TechReport chimes in on the HL2 benchmark

[Dave Baumann]

http://www.beyond3d.com/forum/viewtopic.php?t=7867

 

[K.I.L.E.R]

Thanks for the link Sir.
Everyone seems to be backing Ati now just like a few years ago when everyone started backing nVIDIA.

 

[Chris123234]

He also mentioned that he's seen drivers detect screen capture attempts and output higher quality data than what's actually shown in-game.

This has to be the bombshell of the day. Apparently in addition to recording their own secret demos, 3d card reviewers are now going to have to carry around digital cameras with them???

I'm not that surpriced, though I didn't expect it to go quite that far. I know I mentioned the possibility to do just that some time ago on this forum, though I didn't expect it to actually happend since I guessed it would not be that trivial to implement in drivers.

Actually unless you made another thread or something I'm the one who asked if thats what nvidia could be doing in benchmarks In fact, I think you even posted in it

Edit: you even said it would be hard to implement lol. http://www.beyond3d.com/forum/viewtopic.php?t=6059&start=0&postdays=0&postorder=asc&highlight=. Just messin with ya.

 

[cthellis42]

<laughs> Thanks, Dave. Serves me right for not checking the News forum first.

Must say I don't detect "annoyance" from MS's release, though. Just ringing endorsemen for HL2, which speaks a lot in and of itself, certainly. Do you know what the timing on Gabe's comments vs. Microsoft's announcement were? They both appear to be out the same day, so I can't tell offhand if MS was planning on saying that in general after the event, or decided TO say that simply to offer Valve support for their comments?

Hard to tell exactly, but it certainly says a lot regardless.

 

[Evildeus]

So, if a man was convicted of previous felonies but then *appeared* and *commented* on his intent to change - any charge made against him - even those with no empirical evidence that proves guilt - after that point should be considered true?

Logical contradiction. If there is no empirical evidence of guilt, why would the man "appear" and "comment on an intent to change"? In order to change, one must first acknowledge one's present course is in error. Surely you are not characterizing nVidia here, as I've seen plenty of empirical evidence, but no "comment" on the part of nVidia which specifies an "intent to change." You therefore must mean some other entity.

Well, i think you forgot the time constraint making your reponse non sensical.
1- Nv is found gulty of something (as Vince point out)
2- Valve says they are still cheating
3- Nv answers that's they should use det5* and that there's a bug on fog.
4- Therefore Vince can say that even if Nv have been guilt in the past there's no proof that's the fog bug is a cheat and not a bug as Nv stated they wanted to change (after the premilinary cheatings).

I don't care. Show me proof and then I'll join you. Without unbiased, reproducable, and empirical proof - you've comments have NO validity and can be made by an invidual against any Company, Institution, and Product. How hard is this to comprehend?

Illogical contradiction. Proof has abounded for months from a multiplicity of sources. You have not acknowledged that proof--therefore, whatever proof he could show you would also be ignored. You ask him to show you proof, and then you declare that he cannot do so. You have decided--before seeing it--that such proof does not exist.

It's not because all the ducks that i saw were red, that all the ducks are red. Inductive methodology is not future proof.

 

[bdmosky]

Sorry, but your problem is that you don't know what constitutes proof and what doesn't. Valve is putting its credibility on the line and doing it *before* their game goes on sale--that's enough proof for me. I think the proof is grossly abundant over the past 10 months--but not of course if you don't wish to see it. I'd say Valve has been extremely systematic in its approach to nV3x hardware over the past several months.

Haha, you're right. I should believe the white male plantation owner because he's a respected member of the community. He's telling the truth because slaves are too stupid and they can't help but lie all the time. Yes I'm being ridiculous here, but I hope that at the very least, you see that you're being ridiculous here too.
My point... respected members of the community can lie too, and no my friend, that isn't enough PROOF for me.

Face it: you are not in a postion to prove or disprove anything yourself

Exactly! But neither are you.

All you can do is rely on the word of people who are in such a position, like Valve.

Why do I have to rely on them again? Oh right, because it's important that I immediately bash Nvidia on every free whim. It's obvious I can't hold my tongue and wait for evidence to be presented.

Like I said, Bdmosky, the proof is abundant everywhere--you might say it's ubiquitous...it has been for months...

I agree, according to those screen shots, the 51.75 drivers don't appear to have the best image quality with the AquaMark 3 benchmark, but then again, the assumption is still being made that it is because of color precision issues without offering much in the way of support. Additionally, I've yet to see how the 51.75 series drivers affects IQ in HalfLife 2, if at all. My question is, would a loss in precision in any way affect the variation of colors displayed?

 

[paju]

Seems odd to me that some people insists that nVidia is not cheating but doing everything right. We have "evidence" from Valve and ATi (not official). Naturally ATi cannot make public accusations as it would tarnish their reputation. Valve made it but did not provide sufficient information about the driver version etc. However, why the heck Valve would say that nVidia is cheating when we all can remember FM case from not too long time ago?

IMO, Valve did put a lot of effort to get NV3x line cards run DX9 path fast but ended up with mixed mode. What more can you ask from the company for one IHV? Did they optimise for ATi - not according to Gabe's presentation. And when nVidia is pulling cheats (according to Gabe) such as screenshot detection, missing fog and who knows what then most certainly developer will be pissed. Maybe Valve forces all nVidia products to use DX8 paths with upcoming patches/updates if nVidia keeps on cheating.

How about det5's upcoming shader "compiler"? From what I understood it sounds like that it reverse-engineers the shader code and creates something that is close to original and runs much faster on NV3x than the original. People most likely will remember the guide lines that nVidia presented about driver optimisations. One of them was that the optimisation must not affect the IQ. OK, maybe it does not affect if you're nearly blind (yes, this was sarcastic statement).

Instead of us proofing nVidia is cheating it should be up to nVidia to proof they are not. It's shame that they HAVE to do it because the NV3x GPU is not fast. They cannot admit to the public that NV3x was a failure when it comes to DX9 path. The best way would be like someone suggested that no reviewer (well, there are those who braise nVidia despite what) should use nVidia's cards for anything until drivers' contain no more cheats. Encrypting drivers does not help the cause.

 

[stevem]

While we've known for some time about Nvidia's quirks with WHQL certification, how many would be surprised by "MS having to upgrade the WHQL test to find spec violations." I'd imagine MS not impressed by IHVs cheating their WHQL certification tests...

In case people missed Dean Calver's post yesterday, here's the pertinent link:

http://www.beyond3d.com/forum/viewtopic.php?t=7873

You'll find a large number of devs, ISVs & indeed OEMs feeling the same way...

 

[demalion]

Charge more to nVidia card owners specifically? First: are you serious about that being feasible, or are you just avoiding recognizing a problem with your analogy?

For myself, I still don't think Valve has the option of charging more to nVidia customers specifically, and your proposition otherwise continues to seem nonsensical to me. This still leaves your analogy "all kinds of broken" AFAICS.

Valve can do what they want via Steam, which was the point of the post.

Yes, and I can go jump off a cliff. You ignoring that that this solution might possibly not be a reasonable one (from the viewpoint of the one deciding not to jump off the cliff), and proposing an argument and analogy based on the premise of this occurring, isn't much of a point...besides one of simply speaking nonsensically as an alternative to budging an inch from your position.

If they wish to target N consumers that have purchased i hardware - then they will have to put in the manpower/monetary/temporal resources that will allow it run acceptable.

And they did. But your solution went further to support your mechanic analogy and required that they had the option of charging the nVidia customers more for the extra work. Simply trying to bury the flaw in your argument by talking loads of nonsense around it doesn't work very well, and ends up with you saying "they should have done this" when describing something they did do, using your nonsensical comments to avoid recognizing that contradiction at all.

This is an obvious prerequisite to make a given sale in any marketplace. Thus, as per the origional post - Gabe knew the repercussions of supporting nVidia's hardware and still went down that path.

This reply to my comment says and does this: In trying to support nV, the only possible outcome is success equivalent to any other brand, and any discussion related to hardware capabilities are just made to disappear by you as far as your replies. But hardware capabilities are what this is about...why do you keep making it disappear and think your discussion is valid at all? What is this infatuation with "brand" being all that matters?

His comments as bdmosky stated, only created a more polar enviroment.

So: Valve shouldn't ignore nVidia card users, "thus" if they try to support nVidia card users, they can't speak about the issues they encounteered or of consumers being mislead by whatever the IHV can do?

First: Where is your logical support for this?
Second: Where is your set of limitations on what an IHV can do? You know, like not cheat in benchmarks and place a software developer in this position, and silly things like that?
Or is it that your stipulation is valid because it is yours, and any other stipulations are irrelevant because they're not? Just trying to make some sense of your statements here and commentary...

They did. It just isn't compatible with full feature exposure, because of the hardware's capabilities. This point is still lost on you, I see?

I already stated:
...

Yeah, and I replied to this statement afterward, because that is where it occurred in your post. Moving your statement up here as if I didn't already point out problems with it, and ignoring the problems pointed out, is rather pointless, yes?

Let's see what results from your mechanism of proposing your statements as a proof while ignoring any problems pointed out with them:

So, they didn't make it playable within the limits of the hardware? Or does the hardware not have any limits?

You're missing what I'm saying. Of course there are upper bounds on absolute preformance - yet, you can offset this to a large extent by investing more time to run a given application more effeciently. Which is something Gabe stated they did, but used it in a negative context.

Err...because 1) the hardware still failed to perform after they did it, 2) nVidia has and continues to extend effort to lie to consumers about the hardware's performance, and to instead specifically blame the software developers for poor performance.

Vince...that is a "negative context"! The truth can be negative, even for nVidia ...why do you go on about "they shouldn't say something negative", and ignore the issue of something being true completely? Consider perhaps that this is related to"negative" responses to your own words, if that "lesson learning" thing is still possible.

Which comes back to my main point, which Joe agreed to, that the consumers bought that partticular hardware. Thus, if you're targetting and chasing a given group of consumers with i hardware, then be prepared to put front-end work and money into creating an application that runs acceptably.

And they did. nVidia's lie is about saying it can run as acceptably (well, they try to propose better, I think) as other products when enabling full feature support (DX 9), which it cannot actually do. Why isn't anyone allowed to react to the issues resulting from their acting this way?

BTW, I covered this already in the large section of text you ignored to quote yourself...all that "negativity" that you seem to prefer not be directed against things you favor (AFAICS, this includes yourself and nVidia ) regardless of evaluating the truth of statements.

Don't go out of your way to critisize a company for making you do your job as a developer.

OK, by what principle does this follow, and what "job as a developer" is it you are proposing Valve did not do?

What about the "job" of an IHV to put out hardware capable of what they say?

....

Missed the point, already covered.

You know, maybe I should have just quoted your post and replied like this, and skipped showing the "why". This seems the type of conversation you prefer. However, I doubt you'd respond to it like you expect other people to respond to it when you do so.

Heaven forbid consumers buy hardware because it is actually better. No, they shouldn't conform to performance capability recognitions, simply their brand loyalty and ignorance of such information. Shame on a developer for challenging that, and seeking to inform contrary to consumer ignorance.
In short: hardware can't actually be better, people should be able to pick the brand name they want?

This is absurd. As a supplier, your goal is to sell a given product which has a set-piece value to a given group.

I addressed your original mechanic example. Joe addressed this new one. Your accusations of absurdity neglect responding to these...in fact, it seems to depend on ignoring them. Another item I think you could consider for that "lesson learning" thing.

To stick with the earlier automobile example, if I was producing a given brakepad designed with a Dodge Viper in mind: Should I tell the other 99.99% of the marketplace who own POS cars: "Heaven forbid consumers buy hardware because it is actually better"?

Well, my brief statement on the problem here is that the "Viper" is the same price as the "POS cars" (well...actually less). And, in reality, mechanic recommendations do influence the purchasing decisions of consumers (if they look at things other than brand, that is). Does "all kinds of broken" sink in yet, or are you proposing that all consumers should just be buyng by brand, and that this makes sense for consumers to buy a "POS car" for the same price or more than a "Viper"?

No, a mechanic who could only charge the same, and would be blamed if they failed to get good performance out of the car, would not recommend people get Vipers (or whatever would let them tune up with less effort, and give the consumers higher performance) instead of POS cars, even with the former at the same or lower price. This makes perfect sense, you're saying?

Common' here. The seller conforms to the prevailing consumer and marketplace - the marketplace doesn't adapt to what I see.

Yes, I understood the "hardware doesn't matter, vendor does" reasoning you propose, which is why I discussed it. It is just that your dismissal of that discussion doesn't hold together.

The rest of your post about informing the consumer is irrelevent to a discussion on producing a product that has a wide market appeal from a preexisting userbase.

Yeah, like this. Your automechanic examples are "irrelevant", and I've actually shown why.

 

[Humus]

Actually unless you made another thread or something I'm the one who asked if thats what nvidia could be doing in benchmarks In fact, I think you even posted in it

Edit: you even said it would be hard to implement lol. http://www.beyond3d.com/forum/viewt...p;postdays=0&postorder=asc&highlight=. Just messin with ya.

I recall a much longer thread in which I made a comment like that, though I'm not sure either. Not sure if that thread would be newer or older than that one. Not that it matters anyway.

 

[WaltC]

Well, i think you forgot the time constraint making your reponse non sensical.
1- Nv is found gulty of something (as Vince point out)
2- Valve says they are still cheating
3- Nv answers that's they should use det5* and that there's a bug on fog.
4- Therefore Vince can say that even if Nv have been guilt in the past there's no proof that's the fog bug is a cheat and not a bug as Nv stated they wanted to change (after the premilinary cheatings).

You know, you should check the context of Vince's remark. It was in response to OGl guy's talking about the fact that nVidia had been caught "red-handed" cheating on several occasions. Valve wasn't mentioned in the exchange.

But...in response to your remarks:

1. What was nVidia "found guilty" of that nVidia admitted to and pledged to correct?
2. Valve said that nVidia--and other IHVs--have cheated in the past and that the current 50'xx Det *betas* aren't recommended, because they are beta and contain things Valve didn't like
3. nVidia recommended Det 5 *beta's* and the fog problem was discovered before nVidia commented on it. In all other instances that I am aware of nVidia constantly tells end users NOT to use beta drivers--but to use officially released drivers, instead. What's the difference? Is it Valve's job to pimp nVidia's products? I think not.
4.Vince said that nVidia had commited to a "change"--that has not happened--if it has--please provide me a link to the statement because I've never seen it.

I don't care. Show me proof and then I'll join you. Without unbiased, reproducable, and empirical proof - you've comments have NO validity and can be made by an invidual against any Company, Institution, and Product. How hard is this to comprehend?

Illogical contradiction. Proof has abounded for months from a multiplicity of sources. You have not acknowledged that proof--therefore, whatever proof he could show you would also be ignored. You ask him to show you proof, and then you declare that he cannot do so. You have decided--before seeing it--that such proof does not exist.

It's not because all the ducks that i saw were red, that all the ducks are red. Inductive methodology is not future proof.[/quote]

Huh?... He invites Ogl guy to "show proof" and then himself declares Ogl guy's comments as "invalid" without ever attempting to PROVE them invalid himself. Circular argument.

 

[WaltC]

Haha, you're right. I should believe the white male plantation owner because he's a respected member of the community. He's telling the truth because slaves are too stupid and they can't help but lie all the time. Yes I'm being ridiculous here, but I hope that at the very least, you see that you're being ridiculous here too.
My point... respected members of the community can lie too, and no my friend, that isn't enough PROOF for me.

Why would you pollute the discussion with racism? As well, last time I checked, Valve does not own a plantation, and holds no slaves....*chuckle* Here's a clue: it is now 2003--it hasn't been 1860 for an awful long time, and a lot has happened in the meantime of which you seem blissfully unaware. Do yourself a favor and try and catch up.

Trust is an essential part of any civilization. When trust breaks down, so does civilization. The issues here are precisely those of trust--as in, who do yout trust and who don't you trust? Valve is in a position to know certain things, and they make their money selling their software to people with hardware manufactured by every single IHV in existence. You tell me--why should I not trust them? If they set out to deliberately favor one IHV over another, they hurt only themselves. However, if they program to an API so that presumably everyone's API-compliant hardware can run it--they certainly cannot be held responsible for performance differences among IHV products when running their software. Especially, when they demonstrate a willigness to spend inordinate amounts of time and money to create custom code paths in their software for a specific IHV in order to *help* the performance of the IHV's products in their software. Simply put: I have no reason to distrust Valve. OTOH, I have ample reason, ample precedent, to distrust a certain IHV at this point in time.

Exactly! But neither are you.

Ah, but you see you are the one doubting, not me. I am satisfied thus far that Valve is telling the truth. I consider it highly unlikely that Valve would risk the success of its upcoming game sales by deliberately standing up and lying about an IHV's hardware before the game begins shipping! That would be an enormous blunder of huge proportion for Valve. I do not think Valve is that stupid--sorry.

Why do I have to rely on them again? Oh right, because it's important that I immediately bash Nvidia on every free whim. It's obvious I can't hold my tongue and wait for evidence to be presented.

Other than inapplicable racist quips, you haven't provided any "proof" whatsoever as to why Valve should not be trusted. Please do explain in concrete terms how it is in Valve's interests to lie about this--please "prove" your case.

I agree, according to those screen shots, the 51.75 drivers don't appear to have the best image quality with the AquaMark 3 benchmark, but then again, the assumption is still being made that it is because of color precision issues without offering much in the way of support. Additionally, I've yet to see how the 51.75 series drivers affects IQ in HalfLife 2, if at all. My question is, would a loss in precision in any way affect the variation of colors displayed?

If the loss in IQ is perceivable and can be captured in a legitimate screen shot, does the precise *cause* of the loss in IQ really matter as to proving whether it exists? I don't think so...seeing is believing, as the saying goes. The other thing is that nVidia is exceptionally tight-lipped about what it's doing, up to and including encrytping its D3d drivers, apparently. It's unlikely the precise cause will be revealed by nVidia, so all that's left is for people to make educated guesses about what they've done. The only thing nVidia has admitted to in its recent statement is rerouting some shader code from 2.0 to 1.4.

The onus here is on nVidia--not Valve. That's what Valve wants to explain. Also, you apparently have forgotten to think about what would have happened if Valve had waited until the game shipped so that people would have discovered these performance discrepancies on their own. If Valve had kept silent it would have been assumed that, in accordance with the ATi bundling deal, Valve had sold its soul to ATi and rigged the game to run slower on nVidia's hardware. It would have been much tougher for Valve to come back with this info at that time, IMO. Doing it now is preemptive, and clearly places the onus where it should be--on nVidia. Valve simply wants people to know that they tried, and tried hard, because they recognize it is not in their interests for their softwarte to perform so poorly on nV3x running the DX9 code path. At the same time, despite what they have done, it does. I agree completely with their handling of the situation as they have done--this is much better for Valve than trying to explain it all *after* the game ships. Valve understands full well that by revealing this now they may lose sales to nV3x owners who decide their nV3x hardware is more important to them than HL2.

Edit: typos

 

[bdmosky]

Why would you pollute the discussion with racism? As well, last time I checked, Valve does not own a plantation, and holds no slaves....*chuckle* Here's a clue: it is now 2003--it hasn't been 1860 for an awful long time, and a lot has happened in the meantime of which you seem blissfully unaware. Do yourself a favor and try and catch up.

You see, the eye roll was to demonstrate my sarcasm which apparently doesn't catch on easily to you. Next time I'll [sarcasm] [/sarcasm] it for you.

You tell me--why should I not trust them?

You're skirting issues again. I mentioned nothing about trust. Trust != Proof. You can trust them all you want. I have trust in people's judgements too, but I need that trust backed up with evidence and proof.

Ah, but you see you are the one doubting, not me. I am satisfied thus far that Valve is telling the truth.

I'm baffled by your twisting of words here. Need I remind you what that quote was a response to?

You told me:

Face it: you are not in a postion to prove or disprove anything yourself

By saying this, you are actually implying that you are in a position to prove or disprove just because you believe Valve is telling the truth, and I am not because I'm not willing to believe Valve without evidence? Am I correct on this implication?

Other than inapplicable racist quips, you haven't provided any "proof" whatsoever as to why Valve should not be trusted. Please do explain in concrete terms how it is in Valve's interests to lie about this--please "prove" your case.

I didn't start posting here because I was trying to "prove" a case that Valve shouldn't be trusted. Once again, YOU brought up trust. Let me restate my point again, since apparently you must have missed it last time.

My point... respected members of the community can lie too, and no my friend, that isn't enough PROOF for me.

Can doesn't mean are, it only means they have the ability to. I'm merely stating the possibility. I'm even confused as to why you would think I would want to start a crusade against Valve. Last time I checked, they never even made direct allegations against the detonator 50 series of drivers, YOU did.

If the loss in IQ is perceivable and can be captured in a legitimate screen shot, does the precise *cause* of the loss in IQ really matter as to proving whether it exists?

YES! If you are trying to make a case against Nvidia, a precise *cause* of the IQ loss must be made. Just because a scene in any game isn't rendered correctly doesn't automatically make it a cheat. If you are assuming this, then hell, all IHV's have been cheating since their creation. As far as the rest of the crap you wrote, encrypted drivers aren't proof, and neither is silence. That's why in the United States we have the 5th amendment to be able to remain silent without incriminating ourselves. Any implications you may draw from silence cannot be used to PROVE a case.

The onus here is on nVidia--not Valve. That's what Valve wants to explain. Also, you apparently have forgotten to think about what would have happened if Valve had waited until the game shipped so that people would have discovered these performance discrepancies on their own. If Valve had kept silent it would have been assumed that, in accordance with the ATi bundling deal, Valve had sold its soul to ATi and rigged the game to run slower on nVidia's hardware. It would have been much tougher for Valve to come back with this info at that time, IMO. Doing it now is preemptive, and clearly places the onus where it should be--on nVidia. Valve simply wants people to know that they tried, and tried hard, because they recognize it is not in their interests for their softwarte to perform so poorly on nV3x running the DX9 code path. At the same time, despite what they have done, it does. I agree completely with their handling of the situation as they have done--this is much better for Valve than trying to explain it all *after* the game ships. Valve understands full well that by revealing this now they may lose sales to nV3x owners who decide their nV3x hardware is more important to them than HL2.

[sarcasm]Well, thanks for speaking on Valve's behalf. I'll let them know that their PR staff is no longer needed because you will do it for them for free.[/sarcasm]

Please note the statement directly above was only intended for sarcasm and is not to be taken as a serious belief of mine.

 

[TheMightyPuck]

P.P.S. Since you brought it up: Many cases are won on circumstantial evidence alone. You don't always need habeus corpus, sometimes means, motive and opportunity are enough.

res ipsa loquitor baby

 

[WaltC]

You're skirting issues again. I mentioned nothing about trust. Trust != Proof. You can trust them all you want. I have trust in people's judgements too, but I need that trust backed up with evidence and proof.

Ah, but I think trust is precisely the issue. If you take your car in to get a valve gasket replaced, return in the afternoon to pick it up and pay the bill, and your car drives fine--are you going to get home and disassemble the engine to get to the valve gasket just to see if it was really replaced with a new one so that you can "prove" it--or are you going to trust that the work outlined on the bill you paid was actually done? Most people will trust that the work was done and won't require "proof" beyond the word of the shop (and of course that the car drives satisfactorily.)

I bring this up merely to illustrate what a common, ordinary thing trust is and to show how much we all rely on it daily, many times without thinking about it.

In this case, Valve is the mechanic and HL2 is their engine. Yes, I'm going to trust what they tell me about it up until I have a reason not to. I am not going to insist that Valve dissasemble its HL2 code and send it to me so that I can "prove" whether or not what they say is true. As they are releasing their benchmark in two weeks or so that will become evident shortly. I'm not going to waste my time distrusting them in the meantime.

Again, what is your reason to doubt what they've said? Either you have a reason to doubt it, or you don't.

[sarcasm]Well, thanks for speaking on Valve's behalf. I'll let them know that their PR staff is no longer needed because you will do it for them for free.[/sarcasm]

Please note the statement directly above was only intended for sarcasm and is not to be taken as a serious belief of mine.

Well then, let me respond to the sarcasm...

If you'll go over the written accounts of what Valve said, and review the reproductions of the slides they used which are available from certain web sites, you'll see that I was only reiterating what they said in their presentation. And so yes, you're right, they don't need a parrot for their PR...

 

[bdmosky]

Ah, but I think trust is precisely the issue. If you take your car in to get a valve gasket replaced, return in the afternoon to pick it up and pay the bill, and your car drives fine--are you going to get home and disassemble the engine to get to the valve gasket just to see if it was really replaced with a new one so that you can "prove" it--or are you going to trust that the work outlined on the bill you paid was actually done? Most people will trust that the work was done and won't require "proof" beyond the word of the shop (and of course that the car drives satisfactorily.)

I bring this up merely to illustrate what a common, ordinary thing trust is and to show how much we all rely on it daily, many times without thinking about it.

Actually I think your illustration better supports my point than yours. You don't take the mechanics word for it. Taking the car home and discovering that the car runs fine is evidence that the car was fixed. I trust the mechanic enough to believe that if my car is running fine, he must have fixed the problem, but you can sure bet I wouldn't loan out my car to a friend if I hadn't test driven it myself first. My point... even you probably don't trust completely blind.

Again, what is your reason to doubt what they've said? Either you have a reason to doubt it, or you don't.

Because they aren't really saying anything. They are implying things, but they're not outright stating them, nor are they showing evidence of these implications. Oh, and btw... apparently this screen grab "feature" they mentioned (but didn't cite where it could be observed) is broken for AM3 with the 51.75 drivers because it shows lots of rendering errors in it's screen grabs and I wish it to render at least the screenshots correctly.

*edit* paragraph below added...

What I wish Valve would have/still would do is to outright state whether or not the reason why they 51.75 should not be used for these previews is because they exhibit inappropriate behavior and then demonstrate and provide evidence of where this occurs. Yes, from what we've seen of the 51.75 drivers, they exhibit a few issues in AM3, but are they also apparent in Halflife 2 and if so, what and where? If you want to just take Valve's word for it without evidence... fine, but the least they could do is state exactly what they are implying so you don't have to do it for them.

 

[martrox]

This whole "discussion" is bordering on absurdity. First...this isn't a court. Second.... If those of you supporting nVidia in this "discussion" would just try a moment to look at it from a different point of view........ just answer the following questions.....

1) Has nVidia been acused of cheating in their drivers within the release of the GFFX series of video cards?
2) Has there been any proof of said cheating?
3) Just how far can you go back and still find said cheating?

Now, after you have answered these questions, just ask yourself this:

In MY opinion, after answering the previous questions, what are the odds that the latest drivers might have cheats in them?

Now, try to ask yourself the following question:

If I had caught my employer cheating me out of 10% of my pay in each of my last 6 paychecks, would I trust him to give me correct amount on the 7th paycheck?

 

[Dio]

Actually I think your illustration better supports my point than yours. You don't take the mechanics word for it. Taking the car home and discovering that the car runs fine is evidence that the car was fixed. I trust the mechanic enough to believe that if my car is running fine, he must have fixed the problem, but you can sure bet I wouldn't loan out my car to a friend if I hadn't test driven it myself first. My point... even you probably don't trust completely blind.

There are people - I hesitate to say so but my brother is probably on this list - who wouldn't know if their car hadn't been fixed properly unless the engine fell out on the way home. Who are they to trust?

Most people know an awful lot about cars, compared to what they know about 3D particularly on the image quality side. There are people I work with - I ain't one of 'em - who can tell you the monitor gamma from glancing at the screen, and spot single-pixel rendering errors. I can only do that on Quake 2 and Quake 3 timedemos and even that's probably lapsed by now.

 

[bdmosky]

This whole "discussion" is bordering on absurdity. First...this isn't a court. Second.... If those of you supporting nVidia in this "discussion" would just try a moment to look at it from a different point of view........ just answer the following questions.....

I'm not sure if you were implying me in this, but I'd just like to state that in no way have I defended Nvidia's actions here or even stated that they are innocent. I'm merely stating that I think it's absurd to make implications into fact when exact accusations haven't even been made (by Valve... it's fairly obvious that people here are making accusations). Just because I'm not riding down Nvidia's back over this thing doesn't mean I'm supporting them. I'm merely trying to stay impartial and please feel free to correct me if I ever fall off the fence.

1) Has nVidia been acused of cheating in their drivers within the release of the GFFX series of video cards?
2) Has there been any proof of said cheating?
3) Just how far can you go back and still find said cheating?

I think these statements could be applied to any IHV...

If I had caught my employer cheating me out of 10% of my pay in each of my last 6 paychecks, would I trust him to give me correct amount on the 7th paycheck?

Yeah, that's great analogy and all, but this actually happened to me. For an extended period of time it went unnoticed by me that one of my previous employers hadn't been paying me for the extra half hour of work I had been putting in every day that they had requested I put in. You bet your ass that as soon as I confronted them with the issue they immediately resolved it, but then again, that was my former employer and not Nvidia. You can make your own opinions about them, but please don't try to pass speculation as fact no matter your opinion of those involved.

 

[banksie]

Additionally, I've yet to see how the 51.75 series drivers affects IQ in HalfLife 2, if at all. My question is, would a loss in precision in any way affect the variation of colors displayed?

Yes.

We saw this with 3Dmark03 where shader replacement of the sky texture for Game Test 4 reduced the sky from a 'sunny day with clouds' to a much duller 'overcast day' look. The colour palette was much more muted and it was one of the giveaways that something dubious was going on.

Accordingly the muted colour palette, blurryness of textures and changes in distance rendering (like the mountains behind the sub in the oft referenced screenshot) show that a variety of work reducing tricks are being used. One of which is likely to be a lower precision shader for colouring the sub.