Hanners said:
No, the mip map transitions are not obvious, what is apparent is that I can tell there are different mip map levels in use. Its a subtle difference, but one that Trilinear prevents.
OMG HARDOCP REVIEW OF UT2003 AND FILTERING
- Thread starter bloodbob
- Start date
- Status
No, the mip map transitions are not obvious, what is apparent is that I can tell there are different mip map levels in use. Its a subtle difference, but one that Trilinear prevents.
I'm sorry but Kyle and Brent are purposely trying to help Nvidia perpetuate consumer fraud. It's that simple Nvidia for years now have had a plan on how to maintain there performance lead over every card maker out there, and this plan includes driver cheats and grass roots campaigning by influencing and bribing hardware review sites to do the marketing PR for them.
They are part of this and they need to be stopped plain and simple. Nvidia has single handily ripped off consumers for millions of dollars worth of video cards and ripped off oem's for millions of dollars by selling video cards that simply are not nearly as good in reality as they are on benchmarks which consumers and oem's base there buying decisions on.
That in a nutshell is consumer fraud and any party that contributes to this fraud needs to be taken down.
It may not be politically correct to have these personal attacks but Kyle's entire business model relies on the fact that he have an image of unbiased hardware reviewer which is obviously not the case. This image of his needs to be tarnished, people need to be taken off there high horse and brought down to reality. For this industry to move forward these events of the past few months need to have long lasting consequences so that they are not repeated again.
By the way things are going it seems Nvidia and [T]ardocp are going to get away with this by skating the issues putting on there pr face and trying to bs the unsuspecting public away from the facts. They will rely on the fact 90% of consumers have no clue about these type of forums or the extent of what's going on. And it seems they will get away with it scott free make a couple of pr statements and just dig deeper and find new ways to rip off the consumer.
This needs to be stopped by any means.
They are part of this and they need to be stopped plain and simple. Nvidia has single handily ripped off consumers for millions of dollars worth of video cards and ripped off oem's for millions of dollars by selling video cards that simply are not nearly as good in reality as they are on benchmarks which consumers and oem's base there buying decisions on.
That in a nutshell is consumer fraud and any party that contributes to this fraud needs to be taken down.
It may not be politically correct to have these personal attacks but Kyle's entire business model relies on the fact that he have an image of unbiased hardware reviewer which is obviously not the case. This image of his needs to be tarnished, people need to be taken off there high horse and brought down to reality. For this industry to move forward these events of the past few months need to have long lasting consequences so that they are not repeated again.
By the way things are going it seems Nvidia and [T]ardocp are going to get away with this by skating the issues putting on there pr face and trying to bs the unsuspecting public away from the facts. They will rely on the fact 90% of consumers have no clue about these type of forums or the extent of what's going on. And it seems they will get away with it scott free make a couple of pr statements and just dig deeper and find new ways to rip off the consumer.
This needs to be stopped by any means.
Doomtrooper
Veteran
I tested a 5600 U with the 44.03 drivers and played some custom CTF maps and could see the mip map boundries quite often..especially in maps with alot of grass/stairs like CTF MUSE.
Reminded me alot of the old 8500 Days and Serious Sam 2...for some it may not be noticeable, but when running the anti-detector script and enabling trilinear you could see it night and day..and the performance hit too.
CTF MUSE
Reminded me alot of the old 8500 Days and Serious Sam 2...for some it may not be noticeable, but when running the anti-detector script and enabling trilinear you could see it night and day..and the performance hit too.
CTF MUSE
Fred da Roza said:
He could have. He chose not to. Does that mean there is a shooter on the grassy knoll? [H] has apparently been in a weeks worth of heart to heart discussion with Nvidia talking about many of the very "issues" talked about on this board and many others. From all indications Nvidia is paying attention. Again, regardless of the path, if the final destination improves our experiance as users, more power to em.
digitalwanderer said:What I'm wondering about now is if he was in secret discussions with nVidia fixing the "problem" than why did he and Bent write up that justification for the "almost-trilinear" filtering?!?
I hate it when he contradicts himself and sends mixed messages, I get so confused!
The write up was not a justification for "almost trilinear." It was a review of IQ between cards if anything.
Blackwind said:
Agreed, if it benefits us as users in the end, then it's not all in vain. The problem is that if [H] knew about issues with UT2003, why did they insist on still using it for benchmarking in all their shootouts? In the short-term, that's damaged both [H]'s reputation, and resulted in what could be perceived as inaccurate benchmarks being used by them and a whole host of other sites.
I don't see why it has to be a one or the other approach between talking to nVidia about an issue, and informing the public of an issue. If they do both, then everyone that [H] should care about wins.
Blackwind said:
Yes, but knowing full well that your comparative benchmarks are not being conducted under apples to apples settings and justifying it through subjective visual checks does not benefit the industry. If one of the major hardware sites will not only condone but, for all intents and purposes, knowingly enable such practices for one IHV, the others will certainly feel pressured to engage in such standards themselves, and in the end the consumer will only see a continuation of popular games used as benchmarks displayed with lowered image quality. Less power to 'em, IMO.
And on a site note, seeing Kyle not only refuse to answer Rev's politely worded questions on his forums and instead launch into a personal attack accusing Anthony of failing to have a positive influence on the 3D graphics market (describing him as a "wasted talent") is my new personal definition for the word ironic.
Hanners said:
If perception was what we all had to be measured by, we'd all be screwed.
I don't believe there was any insistance to use UT2003 or any other tool [H] presently uses for reviews. You work with what you have to use. In this case in question it was specifically used due to ATI's accusation/concern. Would kind of make it really pointless of them to address it using Sponge Bob Square pants to prove otherwise. [H] does apparently still use it in their video card reviews but again, what else can they use? It's not as though there are tons of tools out there. That and apparently it's still pretty popular to play.Kyle and Brent have expressed time and time again their frustration as a review site and just how much more difficult the job is to do that, review. It's not as cut and dry as it once was when all you had to do was look at a FPS and you knew. (misses 3dfx) I find current standings much more exciting simply thinking of the potential. ATI and Nvidias hardware are apparently very differant. They also choose to render differant. In the end, I'm willing to bet, the consumer will win. Very rarely in the history of business has that failed to happen.
Blackwind said:
Therein lies my point. In the situation it appears [H] were in regarding what they knew, the only two choices they had were to:
a. Stop using UT2003 as a benchmark, or
b. Inform the public that their UT2003 benchmarks were not a fair comparison, and explain the reasons why this wasn't the case.
Hanners said:
Unfortuantly I find that to simply be a matter of opinion in regard to appearance. Not one I can say I share. I think they did a good job of stating that already. They stated, its not an apples to apples comparison. How many more times must they say it?
John Reynolds said:
Never had a shitty day?
I know I have. I can't say I'd be too receptive either sitting in his shoes especially as of late. I'm not excusing him but I think in review what you are asking is for him to have is the kind patience of Buddha. I think he’s done pretty well. I’d have hunted em all down and proceeded with Chinese Water Torture. They did state it wasnt an apples to apples comparison and why.Blackwind said:
To be honest I don't remember reading that in any of their reviews of the past couple of months when it came to the specific case of UT2003 (this last article which dealt directly with the issue notwithstanding). And if they did say 'we can't do an apples to apples comparison in UT2003', I don't recall them explaining the reasons why.
As for Kyle's behaviour, I think that he needs to remember that he presides over one of the most popular hardware sites on the Internet, and when you are in that enviable position you are going to be subjected to a great deal of scrutiny. Unfortunately, I get the impression that Kyle wasn't, and never has been, prepared for it, hence his reactions whenever the going gets tough. It's a fact of life that when you're at the top people will want to knock you off your pedestal and pick up on every error, and you need to react to that in the right way. In my opinion, Kyle hasn't.
Fred da Roza
Newcomer
Blackwind said:
After 6 months of denial. In other words he practiced a double standard, which is exactly what I originally said.
You don't find the statement that he knew about this "two months ago" a little spurious? Doesn't that mean he knew about this the moment NV35 was available to reviews (or even before)? If so, how did he know? Evidently [H] can't tell the difference between NVIDIA's hacks and real Trilinear and nobody knew how to enable the mipmap colors in ut2003 until this site pointed it out. So, how on earth could he know this "two months ago"???
Hanners said:
I 100% agree. I do not think he was prepared for just how well and noted a site [H]ardOCP and [H]ardforum have become. Being human as well, we can only recommend some area's of improvement and hope for the best.
Fred da Roza said:
I'm not following your double standard comment. How is that a double standard?
whql said:
I assume this was directed at me so I'm responding. I don't find anything suspicious, fishy, cagey, green-eyed or suspect about it. For as long as I have read [H]ardOCP they have been consistent in doing things behind the scenes to try and help us all. At an appropriate time, telling us all about it. It’s only been recently (last year or so) where I’ve noticed everything they do being put under the microscope.
Hanners said:
When their 5900 review came out and Brent got raked over the coals about the whole UT2K3 thing, his reactions to it definitely indicated to me he was unaware of the issue, or at least unaware of the scope. He certainly made no mention of [H] already discussing the issue with Nvidia.
Dave H said:
On the contrary--what the game asks for is paramount. I think this is where you've gotten a bit off track here.
The ideal situation is that all 3D games allow for the user to set up his preferences with respect to FSAA, AF, filtering, resolution and color depth from within the application. That is the goal and has been for a long time.
Use of control panel settings to *force* settings changes for these parameters has been implemented by IHVs because of the lack of such configurability in the great majority of 3D software which has shipped to date. Forcing doesn't always work, though, because of the application and not specifically because of problems in the drivers--witness the inability to force FSAA in Splinter Cell, for instance. Sometimes the issues have to do with the API and get more complex, but that's really not the point here. The point here is that forcing these settings through driver control panels is not the best approach and never has been--setting games internally to instruct the 3D card drivers to initiate various display options has always been the "ideal." The IHV's have simply taken up the slack left by developers unwilling or unable to implement simple application controls to instruct the driver how to render, and in cases of older 3D software which predates the implementation of AF/FSAA, etc.
Bottom line is that the application should be the ultimate authority. Hence, even if full trilinear filtering does not occur when the option is selected in the control panel, it should occur when the game itself instructs the drivers to provide it. The fact that it does not do so with the Detonators is proof that nVidia has simply hard-coded its drivers not to provide full trilinear filtering when the UT2K3 game is running and is recognized by the Dets. And that's the problem.
Brent said:
Sorry if this sounds like the Spanish Inquisition, but does that imply that you were unaware of Kyle's discussions about this matter with NVIDIA, which presumably occurred before Dave's revelations here on B3D? If so, why would Kyle not inform his reviewers about an ongoing issue of such relevance, especially with your NV35 review about to be posted?? :?
This almost deserves its own thread, but I thought it was interesting and quite on the topic here:
http://www.halflife2.net/forums/sho...f74af40&threadid=3071&perpage=15&pagenumber=1
Specifically, this one quote I found demonstrative of the topic in this thread:
http://www.halflife2.net/forums/sho...f74af40&threadid=3071&perpage=15&pagenumber=1
Specifically, this one quote I found demonstrative of the topic in this thread:
(emphasis mine)Gabe Newell said:
- Status