X800's Irrelevant?

Price I can NOT argue with as they're both just god-awful expensive toys, and while I keep hearing about availibilty problems the fact of the matter for me is that when my 9700 pro died I was able to find and order the card I wanted on the same day and it shipped to me in under 40 hours.

Now I'm sure that ain't the case for everybody, but I really didn't have any trouble finding one.

EDITED BITS: Oh, and I priced out and had lined up a 6800 to get too....I found both but finally decided on the X800.

 

So why can't the reason for 'nvidia fans' passing it over be the same as the reviewers' ? You can't just tell the masses 'I think it looks better' and expect them to believe you. This is much different from the AA battles of yore where the superiority of ATI's RGMS and 6x support were easily demonstrable.

Didn't know a method was developed for quantifying IQ

 

Sure, it's called "subjective opinion".

 

I think a lot of the lack of respect for TAA comes from confusion over how it works. Even Dave and Neeyik claimed that you have to maintain performace equal to your refresh rate for it to work, and if that was the case then TAA would be a lot less useful. Also I have seen a lot of people with the missunderstanding that one you fall below the threshold TAA turns off and doesn't ever come back on, which leads to claims such as "its no good if you can't keep framerate above the threshold"; when in fact you might as well leave it on even if you barely ever go over the threshold becuase it doesn't hurt anything and in those few moments you are above the threshold it will help image quality. Yet another gripe I have with reviews of TAA is that people will benchmark it and compare the scores to those taken without vsyc. Granted, benchmarking should generally not be done with vsync; but if you are going to compare peformace of TAA which forces vsync it is only fair to enable vsync for the non-TAA benchmarks as well.

 

Question: why?

Why benchmark without v-sync when most people who have a high-end card will use it?

 

Same reason why nvidia fans wont let ati fans pass over sm3.

Sm3 does jack shit (parden my french) It just speeds up the 6800s which are slower than the x800xt pe in sm 2.0 and higher games and it doesn't let the 6800 pass the x800xt pe (in the examples we have)

So whats hte point of bringing it up. IF it was superior then surely the 6800s would be faster than the x800 which is just not the case .


Also I can easily show you the diffrence between ati's temporal and nvidias normal fsaa .

Just come over and look.

I never understood the point of comparing screen shots anyway. I don't play a screen shot. The game can look perfect in a ss , but in motion it can look like a mess (look at the fx fiasco) .

Sure there is .

We already know that ati's 2x fsaa looks worse than its 4x fsaa .

So right there we are quantifying IQ .

We know that 2xt is better than 2x . We know its not quite as good as 4x but very close to it (i think it looks the same but playing it safe here)

We know that 4x looks good but 4xt looks better .



Not to mention that I can go to 2x3 (regular temporal is 2 frames) and get almost 6x quality at the speed of 2x fsaa .

As i said this should be talked about over and over again.

Because it is better iq at the same performance.

Yet as i said nvidia fans will fight it to the death because the x800s and 9800pros will be doing les s work than the nvidia cards. And we all know from the fx fisaco only nvidia is allowed to do less work and get away with it

Only catch is this time ati is doing less work and giving higher iq where as nvidia did less work and gave lower iq .

 

I do kind of agree with HB though. ATIs card on paper look like the should be faster. But in real life they are not much/if any faster. And I know there are many, many, many things that effect the overall profromance. And I can see why HB is upset with them, ATI was the clear leader with the R300 cores. Now its a toss up..which is good for us buyers

 

Next time I cross the river I'll be sure to look you up

Those are all qualitative statements. I was referring to your twice the IQ statement earlier.

What are they fighting? The only thing they have to go on are the opinions of ATI card owners (some of whom are a tad zealous). TAA has not been widely regarded or investigated by the gaming or review press so there is nothing for 'nvidia fans' to fight against. On the other hand SM3.0 has had a lot of exposure and non-6800 owners can base their opinions on the screenshots / reviews etc.

 

Heh. well , i consider twice the samples twice the iq quality.

But hey whatever.


I will ammend my statement.

You will get iq equal to twice the samples of the fsaa sample rate you picked. So if you pick 2x you would get 4x fsaa quality.

Which is an incease in iq and does save alot of speed.


Is that better ?

 

The same reason you normally don't want to bench with any other framerate cap, the results of any card that can exceed the cap will be limited by that cap. So take a card that gets a sold 90fps in situation and another that gets 250fps in the same situation, yet with vsync at 85hz they both look like they are equal at 85fps. Granted, for gameplay it doesn't rightly matter if you can meet your refresh rate; but but benchmarking with vsync is useless for any attempts to form a comparison of the abilities of various cards.

On the other hand I have often though it would be nice someplace like HardOCP, who claims to be benchmarking gameplay, were to do so with vsync on. It would nice to see something along the lines of "While benchmarking without vsync Nvidia's offerings are notable faster, but due to Nvidia's poor support for triple buffering their hardware stutters down from 85fps to 43fps and then to 28fps, while Ati's drivers allow you to enjoy smooth framerate transitions while benefiting from the image quality provided by vsync." Kyle could go off on one of his little tyrades about it and actually do something good by motivating Nvidia to provide better support for triple buffering.

 

Quoted for truth. 8)

 

Much

 

DW the reason the prices are so high still is that there is low availability. If people were over flowing with x800's and NV40 GT/Ultra the prices would go down at least a bit.

I just wonder if it is intentional, but I think not ...

 

To be honest I have been long arguing that both the 6800 series and the X800 series are stop-gap cards.

With the shift to PCIe and SM3.0 both are stop-gaps intended to give the respective companies time to prepare their next real step up. nVidia has historicly had one hardware version introduce new features and then refines it to make it fully useful in the next release. I see SM3.0 on the 6800 as being precisely that, by the time we get games that utilise SM3.0 they will be targeted at the next stepping of cards which will have SM3.0 performance well above what the 6800 delivers.

ATi seem to have spent the bulk of their effort preparing their SM3.0 part and doing the transition to PCIe. Given that they have already said the R520 is going to be a new architecture different from the R300 core then this will be taking a fair amount of effort to get right. ATi has shown they can deliver stunning parts that are well future proofed - the R300 core is a good example of that. Who would have thought that features would still be being enabled in the hardware like centroid sampling and the temporal AA a goodly two to three years on from the launch?

For me it is the next round where the real goods from both companies lie that will be the big deciding factor in graphics for the next three odd years.

 

I don't think it is intentional either, if they had more to sell both companies would have 'em on the shelves since the demand is still high.

I'm pretty sure high-end memory shortages are the problem for both ATi and nVidia right now.

 

I swear you get weirder every day, give it a rest.

 

Ok before I knew of this forum I had my personal opinion.

Coming to this forum has stregnthend my opinion...

DONT SPEND A FECK OF ALOT OF MONEY ON A VIDEO CARD!

Seriously! We have geforce 6800 ($300 and up) users complaining about texture aliasing.

And now this, benchmarks proving a $500 video card is inferior to a $400 card.

 

All that is true, but I can also happily attest to the fact that you get a FECKLOAD worth of graphics power for your money and it is hella cool. 8)

 

or you buy a 6600GT or X700XT and be happy this generation.

 

Are those about on par with the 9700/9800? (I haven't been keeping up. )