDIA's similar trilinear optimizations turned off compounded the offense. Any points ATI has scored on NVIDIA over the past couple of years as NVIDIA has been caught in driver "optimizations" and the like are, in my book, wiped out.

    So whats your view guys?
     
  2. No_Style

    No_Style Styleless Wonder

    Joined:
    Jun 18, 2002
    Messages:
    6,027
    Likes Received:
    0
    Trophy Points:
    46
    They will fix it.

    I still have confidence in ATI. If they were really cheating, they would fix it in the next Catalyst release. (Similar to the action taken by ATI during the 3DMark optimizations fiasco) To say that this optimization is on the same level as NVidia is just wrong. To my understanding their method of trilinear filtering is "adaptive" in which it "detects" what kind of mip map level the filtering is applied on. In my opinion, that is a great idea to be able to flexibly apply the right filtering to the appropiate mip map level. The only problem with their method is that it is still flawed and that is what the people have caught ATI on. This wouldn't have surfaced if ATI's "adaptive trilinearing method" would have gone well.

    Take for example MP3 compression. If you convert a wav file to MP3 and you're losing significantly, people will notice. However, if you have perfected it, then no one will care.

    So concluding, I do understand TechReport's stance. When you look at all the slides and documentation about how ATI claims Trilinear filtering is on by default. You can't help but feel a little betrayed. I do, however, believe that ATI will rectify their mistake in the next release by either perfecting their method or forcing full blown trilinear until their method can be perfected.

    I hope that made sense.
     
    Last edited: May 21, 2004
  3. Huigie

    Huigie New Member

    Joined:
    Oct 27, 2002
    Messages:
    44
    Likes Received:
    0
    Trophy Points:
    0
    I couldn't care less. ATI it is, ATI it will be.
     
  4. ReDGuiTaR

    ReDGuiTaR New Member

    Joined:
    May 5, 2004
    Messages:
    746
    Likes Received:
    0
    Trophy Points:
    0
    Still ... they got the best IQ around .... how can we complain about it ???
     
  5. HardwareHeaven

    HardwareHeaven Administrator Staff Member

    Joined:
    May 6, 2002
    Messages:
    32,274
    Likes Received:
    163
    Trophy Points:
    88
    We have a very strong ATI userbase here, personally im curious to hear what you all think about everything thats happened lately.
     
  6. Chaos

    Chaos Number Nine

    Joined:
    May 9, 2002
    Messages:
    5,260
    Likes Received:
    95
    Trophy Points:
    0
    After reading their official response and reading the chat logs, I think they have come pretty much clean. I give them merits for admiting they changed the way the data is interperated and if there is more to the story; they seem like a company that will either remove it or fix it
     
    Last edited: May 21, 2004
  7. No_Style

    No_Style Styleless Wonder

    Joined:
    Jun 18, 2002
    Messages:
    6,027
    Likes Received:
    0
    Trophy Points:
    46
    Re-read my post

    :D I did some editing. Am I getting this right?
     
  8. Jalle64

    Jalle64 New Member

    Joined:
    May 11, 2003
    Messages:
    231
    Likes Received:
    0
    Trophy Points:
    0
    OT: Well first I want too say thank you! and a GREAT JOB with the site!:D
    I can´t say I´m surprised, but the way ATi: s been acting about it, so righteous an all, they look a little shadier too me! sorry to say.:( They have also to "wash their messy laundry" before they can come clean I think! And thats what we are here for!;)
     
  9. mike2h

    mike2h New Member

    Joined:
    Nov 11, 2002
    Messages:
    6,359
    Likes Received:
    69
    Trophy Points:
    0
    i beleive it is being blown out of proportarion, but i also beleive ati got caught with their pants down. i also beleive they will 'fix' the problem with little or no loss to performance/iq due to their excellent team. i also beleive neither team has a right to cast stones at the other.
     
  10. HawK

    HawK Banned

    Joined:
    May 13, 2002
    Messages:
    2,092
    Likes Received:
    0
    Trophy Points:
    0
    Well i'm not as much in 3D techniques as i used to, but ;
    If Ati can do optimisations without the loss of IQ, I see no problem & has been done on 3D cards for years
    if NVidia cannot, that is their problem..
    Digging/scraping out a thing that could have been discovered over a year ago, is "techreports" problem :bleh:
     
  11. ReDGuiTaR

    ReDGuiTaR New Member

    Joined:
    May 5, 2004
    Messages:
    746
    Likes Received:
    0
    Trophy Points:
    0
    are you in my head ?! :eek:
     
  12. zerodamage

    zerodamage New Member

    Joined:
    May 16, 2003
    Messages:
    3,478
    Likes Received:
    0
    Trophy Points:
    0
    ATI screwed up on this one. The TechReport article is pretty much dead on. This type of optimization is fine. The problem is claiming to use full trilinear filtering all the time when the X800 series is not. Now if ATI admitted they were using this new Algorith at the time of the reviews, then there wouldn't be this big deal. Also, According to that chat, this is being done through the drivers and it is not hard coded into the x800 hardware. ATI can fix this by allowing an option in the drivers to disable this algorithm and use full trilinear all the time.

    Edit: To add to this. With an option to turn off the optimizations, true comparisons could be done with it off and on to see if there really is a difference IQ wise.
     
  13. davidnsibs

    davidnsibs 9500pro Forever

    Joined:
    Oct 7, 2003
    Messages:
    273
    Likes Received:
    0
    Trophy Points:
    0
    i agree with zerodamage completely. basically the image quality isnt noticably imparied, and we get better performance (well 9600 and x800 users do anyway). so this is great, well done cat makers for coming up with this clever way of improving the way that our games are rendered. however, its a shame that we had to find out about this clever knew algorithm thing through articles claiming that ati are cheating rather than in a readme of a new driver set or something. my next card will be an ati irrespective of this or any other things as ive been so pleased with my card atm and the support from this site is brilliant, but any press releases from ati about image quality i guess ill take with a pinch of salt from now on as they may not be revealing the full picture

    EDIT: hmm, sorry sounded a bit harsh, i still really like ati and everything that its doing. shame that some people and sites were waiting for anything that they could try to get at ati with
     
    Last edited: May 21, 2004
  14. snake

    snake Active Member

    Joined:
    Oct 15, 2002
    Messages:
    423
    Likes Received:
    6
    Trophy Points:
    43
    What can i add ? ATI RULES :evil:
     
  15. germanjulian

    germanjulian Back in London

    Joined:
    Jul 15, 2003
    Messages:
    1,797
    Likes Received:
    2
    Trophy Points:
    0
    they said one thing and did another. Sure their IQ is still perfect but they never mentioned this filtering! they lied a bit, bit dissapointed but at least they are honest now and try to rectify things. good work ATI little dissapointed
     
  16. WaltC

    WaltC New Member

    Joined:
    May 9, 2003
    Messages:
    245
    Likes Received:
    4
    Trophy Points:
    0
    Here is what I posted at TR:

    Encouraging reviewers to make comparisons to NVIDIA products with NVIDIA's similar trilinear optimizations turned off compounded the offense. Any points ATI has scored on NVIDIA over the past couple of years as NVIDIA has been caught in driver "optimizations" and the like are, in my book, wiped out.

    This is a pretty sad conclusion to have reached, considering how much in error it is...;)

    First, if you know anything about the ATi algorithm then you know that under conditions such as forcing colored mip levels on in games, such as the UT2K3/4 colored mipmap tests, or with tools like ATi recommends, such as 3D Analyzer, the drivers run such code in full trilinear. What ATi is telling people is that when you run colored mipmap software, and you test it against nVidia's products running the same software, you have to turn off nVidia's optimizations manually because ATi's turn off automatically in the presence of such software, so that you can get an apples-to-apples comparison running such software.

    From the ATi chat:

    ...Our target is also to avoid any need to detect applications, and as such we have to try to be sure that our image quality remains high in all cases. To achieve this we spent a lot of effort developing algorithms to make the best use of our quality tuning options. This isn’t a performance enhancement applied to popular gaming benchmarks; it’s a performance boost – without image degradation – applied generically to every game that uses box-filtered mipmaps, which is most of them.
    ....

    If it were the case that we were only performing full trilinear with coloured mipmaps then you might have a point, but this is emphatically not what we do. We take advantage of properties of texture maps for performance and IQ gains. In cases where we are not able to determine that the texture content is appropriate for these techniques we use legacy trilinear filtering. This includes cases such as dynamically uploaded texture maps where we avoid performing analysis so as not to cause any possible gameplay hitches.


    So, it's very difficult to see why TR finds ATi's advice unacceptable, as when running colored mipmap software in the way that ATi recommends for reviewers there is no optimization taking place in either the ATi drivers or the nVidia drivers, and as such it's full trilinear versus full trilinear, which is what you want, right?

    Second problem here is that you've assumed nVidia's trilinear optimizations and ATi's trilinear optimizations are *the same thing* when clearly that's not the case at all. After labeling both as "trilinear optimizations" the similarity between the techniques ends.

    Third, what on earth does "the last two years" have to do with something specific to x800? Beats me...;)

    If I had to guess, I'd say TR has gotten a bit confused on the issue and is thinking that ATi's optimizations *turn on* when running colored mipmap software--when in fact they *turn off* automatically, and therefore any comparative instruction ATi provides towards manually turning nVidia's off to compare products running colored mipmap software is entirely correct. If you ran colored mipmap testing software and left nVidia's optimizations on you'd be comparing them to ATi's drivers running the same software with optimizations turned off. That would be dishonest, wouldn't it, as well as an inaccurate, unfair comparison, wouldn't it? I don't think TR would be in favor of that, and so it seems you've jumped the gun in a knee-jerk fashion and need to consider the issues a bit more.
     
  17. No_Style

    No_Style Styleless Wonder

    Joined:
    Jun 18, 2002
    Messages:
    6,027
    Likes Received:
    0
    Trophy Points:
    46
    But like I said. If their method was near flawless, then none of this would have came out. Nobody would have noticed that they were using a different method of trilinear filtering. The safe route would have been to do what you said and mention the new algorithm. Like they say "Better safe, than sorry."
     
  18. free_refil

    free_refil New Member

    Joined:
    Mar 15, 2004
    Messages:
    100
    Likes Received:
    0
    Trophy Points:
    0
    I also agree with zerodamage. Had ATI toted this new algorithm around at the release of the X800, stating it improve performance without sacrifice of image quality, and also provided a way of disabling it in the ATI control panel, things wouldn't have come about so harshly. I personally feel that they shouldn't sacrifice image quality for performance. ATI was known for rendering full trilinear filtering with impecable image quality faster than NVIDIA and their algorithm. Now that they have done this, they do appear to be a bit on the shady side. But, I will sit back and see how things go.
     
  19. dpagan

    dpagan New Member

    Joined:
    Apr 5, 2004
    Messages:
    249
    Likes Received:
    0
    Trophy Points:
    0
    i have a 9800 pro and as long as i see what i am supposed too with as much eye candy as possible i don't care about any method used to achieve it, ati is doing a great job but they should have stated the method used in the 1st place but other than that i don't care, too many users out there drown in a cup of water, you give them poor performance drivers but with great IQ they'll still complain about the performance, if you give them drivers with good performance and good IQ they'll complain about the method used to get the performance and IQ, as a gamer i have come to accept that for the most demanding games out there, there has to be a trade off somewhere cause if your system isn't up to it and you want it all ON your game will play like crap, i say 'ati & nvidia go on and keep exploring new methods for AA & AF so we the gamer can get good IQ and FPS so we can play our games better and extend the life of our graphic cards" any body making a screen shot of any game magnified 300% just to look for faults and complain when all there doing is to give you more bang for you buck with just a driver upgrade needs to GET A LIFE.
     
  20. OpusP.Gladiator

    OpusP.Gladiator '42'

    Joined:
    Oct 16, 2003
    Messages:
    300
    Likes Received:
    0
    Trophy Points:
    0
    Why Lie?

    ATI Owned up to it's ways as it usually does... Still waiting for the next evolution in systems tho i.e. pci-express on the btx mobo stndrd w/ ddr2 and high def onboard audio on a philips 3d display and Blu-Ray! hahaha Dreaming of a time when PR actually puts the appropriate spin on things... End the Hype of the future (it's here already) Show me all the wonderful things Ruby has in store for us.....
    --oops, sorry just finished dl the latest directx9c zip but don't think it's gonna help since "everyone" else is so far behind with apps/tweaks...Is Nvidia the only co. to make use of the new stuff in 9.0c???? WHY???? Anyways, sorry 4 the rant...it's early still...hehehe

    --btw pleasantly happy with my ati 9800 pro (even tho it's an early model...)

    --as ever
     

Share This Page

visited