calling all 6800 (gt, u, nu) card users....

[micamica1217]

I've been hearing about poor texture shimmering, cralling, texture alizing, and what I call, AA fall off.....
in other words, it looks like nVidia is not doing FSAA.

here are some examples of what I'm talking about:

http://www.beyond3d.com/forum/viewtopic.php?t=16196

http://www.nvnews.net/vbulletin/showthread.php?t=35592&page=1&pp=15

http://www.mitrax.de/?cont=artikel&aid=24&page=14

http://www.hardforum.com/showthread.php?t=809448&page=1&pp=20


and some pics and some AVIs as well.

4xAA and 16xAF, and a 6800gt
http://img55.exs.cx/img55/4444/cocprob1.jpg


6800 ultra
http://img21.exs.cx/img21/790/6800.png

ATI
http://img21.exs.cx/img21/7308/9800.png


also, take a look at the videos on this page.....http://www.oc-zone.com/modules.php?name=Reviews&rop=conteudo&id=87&page=9

and this new one on BF:1942
http://www.abu-el-mot.de/rs/bf.avi


notice the banding they are talking about?
this is with max IQ on the nVidia cards, all optimizations turned off.


since many review sites have not been totaly honest when it comes to IQ lately....
I'm asking all 6800 card users, if you are noticing some of the problems that are discribed in the links and pics.

if you are, what games?
is it all the time? or only at rare camera angles?
if it's only some, but not all your games, please also list what games are fine.
if you going to post pics, please allow them to be viewed at full size, as small or tiny pics may not show what I'm talking about.
(also note that only AA will be noticable in most still pics. and AVIs are needed to show shimmering or other problems.)

as for what to look for,
start reading the links that I have up above....
look for what I call AA fall off (you should look for nice pure lines that at the bottom to mid section of the image, then from the mid to upper image it looks like there are some jaggies.).
I gave some good examples of pics above as to what I'm talking about.
you can't call them FSAA, as they don't do it in the whole pic.


look also for texture crall, shimering, and moreing....
use the movies as a guide as for what to look for.

use any games you like, yet driving games and flight sims will highten the AA problem.
UT2k4 seems to highten the shimering problem too.
you could also try painkiller as it was noted too.
madden or other football games are a great place to look for true FSAA.

please also list what size and type monitor you have.

note: this thread is not aimed to bash nVidia or the 6800gt/u/nu cards...
comments are welcome, yet let's be adults about this.

mica

[ForlornWS6]

thats pretty interesting, honestly I havent really noticed in my games but I'll definitely take a harder look now, hopefully they address this soon

[Ziggey]

I'll do some screen shots of various games and see what I get.

[Paladine]

Great haven't even got my Leadtek Winfast A400 and already issues. Man, did I just get screwed on my system or what?

Anyhow, I will keep you posted as to what I discover this weekend (when my system is to be done and up and running).

[micamica1217]

I'll do some screen shots of various games and see what I get.

thanks alot dude.

by the way, what I'm also realy looking for is user feed back.

.....what games?
is it all the time? or only at rare camera angles?
if it's only some, but not all your games, please also list what games are fine.
if you going to post pics, please allow them to be viewed at full size, as small or tiny pics may not show what I'm talking about.
(also note that only AA will be noticable in most still pics. and AVIs are needed to show shimmering or other problems.)

any feed back you can give would be realy helpfull....
yet if you can anser all the questions then it would be great too.

a big thanks in advance.

mica

[fafnir]

6800 GT - > Ultra here, no problems at all, fsaa 8x16 aa stuff works fine and all, just a bit slow if everything's is on

you sure its not them drivers?

[micamica1217]

6800 GT - > Ultra here, no problems at all, fsaa 8x16 aa stuff works fine and all, just a bit slow if everything's is on

you sure its not them drivers?

that's just it, it seems the complaints are with all the drivers.
beta, or WHQL.

now, while I'll reread the complaints (oh my god, not again. lol), I think that most are talking about 4xAA and 2xAA...not 8xAA, as that is too slow for most games that people play.
but maybe 8xAA doesn't have the problem, since it's a mix of SS and MS.
I also will say that IIRC, AF has nothing to do with the problems, and that only once AA is turned on, does some of the problems arise.

then there is the problem with at least the lower AA modes (2x and 4x settings)...and what I call AA fall off.

so please help me/us, and also try 2xAA and 4xAA and also tell us at least what monitor you use, what res your playing at, and what games your testing.

btw, do you have any of the games that the people are complaining about?

btw, I at least want to thank you for your time, and hope you can retest and get back to me on the questions that are asked.

mica


edit: I was right, it's mostly a problem with 2x/4x AA, not 8xAA

here is two pics with a 6800u and vice city:


4xAA and 8xAF
http://12.22.230.41/MicroTech/Hosted/Images/GTA-vc2004-08-2623-55-30-64.jpg

8xAA and 16xAF
http://12.22.230.41/MicroTech/Hosted/Images/GTA-vc2004-08-2700-03-32-10.jpg

and here is the comment from the author of the pics:

Hey, guys I know whats he's talking about!!

If you look at the pic with the red marks you can see whats going on. Believe me it looks worst in action. Its like a shimmering effect is going on in the background buildings as you walk. It happends at both levels of AA listed in the problem pic and it's even worst at the 2xAA settings.

To solve this problem Terminator use the 8xAA and 16AF settings and you will be good to go!!

I hope nvidia will solve this problem, because my "retired" 9800 Pro did a better job at 2xAA levels -vs- nvidia 4xAA levels within (GTA:VC).

mica

[oc_byagi]

does this issue has to do anything with this article I just found today???
http://www.theinquirer.net/?article=18730

source: Inquirer
Nvidia fails WHQL tests but Catalyst certified

Oh oh

By Fuad Abazovic in Wien: Tuesday 28 September 2004, 14:20
RECENTLY, graphic games are getting more and more intensive and in that light we have some WHQL scores that we can trust.

An Nvidia WHQL FUD document against Catalyst claimed that after Catalyst 4.8 driver, ATI will need to remove their optimisations or it will lose its WHQL stamp.

The Catalyst 4.9 driver is out and ATI still has WHQL, meaning that its driver is fine from Microsoft WHQL test point of view. When we talked with Microsoft we learned that there is a set of tests that you can run in order to see if your driver is WHQL material or not. In other words, you test in house, see whether you can pass the tests and then later submit your driver for certification and get the stamp in less then 24 hours after submitting.

Microsoft can confirm this said us before that all of their nine drivers introduced in 2004 are WHQL and all of their drivers will remain WHQL certified. Nvidia facts are not so bright, they had four partially WHQL while their latest publicly available 61.77 is not WHQL at all.

Our snitch sent us complete report on WHQL test finished at Nvidiai Geforce 5950 card with 61.77. they are failing each and part of WHQL test. They are failing in all fourteen tests including, D3Dlines, Multisampling, Non power 2 conditional RenderTarget, Pixel Shader Precision, Pixel Shader Ver. 1.1, Pixel Shader Ver. 1.2, Pixel Shader Ver. 1.3, Pixel Shader Ver. 1.4, Pixel Shader Ver. 2.0, Point Sprites, Texture address, Texture stage, Update surface and YUV Bit test.

We are not sure what is going on with 6X00 cards and how are they performing but learning from the fact that Nvidia still doesn’t have WHQL drivers on its web page speaks for itself.

Nvidia promised us WHQL drivers a few times already in the last few months since the launch but never released anything yet. What's going on Nvidia?

[Vrykyl]

however, just because a driver isnt whql certified doesnt make it BAD, for example none of my sound card drivers are whql's, yet they/my soundcard sounds great....
Nvidia do need to get their act together tho

[JBistodeau]

Im about to order this video card..do i get something else or keep it?

[PhobMX]

now this is making me think twice if i should get the 6800gt or the x800pro... i guess the doom3 performance difference between both is still apealing ...

[Woozy]

I have not noticed anythign with my 6800 gt, btw what kinda games are we talking about, I play Joint Operations And Delta Force Black Hawk Down. What should I be looking for?

[bobmanfoo]

i have noticed a decrease in IQ with AF optimizations on. really noticable shimmering on textures as you move, like floor and wall textures while running in CS:S. but this is expected as only the primary stage is being rendered correctly with the optimizations on. i don't see these type problems with IQ set to High Quality. as for AA, it sucks, jaggies are only noticable with certain objects, usually simple objects. and crawling and shimmering is just the result of these jaggies moving, note that this is a different shimmering from the kind i see with AF opts. it is obvious that NV does not use FSAA imo. i notice the jaggies and shimmering only on certain objects all the time, and only at certain angles and distances with other objects (objects not on the primary stage), but in all games. some games have alot, some games are barely noticable . im thinking that this are the result of AA optimizationS.
here are some pics of COD at 1600x1200, AAx4, High Quality(no opt.)
PIC 1 (http://www.msnusers.com/myrig/shoebox.msnw?action=ShowPhoto&PhotoID=63)
PIC 2 (http://www.msnusers.com/myrig/shoebox.msnw?action=ShowPhoto&PhotoID=64)
PIC 3 (http://www.msnusers.com/myrig/shoebox.msnw?action=ShowPhoto&PhotoID=65)

[oc_byagi]

however, just because a driver isnt whql certified doesnt make it BAD, for example none of my sound card drivers are whql's, yet they/my soundcard sounds great....
Nvidia do need to get their act together tho
obviously, a driver not being WHQL certified does not mean it's bad, but the article is explaining WHY it's not certified, which is because of the optimizations.

[bobmanfoo]

obviously, a driver not being WHQL certified does not mean it's bad, but the article is explaining WHY it's not certified, which is because of the optimizations.
where is this article you are talking about? thanks

[oc_byagi]

does this issue has to do anything with this article I just found today???
http://www.theinquirer.net/?article=18730

source: Inquirer

previous post by me

[johan851]

That's very interesting. I don't have either card, but I was thinking today about trying out a 6800 NU to replace this here broken 9800 Pro.

If they're not handling AA well (which is very important to me, as aliasing can be very distracting) then I might just get another 9800.

[CandymanCan]

Well going from an Ati 9700pro to the 6800GT i to also have noticed something similer.

The AF on my 9700pro was awhole lot better then the 6800Gt. When my gt is @ like 16xAF and my 9700pro is @ 16xAF the 6800GT looks like its @ 2xAF.

As for AA I'll do some testing o nthat also.


Since my bro's PC is using a 9700pro which i built for him i will test Wolfenstein ET for how well each card uses AA and AF.

Btw this very reaso nwhy i was ganna buy a x800pro instead but everyone said the 6800's were awsome, now i find out nvidia is cheating on thier drivers again... This is sad maybe I'll just stick my 9700pro back in.



EDIT: ok can could someone host my pics since what i have doesnt like me doing that direct linking.

Anyway On wolfenstein ET i didnt notice anything diff, but on BF1942. while on the japs carrier i noticed the runway from far distance was horribly jagged from the distance. When i walked closer that little shimering thing youre talking about happens and the once jaggies in that spot are gone and its as smooth as butter, yet from far way thier are more jaggies on the spot i walked from. That "SHimering" affect i never noticed until now. This total bullsh!t i hope nvidia fixes this because i paid $400 for this card to run like they claimed it and now i find it doesnt. I thought trying nvidia again and giving them a second try was a good idea now its looking bad. :bang head

Remebers the ti4600, i had one when the 9700pro came out nvidia cheated so badly everyoen noticed it the blurryness and whited out screens for exsample. I told myself id never go back to nvidia because of the thing they pulled with the drivers. Now i get a bunch of recomendations saying the 6800 is the king blah blah and now that i have it it totally sucks. Sure its fast but for what price.

Take UT2K4 or exsample. I can play that game on my 9700pro @ 2xAA 8xAF all details o nhighest and 1280x960 and still get good framrates. My Gt i have to run 2xAF 2xAA to even get decent frames. Yet Doom3 runs 1600x1200 4xAA 16xAF no probs...

When u click on the link make sure u expand the pic or else itll look like poop !!

Pic 1 (http://img.photobucket.com/albums/v466/Candymancan/ScreenShot0.jpg)

[micamica1217]

I have stoped bashing nVidia or ATI for quite some time(maybe 5 weeks).....
as some people here don't like seeing or hearing any bad news for a hardware vender that they may like.
I also had a hard time debating if I should even ask owners of the 6800 cards if many of the problems I've heared over the last 4 months were true.

so the main reason I had no choice is because a long time friend who has a 5900u is starting to ask about the 6800gt's.
I've long told him that he will be more then happy with it, since he should no longer be up set with the briliner filtering of his 5900u card.
the performance is great, and every owner loves them.

yet the problems with AA and the 6800 cards don't seem to go away, even with the latest beta drivers.
the funny thing is that most reviews don't do IQ tests, and if they do, they lie about the quality.
maybe they too, just don't know what to look for....even when they post pics.

not that ATI is perfect...
there will always be a game, that at an extream angle you will start to notice jaggies.
yet I've never noticed that half my image was great, and half was not AAed.

I realy just don't know what to say to him on this, as I know that higher res and the use of 4xAA with realy nice framerates is a great selling point for him.
he will not jump ship and get an ATI card, so that is not an option.



i have noticed a decrease in IQ with AF optimizations on. really noticable shimmering on textures as you move, like floor and wall textures while running in CS:S. but this is expected as only the primary stage is being rendered correctly with the optimizations on. i don't see these type problems with IQ set to High Quality. as for AA, it sucks, jaggies are only noticable with certain objects, usually simple objects. and crawling and shimmering is just the result of these jaggies moving, note that this is a different shimmering from the kind i see with AF opts. it is obvious that NV does not use FSAA imo. i notice the jaggies and shimmering only on certain objects all the time, and only at certain angles and distances with other objects (objects not on the primary stage), but in all games. some games have alot, some games are barely noticable . im thinking that this are the result of AA optimizationS.
here are some pics of COD at 1600x1200, AAx4, High Quality(no opt.)
PIC 1 (http://www.msnusers.com/myrig/shoebox.msnw?action=ShowPhoto&PhotoID=63)
PIC 2 (http://www.msnusers.com/myrig/shoebox.msnw?action=ShowPhoto&PhotoID=64)
PIC 3 (http://www.msnusers.com/myrig/shoebox.msnw?action=ShowPhoto&PhotoID=65)

bobmanfoo,

I would like to personaly thank you for your well detailed reply and IQ pics that seem to be a great example of what I'm talking about.

I would also like to thank you for also noting that like I said, the cralling or shimmering is not like we get with textures from time to time, but is caused from the lack of AA applyed to some parts of the final output.
I hope I was clear on that in my original post.
(yet there can still be some texture shimmering such as the examples of UT2k4 and painkiller in high quality mode, that seems to be a LOD issue.)

mica

[Paladine]

Awwwwhh man, this really sucks!!!!!! :mad:


:bang head :mad:

And I thought I had done my research before buying the parts for the system. Oh man. Will Nvidia ever get their head out of their a$$??

I should have gotten a x800 Pro instead. I had to be different! Nutz! :bang head :cry:

[Chowdy]

While i'm not going to run out and sell my card immediately, i'm still pissed off. This had BETTER be fixed with little hit via software, or there'll be hell to pay.

[CandymanCan]

Ok guys changing the LOD in direct3d and Opengl using rivatuner to 0.3 positive fixed all and most of the shimering,

The shimmering is prob thier because the imagei s so clear so Chowdy says and that makes sense, changing the LOD to 0.3 blurs it a little which in turn gets rid of the shimmering

[micamica1217]

Ok guys changing the LOD in direct3d and Opengl using rivatuner to 0.3 positive fixed all and most of the shimering,

The shimmering is prob thier because the imagei s so clear so Chowdy says and that makes sense, changing the LOD to 0.3 blurs it a little which in turn gets rid of the shimmering

reducing one problem by adding another is not what I call a solution.

just my two cents.

mica

[Chowdy]

reducing one problem by adding another is not what I call a solution.

just my two cents.

mica

It's not really a fix. there is little IQ difference, but it's still a little more blurry. It does fix the problem with the ground's "shimmering", which is what all these people call the circular weird pattern? I am not totally clear. Anyways, blurring it via raising LOD to 0.3 makes it go away, while the rest of the game looks 99% as clear as before. We just felt that maybe it implies that it's some sort of strange graphical optical illusion, or something weird like that. I duno. Nvidia's awesome driver support better shine through on this issue.

[micamica1217]

It's not really a fix. there is little IQ difference, but it's still a little more blurry. It does fix the problem with the ground's "shimmering", which is what all these people call the circular weird pattern? I am not totally clear. Anyways, blurring it via raising LOD to 0.3 makes it go away, while the rest of the game looks 99% as clear as before. We just felt that maybe it implies that it's some sort of strange graphical optical illusion, or something weird like that. I duno. Nvidia's awesome driver support better shine through on this issue.

I know what your talking about, "more" patterns.

http://www.extra.research.philips.com/graphics/comparison.gif

mica

[Sentential]

Just to clear it up, mica i know what you are refering to. Before you point the blame SOLELY at nV, rest assured it is a software issue. Very few games recognise the 6800 the way it should. Instead they try to run it like an FX.

Observe. This is America's Army with 8XS AA and 8X AINSO forced:
http://img.photobucket.com/albums/v180/Sentential/AAshot.jpg

^^ I dont see any issues. However when I play farcry I get it when I use the software modes.

Doom3 Using "software":
http://img.photobucket.com/albums/v180/Sentential/shot00009.jpg

http://img.photobucket.com/albums/v180/Sentential/niceone.jpg

(On farcry using "software")
http://img.photobucket.com/albums/v180/Sentential/FarCry0001.jpg
http://img.photobucket.com/albums/v180/Sentential/FarCry0005.jpg
http://img.photobucket.com/albums/v180/Sentential/FarCry0010.jpg

Ill do another round with FSAA forced when I get my 6800GT back from RMA :cry:

SO before people cry bloody murder, how about we wait for software companies to patch their stuff. Does anyone remember how bad Farcry looked on version 1.0?

[Mr.Guvernment]

however, just because a driver isnt whql certified doesnt make it BAD, for example none of my sound card drivers are whql's, yet they/my soundcard sounds great....
Nvidia do need to get their act together tho

As said - look at all of the test's NVIDIA failed as far back as PS 1.1!!!!! - WHY did they fail those tests and WHY is it ATi can pass them......??

[micamica1217]

Sentential,

the 5900 cards had FSAA.
if the 6800 cards are being recognised by the games as a 5900 then we should get FSAA in all games.

but at least farcry (with 1.2 patch) and doom3 is reconised as a true 6800.

btw, the 8xAA is not effected, as it's only a problem with 4xAA and 2xAA.
forced or not.

yet, I'll look more into your ideas.

mica

[Sentential]

Like I said... Ill do same pix once I get my 6800GT back from RMA. Ill do 3 split screens of the same image, showing each level forced into the game.

Also, just cuz they had it doesnt mean they actually *used* it

[Chowdy]

Sentential,

the 5900 cards had FSAA.
if the 6800 cards are being recognised by the games as a 5900 then we should get FSAA in all games.

but at least farcry (with 1.2 patch) and doom3 is reconised as a true 6800.

btw, the 8xAA is not effected, as it's only a problem with 4xAA and 2xAA.
forced or not.

yet, I'll look more into your ideas.

mica

Well, if it really is being read as a 5900FX in games, doesn't the NV40 use a new, different type of AA architecture from the NV3x? Perhaps this is the cause of conflicts.

EDIT: Dang sent, what happened to your 6800GT?

[micamica1217]

Well, if it really is being read as a 5900FX in games, doesn't the NV40 use a new, different type of AA architecture from the NV3x? Perhaps this is the cause of conflicts.


the x800 chips are far from the closest last generation too, if that were the case.
infact it is well known that the x800 has a far diff aligorithum of AF as compared to the 9800 chips.

yet there are plenty of games this stuff happens in that are already patched up for the new chips....
that's why, it must be looked into deeper.

but AA is a image output onto the frame buffer or other part of the chip, it has nothing to do with any software so to speak....
in other words, even if the game has a setting in the GUI, the AA is done on the chip durring the rendering process, not the game.

mica

[Sentential]

the x800 chips are far from the closest last generation too, if that were the case.
infact it is well known that the x800 has a far diff aligorithum of AF as compared to the 9800 chips.

yet there are plenty of games this stuff happens in that are already patched up for the new chips....
that's why, it must be looked into deeper.

but AA is a image output onto the frame buffer or other part of the chip, it has nothing to do with any software so to speak....
in other words, even if the game has a setting in the GUI, the AA is done on the chip durring the rendering process, not the game.

mica

Once again, we will see. I get my GT back from RMA today, Ill repost some screenies.

(chowdy64....dont ask, its a painful story. Ie: Most likely lightning / ??? killed my 6800GT, from the hurricane)

[micamica1217]

a reminder bump for Sentential. ;)

thx in advance.

mica

[Sentential]

Jup got a few for ya. Give me some time tho. Ill post in the other thread for the mean time.

[Sentential]

Just to lay this BS to rest. They do infact do FSAA, and here is your proof.
http://img.photobucket.com/albums/v180/Sentential/LayeredProxycon.jpg

[Paladine]

Well I am stilling just an observer of sorts. Looks like my system (see sig) will not be ready for weeks, since the 6800's (plain, GT, and Ultra) are all on back order here in Canada, with no signs of coming in for awhile.

Sentential, you make a very good point as pictures do speak a thousand words, so I will excitingly wait with much enthusiasim for my Gainward NFX6800GT card to arrive...

[Sentential]

Sentential, you make a very good point as pictures do speak a thousand words, so I will excitingly wait with much enthusiasim for my Gainward NFX6800GT card to arrive...
Im sure you will love it :beer: ...gotta love them Golden Samples. Guarnteed Ultras for $100 less

[Chowdy]

Very sweet post, sent. So it seems it's a problem with a few select games, and not the hardware itself implying a driver / software issue.

[micamica1217]

Just to lay this BS to rest. They do infact do FSAA, and here is your proof.


I'm not quite sure how your IQ pic shows that the 6800 cards do FSAA...???
after all, I'm still seeing better quality AA on some parts of your pic(s) (the 2xAA and 4xAA) that are doing a great job, and some parts that are doing a poor job.

I also understand that the images are not of the same frame, so I'm not suprised that you are getting less AA on the box(s) at the front of the image on 4xAA then on the 2xAA pic.
(but as funny as it sounds, your 2xAA pic has better AA then your 4xAA pic in some parts...take another look, it's a night and day differance.)

but, I am still seeing parts of each image (the 2xAA and the 4xAA) that is better AAed then the rest. (now that is not a night and day diff, and I understand that most won't notice)

here is a IQ comparison of AA in both the x800 and 6800 cards.
the top is ATI, the bottom is nVidia.


this pic is at default benchmark settings, 1024x768, no AA and AF
http://www.bit-tech.net/images/review/359/IQ-returntoproxycon-default.png

note the lack of detail in the textures of the nVidia image behind the bots, and on the bridge to the door.
it looks like your pic on another thread.


1024x768, 4xAA and 8xAF
http://www.bit-tech.net/images/review/359/IQ-returntoproxycon-large.png

note a few things....
first, the spots that had some texture loss in the above pic is now at full detail on the nVidia pic.

next, let's look at 3 places on the image for FSAA detail:

1st, the two gardrails on the bridge....
note how the gardrail closest to us is AAed quite nice, yet the far one is NOT in the nVidia pic.
starting at this point (the far gard rail) that's were what I call "AA fall off" begins.

2nd, look just above the gard rail next to the door way, you'll see the number "2" on the wall (behind the pipes).
the lines in the wall are also not AAed as good as the ATI image.

3rd, the beam that holds up the bottom of the vat that falls in the demo and benchmark.
it too is no longer AAed as good as the rest of the nVidia pic.

(I wish that the examples showed even more of the bottom of the tests since you will see that in some games or benchmarks, this problem seems to be a "near to far" problem.
yet as noted by people who have the cards....
some games are more noticable then others, and some seem to be a problem at certain angles.)

let me be clear, I AM NOT BASHING NVIDIA AT THIS TIME FOR ANY LOSS OF TEXTURE IQ OR SHADOWS, AS IT IS FUTUREMARK'S CHOICE TO ADD VENDER SPICIFIC OPTIMISATIONS THAT HAVE NOTHING TO DO WITH DX9.0c.
(yet if you look at both the ATI shadows on the door, it is much softer then the nVidia ones...DTS, you gota love it.)

but while you have tryed to show how nVidia does FSAA in at least 3dmark05, I see things quite differantly.
the ATI pics, taken at the same frame shows FSAA in all parts of the image...
I can't say the same for nVidia.

mica

[CandymanCan]

I also have AA problems in Ut2k4 and farcry.

Sometimes whe ni enable AA in farcry to high or whatever my screen isnt fully maximized and i cant change it. Also on Ut2k4, when i enable 4xAA it looks like 2xAA, and 8xS it looks like 4xAA

[PhobMX]

helluva post mica, now im right in the middle of the fence whether to buy an x800pro and softmod to an x800XT PE or just getting the 6800gt.... Those pics show clearly ill miss the smooth textures and aa of my r9700, so if nothings stops me (like decent drivers) in the upcoming 4 months, ill fallback to atis side.

If ati would boost d3 performance a little, but sigh...

Its like when you first play at 1280 res, when you try 800x600 its looks like crap, its the same looking at texture quality and shadows between ati and nvidia in those pics...

i used to be so sure id buy a gt because its technicaly better, but now in not sure at all

[micamica1217]

helluva post mica, now im right in the middle of the fence whether to buy an x800pro and softmod to an x800XT PE or just getting the 6800gt.... Those pics show clearly ill miss the smooth textures and aa of my r9700, so if nothings stops me (like decent drivers) in the upcoming 4 months, ill fallback to atis side.

If ati would boost d3 performance a little, but sigh...

Its like when you first play at 1280 res, when you try 800x600 its looks like crap, its the same looking at texture quality and shadows between ati and nvidia in those pics...

i used to be so sure id buy a gt because its technicaly better, but now in not sure at all

you'll get a PM as I want to keep this on topic.

mica

[CandymanCan]

I personally still dont understand why everyone say Nv drivers are better then ATI, I miss my 9700pro already im comming across alot of problems with my Gt, all of which are driver problems. For exsample all i mean all the 66.XX series drivers for me always reboot my pc during games but 3dmark runs perfectly fine.

The 61.77 drivers sometimes my AA and AF settings work but sometimes they just stop working. Now we are basicly talking about NV cheating or as they say " Optimising" thier drivers for 3dmark's.

I personally think Nvidia agenda is to make thier drivers run well on 3dmark but suck on every game out thier. My 9700pro like the person said above hadm uch better image and AA quality it just isnt as fast as the Nv40

[Chowdy]

I personally still dont understand why everyone say Nv drivers are better then ATI, I miss my 9700pro already im comming across alot of problems with my Gt, all of which are driver problems. For exsample all i mean all the 66.XX series drivers for me always reboot my pc during games but 3dmark runs perfectly fine.

The 61.77 drivers sometimes my AA and AF settings work but sometimes they just stop working. Now we are basicly talking about NV cheating or as they say " Optimising" thier drivers for 3dmark's.

I personally think Nvidia agenda is to make thier drivers run well on 3dmark but suck on every game out thier. My 9700pro like the person said above hadm uch better image and AA quality it just isnt as fast as the Nv40

I have no problem with Nvidia drivers, and i'm using betas. I had a helluva lot more problems with ATi drivers, namely restarts as you're talking about. My computer is 100% stable, as it survived a 12hr+ lan party of UT2004 and WC3.

Maybe *some OC instability?

*EDITED

[ForlornWS6]

I personally still dont understand why everyone say Nv drivers are better then ATI, I miss my 9700pro already im comming across alot of problems with my Gt, all of which are driver problems. For exsample all i mean all the 66.XX series drivers for me always reboot my pc during games but 3dmark runs perfectly fine.

The 61.77 drivers sometimes my AA and AF settings work but sometimes they just stop working. Now we are basicly talking about NV cheating or as they say " Optimising" thier drivers for 3dmark's.

I personally think Nvidia agenda is to make thier drivers run well on 3dmark but suck on every game out thier. My 9700pro like the person said above hadm uch better image and AA quality it just isnt as fast as the Nv40

look around and see how many problems there are with nvidia drivers compared to ATI drivers, while its unfortunate your having these issues, I, like chowdy have had zero problems with my card

now back to the topic at hand hopefully I can reinstall all my old games this weekend after my recent format and see if I get any of these AA issues I'll chime back if I notice anything

[PhobMX]

look around and see how many problems there are with nvidia drivers compared to ATI drivers

i dont mean to hijack but atis driver suckness is an old dead myth. Nvidias driver suckness (well, maybe not unstability) in iq as this thread is all about has been here since 5x00 days, i thought 6x00 would iron that but the more i look into it, the more it looks as the same old bs

[ForlornWS6]

i dont mean to hijack but atis driver suckness is an old dead myth. Nvidias driver suckness (well, maybe not unstability) in iq as this thread is all about has been here since 5x00 days, i thought 6x00 would iron that but the more i look into it, the more it looks as the same old bs

I wont deny that ATI has come along way with their drivers, but seeing as the 6 series cards are still relatively new I am pretty sure better drivers are still to come

[johan851]

I am pretty sure better drivers are still to come
I've seen people saying this for a long, long time now. A lot of people even buy nVidia cards, regardless of lower benchmarks, simply because they expect new drivers to improve the situation. As far as I've seen, they really haven't. All I hear is problem after problem and too many driver releases to even keep track of.

[PhobMX]

I've seen people saying this for a long, long time now. A lot of people even buy nVidia cards, regardless of lower benchmarks, simply because they expect new drivers to improve the situation. As far as I've seen, they really haven't. All I hear is problem after problem and too many driver releases to even keep track of.

exactly what i mean

forlon, if you where a 5x00 owner, you would be dissapointed by your own belief. 5x00 are the perfect example of good drivers that never came, for ti4x00 series drivers where good at first, but iq went down to hell when the r9700 came up.

My point is that if good drivers are going to arrive, they better do it quick (namely 5 months, which is when im going to upgrade)

[Flip-Mode]

In 3dmark05 in the Canyon Flight game test there are alot of ropes and guardrail shadows wich don't look very good and even shimmer.....
Take a closer look next time you run the bench, could be just my problem?

[Flip-Mode]

Is this what you ware talking about?
Notice how the AA starts at a single point.
http://www2.freepichosting.com/Images/421557827/1.jpg
http://www2.freepichosting.com/Images/421557827/2.jpg
http://www2.freepichosting.com/Images/421557827/3.jpg

[micamica1217]

In 3dmark05 in the Canyon Flight game test there are alot of ropes and guardrail shadows wich don't look very good and even shimmer.....
Take a closer look next time you run the bench, could be just my problem?

yes, there are far too many jaggy shadows in 3dmark05 even on a x800pro.
remember that we are testing at 1024x768 with no AA and no AF.
they are just softer on the ATI cards.

Is this what you ware talking about?
Notice how the AA starts at a single point.

I hate using then type of tests as they are hard to notice at times like games for me...and I'm only talking "for me".
yet look at sent's pic of farcry as a great example....

http://img.photobucket.com/albums/v180/Sentential/FarCry0005.jpg

the trees on the laft are AAed quite well, yet the ones on the right are not.
click on the pic for a full size image...128x1024 with 4xAA, 8xAF

mica

[Sentential]

Its all in context. But yes, there are AAing issues.

[giga_dude]

so what card should I get the 6800GT or X800PE?

:temper:

[austinbmxnig]

after looking at the screens comparing the x800 and the GT...the x800 just looked so much better but with the GT being faster it makes it a tough decision

[giga_dude]

well my 5900u might have better performance than the 9800 but there's no way that I can crank the AA/AF at 1600x1200 and get decent FPS. I'm still playing at 1024x768......... :eek:


....I just want to avoid spending $400 & get disappointed again but it looks like most ATi cards get at least 60fps (good enough for me) and since I haven't heard any ATi users complain about IQ.....now I have to find one........

;)

[CandymanCan]

Man my 9700pro is still better then the Gt even when i change the AA on the GT to 4x Supersampling.

Here are 2 pics, the first is the 9700pro the second is the 6800Gt, notice at the Entrance thier, the ground in it the black lines are all jagged and the 9700pro is all smooth


http://img.photobucket.com/albums/v466/Candymancan/9700pro.jpg





http://img.photobucket.com/albums/v466/Candymancan/Shot00005.jpg

[Hellion]

i dont suppose these could be fixed with drivers? please? maybe? i know, false hope....

[Paladine]

Man my 9700pro is still better then the Gt even when i change the AA on the GT to 4x Supersampling.

Here are 2 pics, the first is the 9700pro the second is the 6800Gt, notice at the Entrance thier, the ground in it the black lines are all jagged and the 9700pro is all smooth


http://img.photobucket.com/albums/v466/Candymancan/9700pro.jpg


http://img.photobucket.com/albums/v466/Candymancan/Shot00005.jpg


Sad or sorry to say..... I just do not see any real difference in those pictures to warrant one over the other.

[Hellion]

Sad or sorry to say..... I just do not see any real difference in those pictures to warrant one over the other.


I do, he sent me the two pics so i could look up close....the difference is...wow. its like its hard to believe its so bad on the gt. youll notice that a lot of things that arent rendered by the 9700 IS by the 6800, but the lines, the jaggedness, the quality OF the pic, horrible on the GT. Makes me wanna cry over a 400$ mistake...was it?

[giga_dude]

All I'm saying is that my MSi 5900Ultra was a disapointment & I wan't to avoid making the same mistakes again....it seems like different review sites post various numbers & it's hard to know which ones are biased & not......

:temper:

[Paladine]

I do, he sent me the two pics so i could look up close....the difference is...wow. its like its hard to believe its so bad on the gt. youll notice that a lot of things that arent rendered by the 9700 IS by the 6800, but the lines, the jaggedness, the quality OF the pic, horrible on the GT. Makes me wanna cry over a 400$ mistake...was it?

Hellion - Well maybe you are seeing something I don't with those pics he sent you, but the ones posted here.....when playing a game as fast paced as UT4K I can hardly see the difference in those two pictures. Honest...and know I am not blind.... :D I did have my eye done with LASIK so there are better than 20/20. :sn:

But I am agreeing with you on the mistake part. I am scr$w'd as my card has not arrived yet, but it is apparently on arrival for late next week. I had to put a deposit down.....Doh! :mad: Now I have $2000 dollars with of computer parts with no case or video card and the card I ordered apparently sucks bigTIME! :temper:

Thanks Nvidia....you fruit cake SLI pushing mothers..... :bang head

sorry Mod's....just frustrated with the money I have spent and I am still unable to use the system!

[Hellion]

Hellion - Well maybe you are seeing something I don't with those pics he sent you, but the ones posted here.....when playing a game as fast paced as UT4K I can hardly see the difference in those two pictures. Honest...and know I am not blind.... :D I did have my eye done with LASIK so there are better than 20/20. :sn:

But I am agreeing with you on the mistake part. I am scr$w'd as my card has not arrived yet, but it is apparently on arrival for late next week. I had to put a deposit down.....Doh! :mad: Now I have $2000 dollars with of computer parts with no case or video card and the card I ordered apparently sucks bigTIME! :temper:

Thanks Nvidia....you fruit cake SLI pushing mothers..... :bang head

sorry Mod's....just frustrated with the money I have spent and I am still unable to use the system!

Thanks Nvidia....you fruit cake SLI pushing mothers..... ? nice. i like it. haha.

[CandymanCan]

Sad or sorry to say..... I just do not see any real difference in those pictures to warrant one over the other.

When you look at the pictures @ 1280x960 you will see the diff, photobucket made the pics smaller cuss of the size.

I still noticed the diff and if u cant then you must be blind

Here are the picsj ust bigger i had to use freeservers.com so if the pics go away saying link forbiddon thats the reason why.Now same order 9700pro first then GT, now if u cant see the diff in these pics then you really are blind.

Click on the picture then expand it to get the full quality


http://residentevil.mods.freeservers.com/images/9700pro.jpg


http://residentevil.mods.freeservers.com/images/shot00005.jpg

[Paladine]

Ok...I can see there is a difference. But I must say that with those views it is hard to say if you would even notice them. Since I am not view them with the game/system/video card setup it is hard to say if I would really notice them or even care.

But your point is made, and I do agree the GT looks worse. Bar None!

:(

Now I am even more upset at having purchased/ordered a GT. Scr$w'd again!

[CandymanCan]

Besides this problem i think even tho youre mad now you will be happy when you get youre Gt, iv been told by PNY and others that Nv is trying to fix this.

[Paladine]

Besides this problem i think even tho youre mad now you will be happy when you get youre Gt, iv been told by PNY and others that Nv is trying to fix this.


That is some of the best news I have heard all day!

Thanks....

[micamica1217]

Besides this problem i think even tho youre mad now you will be happy when you get youre Gt, iv been told by PNY and others that Nv is trying to fix this.

first off, PNY will only use drivers based off of nVidia refrance drivers...that is if you use PNY drivers. :eh?:

as for if nVidia will ever fix this, well, you might want to ask what happened to 5900 owners each time nVidia fixed a IQ problem.
hint: the performance tanked.

I will say, the 6800gt/u cards are good, they just need some WHQL drivers that don't take short cuts (in order to help sell thier products).

mica

[CandymanCan]

Wish i could get the best of both worlds, ATI's driver quality and Nv's speed in games. The only problem i see with Nvida is the AA/AF isnt as good as Ati nothing more but thier cards are faster, and this is why oh well.

[Chowdy]

Wish i could get the best of both worlds, ATI's driver quality and Nv's speed in games. The only problem i see with Nvida is the AA/AF isnt as good as Ati nothing more but thier cards are faster, and this is why oh well.

Nvidia's AF is better than ATi's AA, although not by much. Nvidia's AA is also practically as good as ATi's AA, but a few visual anomalies are twisting perspectives much. The IQ difference, as a whole, is not drastically different. Sure, i'm pretty pissed about the issues with games like UT2004, but it's not all taht bad because i'm extremely confident that it'll be dealt with. And despite mica's coining of the term "AA falloff" i can guarantee there is AA in very far distances in UT2004, or at least farther than where the visual anomalies occur. It's not a lack of AA, but a combination of AA technique and texture optimzations that seems to be causing the problems. And i don' tknow where i read this, but the 6800's AA technique is different from the 5900's, and aren't many games detecting the 6800's as 5900's? I'm not sure what that means, but it's just a guess and i'm throwing things out there.

[micamica1217]

Nvidia's AF is better than ATi's AA, although not by much. Nvidia's AA is also practically as good as ATi's AA, but a few visual anomalies are twisting perspectives much. The IQ difference, as a whole, is not drastically different. Sure, i'm pretty pissed about the issues with games like UT2004, but it's not all taht bad because i'm extremely confident that it'll be dealt with. And despite mica's coining of the term "AA falloff" i can guarantee there is AA in very far distances in UT2004, or at least farther than where the visual anomalies occur. It's not a lack of AA, but a combination of AA technique and texture optimzations that seems to be causing the problems. And i don' tknow where i read this, but the 6800's AA technique is different from the 5900's, and aren't many games detecting the 6800's as 5900's? I'm not sure what that means, but it's just a guess and i'm throwing things out there.

first off, games don't controle how a chip does it's FSAA.
it's the GPU that adds the arithigorum to the image durring the rendering faze and then the final output is displayed on your monitor.

drivers at times can improve or diminish the amount of AA used with application detection as noted with the 5900's and MS FS2004.
I have no idea if it's the drivers that are doing the damage to FSAA and giving users less then FSAA in the prosses.
it could be a hardware problem....I highly dought it, IMHO.

the lack of FSAA runs across ALL games, as noted by most users.
they also said that in some games it's harder to notice then others...I agree with that statement too after seeing more pics then I can count.
I'm also glad that it seems you are only having problems with UT2004, if that.

but with so many users from this forum, as well as the many forums that I listed, it is aparent to me that FSAA is NOT done on many if not all games.
note the use of FS in FSAA...yes, some AA is being used in all examples that I posted.

I also want to point out that most of the complaints are made from people who were using the max IQ settings of the 6800gt/u in the CP.
meaning that all the optimizations were turned off and are still getting the same lack of FSAA as well as other complaints in IQ.

now I also understand that nVidia has changed the arithigorum of 8xAA in order to make it faster, about the same time that the first 6800 cards hit the USA.
this is why I feel that this lack of FSAA can be changed in just the drivers.
but rarely have I seen nVidia change such things till most of the web site benchmarking is done....
mostly because it will lower the scores of the cards.

then you have to ask, "will users even use slower drivers that will help with this problem"?
most don't even know that there is a problem to begin with.

then there are what seem to be two types of nVidia drivers:

gaming drivers - plays games great with awesome IQ.

benchmarking drivers - has IQ issues in some or many games, but gives great performance increases in benchmarks and possably games too.


here is a quote from one of our members on 3dmark05:

I agree. I will tell you that the website also says Forceware 66.51 drivers are the only "approved" by Futermark....they suck. The IQ difference between the 61.77's and the 66.51's is huge, in a bad way. The 61.77's gave me a 4333 score, the 66.51's gave me just over 5k, I can see the missing textures while 05 runs with the 66.51's....I cant believe Futuremark would sanction something so blatantly bogus.....

talking about 3dmark, did anyone notice that we are still waiting for an updated approved driver for 3dmark03 that is newer then the original one when the 6800 cards came out?
my god, it's still 60.85 for the 6800's only.
all other cards are 52.16
WTF is up with nVidia in thier driver dept., are they really run by the same marketing/managering jurks that were around durring the 58/5900 cards?

this might sound like one big rant by me, but think about something first...
why would nVidia change anything when your not "demanding" them to?
(with nVidia card users, always, repeatedly saying that nVidia is putting out the best drivers in the world, and not complaining about a lack of WHQL drivers....why would they change a thing?)

the next question would be, how to demand anything from them.
that's an easy one, and I don't think even I need to go into that.

mica

[giga_dude]

Hmm. Well after a huge disappointment with my 5900u and reading recent posts I'm heading over to ATi and the X800 XT card.

What brand should I get or should I wait a bit more ?

[Chowdy]

Good response, mica.

I'd like to note, though, that it is only a problem with textures, isn't it? The problem does not seem to involve any hard polygons, which is what MSAA anti-aliases. It's quite a striking difference. In UT2004, i can be standing by that corridor that Candyman posted about, and i can look through it and see the jaggies, but things in the distance are perfectly AA'd, but those things are polygons and not the middle of textures. Doesn't this imply an AF + AA conflict / issue?

[micamica1217]

Good response, mica.

I'd like to note, though, that it is only a problem with textures, isn't it? The problem does not seem to involve any hard polygons, which is what MSAA anti-aliases. It's quite a striking difference. In UT2004, i can be standing by that corridor that Candyman posted about, and i can look through it and see the jaggies, but things in the distance are perfectly AA'd, but those things are polygons and not the middle of textures. Doesn't this imply an AF + AA conflict / issue?

I'm seeing problems from hard polygons that textures are layed onto (isn't that always the case), as well as, what, soft polys.

now if the edges of textures seem like they are not AAed, then that is briliner at work, and is a whole differant story.

again, the whole FSAA issue seems to effect each game differantly....
and is at times less noticable then others.
I highly recamend you reread the complaints and relook at many of the pics to see what is going on....this is not a UT2004 issue only.

mica

[Sentential]

...Mica before you start on the AAing / AINSOing parade again. Please dont make me post this:

ATi FSAA is NOT done on many if not all games.
note the use of FS in FSAA...yes, some AA is being used in all examples that I posted. (ATi)
http://www.reflectonreality.com/images/3dm05/1x800.png

nV
http://www.reflectonreality.com/images/3dm05/16800.png

first off, games don't controle how a chip does it's FSAA.
Oh... funny. Sure looks like 3d05 is.

ATi
http://www.reflectonreality.com/images/3dm05/1x800gt2.png
then that is briliner at work (instead its now ATi)
http://www.reflectonreality.com/images/3dm05/2x800gt2.png

nV
http://www.reflectonreality.com/images/3dm05/16800gt2.png

nV
http://www.reflectonreality.com/images/3dm05/6800canyon.png

ATi
http://www.reflectonreality.com/images/3dm05/aticanyon.png

Once again.....BOTH have filering issues. Whether you like it or not.

[Section8]

I was trying to stay out of this but OMG!!!!!!!!!! I just want to say STOP it GIRLS your both pretty.... ATI and NV are neck and Neck. I have owned all 4 of the top cards and unless you spend 4 hours looking over every pixel you can't see a real world difference when you play the game or games. I am sure whatever you like or don't like with either card will or can be fixed with driver updates. Sooner or later...:(

[TimBlake84]

Alright, after looking at all the other posts, I figured this was an appropriate thread to post my own question. I have all the gear listed in my signature still with that current OC coonfiguration. I run 3DMark03 and I am only getting back 8170 marks. I thought this card would do at leats 10K. What is going on here? are my settings off? If everything is okay, what upgrade would boost my mark score the most for the money? Thanks for the help.

[Sentential]

back 8170 marks. I thought this card would do at leats 10K
Yes when overclocked.....

#1 Is your card a 6800 Non-Ultra or GT?
#2 have you set *everything* to performance?
#3 Are you at stock?
#4 Drivers?

If you are at stock, those scores look about right. 8500 or so is what a stock Non-Ultra does. If you want 10k+ you gotta break about 400 | 850. I dont have a link, but there is a way to "unlock" the remaining pipelines with a softmod (to a full 16 pipes, vs your 12). The program is called "Rivatuner". Ill try to find a link

[PhobMX]

nice post sentential first sample i seewhere nvidia outperforms ati in AA in a lot of time, but im not sure what to look for in the trees sample

[micamica1217]

...Mica before you start on the AAing / AINSOing parade again. Please dont make me post this:

ATi (ATi)
http://www.reflectonreality.com/images/3dm05/1x800.png

nV
http://www.reflectonreality.com/images/3dm05/16800.png


Oh... funny. Sure looks like 3d05 is.

ATi
http://www.reflectonreality.com/images/3dm05/1x800gt2.png
(instead its now ATi)
http://www.reflectonreality.com/images/3dm05/2x800gt2.png

nV
http://www.reflectonreality.com/images/3dm05/16800gt2.png

nV
http://www.reflectonreality.com/images/3dm05/6800canyon.png

ATi
http://www.reflectonreality.com/images/3dm05/aticanyon.png

Once again.....BOTH have filering issues. Whether you like it or not.

your reply seems to be almost an exact copy of tranCendenZ, using the same images and all to give examples of how ATI is now messed up in 3dmark05.

so in turn I'm going to use almost the same reply (copy and paste) with some additional examples that are directly related.
I might miss a quote or two, as I just don't have the time to go over this again, and again, and again......when this thread is all about DST (and soundstorm).



when your going to post facts about this test, please include ALL the FACTS.

like how there is not one driver for any IHV at this time, that is official for anything but the default tests.
how about that all the test pics you show are using the app and not the CP of the cards....
btw, when using the CP for ATI cards, all banding and IQ problems are gone.
(I'm still wondering about the fog issue as I don't seem to have this problem...
and it looks like the fud that was started about aquamark bench, and I don't have problems with missing smoke in that as well.)

(as I've said later in this thread, I have started noticeing some IQ differances between SM2.0 and SM3.0, mostly with the water.
while my "SPLASH" is maybe more detailed then the last example of tests, it is less then the nVidia test.
this as well as other examples I can give, only show the rare differances between some SM3.0 and SM2.0 shots.
it seems that futurmark wanted this type of differance, instead of a seporate SM3.0 test...and could also be used as an example, that this test can no longer be used as a "apples to apples" comparison again.)


the reason I've even done this investgation thread is because I've seen tons of threads on this topic such as nVnews, bjorn3d, and others.
many of them are not investigating anything, but spouting gosip and rumors.
even if they have been proven wrong.
btw, bjorn3d has already shown that they didn't use the same AA setting on the ATI and nVidia cards....realy lame. :-/
did you miss this quote by me?

you also failed to note that the author of bjorn3d (edit:my mistake, it was NVnews...lol) could not get 3dmark05 to work with AA when set in the CP, so he thought it was the same way for ATI.
(again, lower IQ results for both cards will result when done in the GUI instead of the CP in this tests...again, isn't that crazy?)
did you miss this quote by me?


you also don't note that not one person could find any more banding in the sky pic with a ATI card, as compared to nVidia...both have realy, realy, realy, small amounts of banding...just in differant spots.
(again, you'll get no banding if done in the CP of the card(s)...and no driver is verifide for anything but the default test.)
did you miss this quote by me?


I will repeat my self, nVidia is not cheating in this test due to DTS...this is a problem that futurmark needs to anser to if you ask me.

I could go on and on, but at least you show where you got your information from...(or lack of it.)
did you miss this quote by me? and now your repeating it? why?

I would post links, or quotes from the threads from them forums, but it would be a waste of time....and anyone here can easily look it up to see how truthfull I've been.
I'll add one or two quotes now...

Well, there's clearly something odd in the first set of screenshots, but IMHO banding is equally evident above in both the ATI and NVidia screenshots. That it occurs at different places in the image is the only difference I could see and that's irrelevant.

Bjorn3d shows Ati's 2xAA against Nv's 4xAA, i've mentioned this already on Nvnews.


looked at the differences between the 2 images posted above...i agree with the moire pattern, but for the other red outlining, i would have to say the ati one is better, it seems more defined to me and if you look at the flooring texture, its clearer as well

i was a bit more curious so i went through the images in bjorn3d and found many differences in each image, which may need a reference rasterizer (i do realize that a ref. rast. may not produce 100% accurate image, but the differences between the cards is big, so at least it will help to determine which card is more correct) to compare against...there are some large lighting and some texture quality descrepencies between the 2 cards....

...edit: someone posted on the futuremark forums and compared between applying AA/AF through the CP and APP and found that when forced in the CP, the moire pattern is gone, and with the app, its there

Screenshot X800XT with AF forced by Ati's CP:

http://members.chello.nl/h.appel1/X800.jpg

I think Ruined's case is closed.

In the default game test there is no moire on Ati card's. Ruined, am I writing in Chinese because i've stated this on page 1 of this thread.



I agree with cowboyX, you could start another thread, as it seems you just want to start something, and not post all the facts in order to help others.
did you miss this quote by me?


but I at least understand why futurmark wants to use DTS as a feature...
you see, many games that will come out will use it.
in fact, Farcry uses it too.

http://img9.exs.cx/my.php?loc=img9&image=Nv40PathGun.jpg

http://img9.exs.cx/my.php?loc=img9&image=ATIPathGun.jpg

http://img19.exs.cx/my.php?loc=img19&image=NV40PathCar.jpg

http://img19.exs.cx/my.php?loc=img19&image=ATIPathCar.jpg

guess what pics are ATI and what are nVidia.

mica

btw, why are we going on about this again, in the wrong thread no less?

mica

[Chowdy]

Why does it matter where you set it, CP or not? I thought the game has nothing to do with it, and it's the drivers at work? Or does this apply to AA only? :confused:

I know it's not a UT2004 only problem, but UT2004 is the only game i see it in.

[Hellion]

another point....is it neccesary to quote the pics too???? waste of bandwidth for the NON cable/ dsl or above users.

[Sentential]

Yes, mica it was essentially a copy/paste, once again to prove the point that both nV and ATi is guily of cheating. Ill see what else I can find.

I will repeat my self, nVidia is not cheating in this test due to DTS...this is a problem that futurmark needs to anser to if you ask me.
AND THE TRUTH SHALL SET YOU FREEEE!!!!!! :D

Finally an unbiased anwser. :attn:

[Sentential]

Conclusion:
When I first read the reviews of the 6800 Ultra I was very impressed to see the problems which affected the NV3x being resolved, there were a few issues with IQ mentioned – such as those in Farcry however these were to be fixed in a future driver and appear to be legitimate bugs. I feel this is not the case however for the issues described above. 3Dmark as we all know is a synthetic benchmark, the purpose of which is to compare cards on a level playing field. Based on the changes in mipmap/textures on the NV40 this is not the case. It is disappointing that with such a technology lead over anything currently released that Nvidia have driver issues such as those described above

Driver Heaven (http://www.driverheaven.net/articles/driverIQ/)

http://www.pcper.com/article.php?aid=40&type=expert&pid=3

http://www.pcper.com/image.php?aid=40&img=N8XAA.jpg
http://www.pcper.com/image.php?aid=40&img=6XAA.jpg
Once again, we see the two cards producing very competitive images with one another once the maximum settings for FSAA are selected. Here, the images produced on the GeForce 6800 cards manages to beat the ATI cards in overall image quality by a narrow but evident margin. Looking under the brim of Tiger’s hat, we see that the NVIDIA cards are able to smooth those jaggies out while the X800 cards still show aliasing in their image.

(Fantastic example of "digital vibrance")
http://www.pcper.com/image.php?aid=40&img=AASlide.jpg

With the maximum level of Anisotropic Filtering enabled for each card, we see the differences between the two images nearly disappear. Interestingly enough, the variance in lighting seen throughout testing is now nowhere to be seen aside from a slightly brighter glow on the top of the door on the ATI cards.
http://www.pcper.com/image.php?aid=40&img=N16XAF_big.jpg
http://www.pcper.com/image.php?aid=40&img=16XAF_big.jpg
http://www.pcper.com/image.php?aid=40&img=AFSlideshow.jpg

NVIDIA-Related Issues

For one reason or another, the reference GeForce 6800 GT board we received refuses to run a handful of games when HyperThreading is enabled. In order to have games such as FarCry or UT2004 function at all, we are forced to dive into the BIOS and disable HyperThreading. Although we asked the question some time ago, we have yet to receive any answer as to what the issue is or when a resolution is expected.

The second issue we face with NVIDIA is a strange bug with the flagship GeForce 6800 Ultra OC. When attempting to use the driver's built-in "Auto-overclocking" feature, the system will reboot during gameplay. Given how we're dealing with an early reference board and ultra-high frequencies, we can be optomistic and register this as a temporary bug. Hopefully, this will be cured once people start spending money on retail versions of this card.



ATI-Related Issues

Aside from an isolate glitch with Splinter Cell: Pandora Tomorrow, we only had one issue with ATI. However, this issue is by far the one which conerns us most. After receiving the cards and exploring through the Beta Catalyst 4.5 drivers, we realized that the OverDrive auto-overclocking options was nowhere to be found. A quick phonecall later and we learned that although the BIOS of the cards supported OverDrive, the feature was not enabled within the driver we received. Although the last year in the graphics industry might have made us pessimistic, we were a bit curious why the option which was present now for several drivers suddenly disappeared for our version. Playing devil's advocate, we installed a Radeon 9800XT with the Beta drivers and saw the familiar OverDrive tab. To make things a bit more suspect, a quick inquiry of the default clockspeeds seen in X800XT reviews around the net yielded several different results. Although Kyle at HardOCP received a board with correct speeds (520/560) as we did, others were above and below those speeds by a considerable margin. Anandtech's board clocked in at 500/550, whereas Tom's Hardware Guide saw 526/574 and ExtremeTech saw 520/600. Our efforts to overclock the X800XT through PowerStrip were limited at best as any increase to the core frequency caused visual distortion of the desktop resulting in a necessary reboot. Looking around the web, we realize that we aren't the only ones having a tough time overclocking this card. So for now, we are left with a unsatisfied urge to push the envelope and a handful of questions. Fortunately, all the questions regarding OverDrive and frequencies will be answered once ATI gives us the option to manually disable the auto-overclocking feature and visually verify that it is not in action.


Techlounge (http://www.thetechlounge.com/review.php?directory=ati_radeon_x800_pro_256&page=12)
Image Quality

To compare the image quality of the X800 Pro and 6800 GT, we took screen captures using FRAPS from a few different games With Doom 3, HL2 Stress Test, and Far Cry, we were able to take the exact same shot with each card. However, with Need for Speed Underground, since there is no way to save in the middle of a race or record a demo, there was no way to take the exact same screen capture every time. In that case, the picture was taken right after a race had started on the Olympic Square map with the car in the same starting position each time. The Half-Life 2 Stress Test posed a similar problem. The Stress Test runs through a map on a predefined course and you can not stand still or save the ‘game.’ Because of this, the screen captures were taken while the camera was moving. While the captures for these two games are not exactly the same pixel to pixel, they are close enough for a good comparison.

http://www.thetechlounge.com/reviews/ati_radeon_x800_pro_256/doom3.jpg
http://www.thetechlounge.com/reviews/ati_radeon_x800_pro_256/nfsu_4x8x.jpg
When AA and AF are enabled, we see once again that the X800 Pro seems to be doing a little bit better than the 6800 GT, but the difference isn’t too noticeable. However, the 5950 Ultra seems to do a much worse job than both the X800 Pro and 6800 GT.

[micamica1217]

Yes, mica it was essentially a copy/paste, once again to prove the point that both nV and ATi is guily of cheating. Ill see what else I can find.

I will repeat my self, nVidia is not cheating in this test due to DTS...this is a problem that futurmark needs to anser to if you ask me.


AND THE TRUTH SHALL SET YOU FREEEE!!!!!! :D

Finally an unbiased anwser. :attn:

no, what I said if you read my reply was that tranCendenZ's examples were wrong and misleading...and lacked the rest of the info that I posted.
cheating? I never said that ATI or nVidia was ever cheating in 3dmark05...what are you talking about?
where are you trying to take this thread?

this thread is all about FSAA (the rest is minor)...and if nVidia even does FSAA in the 6800gt cards.

as for your other reply with reviews....
if I'm asking users to reply with thier findings because "WEB SITES LIE AND MISLEAD US" all the time on this subject...why should I even listen to some sites such as bjorn.com when the AA tests you showed was done at 2xAA for ATI, and 4xAA for nVidia.

now I know that not only some of my replys are not biased, but I know I'm speaking for my self, ALL MY REPLYS ARE NON BIASED.
they are filled with info that can be backed up, and replicated...how is that biased?

and that is my quote, btw...so what truth will set me free?

dude, again, were are you going with this?

mica

[Dragonprince]

micamica1217 can you do a capture on the same frame that I am posting? I set all of the sliders to max details in the CP, turned off all optimizations manualy, and enabled vsync. I can not set details any higher on the GT than they are. In particular I am interested in seeing what the X800 looks like when set to max quality while looking at the two areas that I boxed in black on the image. It looks to me like the GT is applying AA only in certain areas, as you can see by looking at the image.

http://img17.exs.cx/img17/128/Image739.jpg

[micamica1217]

micamica1217 can you do a capture on the same frame that I am posting? I set all of the sliders to max details in the CP, turned off all optimizations manualy, and enabled vsync. I can not set details any higher on the GT than they are. In particular I am interested in seeing what the X800 looks like when set to max quality while looking at the two areas that I boxed in black on the image. It looks to me like the GT is applying AA only in certain areas, as you can see by looking at the image.


send me a PM with your email and I'll do a test for you.
sending a full size IQ pic as requested.

mica

[micamica1217]

Why does it matter where you set it, CP or not? I thought the game has nothing to do with it, and it's the drivers at work? Or does this apply to AA only? :confused:

I know it's not a UT2004 only problem, but UT2004 is the only game i see it in.

see, this is why the 3dmark05 stuff should have been posted in the other forum, as you now seem to be confused about the issues.

first, sentanals reply is not valid in both the FSAA topic of this thread, nor has he showed a complete and truthfull comparison with all the facts.
that's why I corrected or added the rest of the info.

really, if your going to post IQ pics on FSAA, maybe you should use the same amount of AA.

now as for your question...
there is a differance, when using the CP instead of the GUI in 3dmark05, in the way AF is done.
when using the CP, more then what is asked from the benchmark is filtered.
this happens with both cards.
and seems to be normal.
(the big problem comes when testing for AA with nVidia...you must use the GUI for AA to work. like in farcry. IIRC)

mica

[JeffP]

really, if your going to post IQ pics on FSAA, maybe you should use the same amount of AA.
mica

Out of all the links posted I only see one thats not a direct comparision. Is there something else I'm missing?

You sure do seem to be a hardcore ATI Defender. I know you say non-biased but everytime something negative is metioned about ATI your quick to jump on the bandwagon. If your non-biased you should be camping on both sides. Just my observation.

[Sentential]

Out of all the links posted I only see one thats not a direct comparision. Is there something else I'm missing?

You sure do seem to be a hardcore ATI Defender. I know you say non-biased but everytime something negative is metioned about ATI your quick to jump on the bandwagon. If your non-biased you should be camping on both sides. Just my observation.
Thankyou.... the voice of reason finally arrives. :thup:

[CandymanCan]

I was about to say this is kinda getting outa hand now, i mean Mica, really you dont have a 6800 so why are you so worried and eager to bash the card ? Personally when i first read thsi thread it ticked me off that i got the 6800gt but now that i relise the diff between the IQ on my ATI and NV is so minimal it doesnt bother me really. The only thing im mad about is that the VP doesnt work on my card.

Other then that i think you should lay off and consentrate youre efforts on ATI's "Optimised" drivers for 3dmark05, ya you know the ones that increased the scores by 1500-1800 points on the same day the bench was released

[micamica1217]

Out of all the links posted I only see one thats not a direct comparision. Is there something else I'm missing?

You sure do seem to be a hardcore ATI Defender. I know you say non-biased but everytime something negative is metioned about ATI your quick to jump on the bandwagon. If your non-biased you should be camping on both sides. Just my observation.

as I've said before.....

if your going to post IQ pics with FSAA in mind, they both need to be the same.
posting a pic with 2xAA with an ATI card, and 4xAA with a nVidia card, and not telling anyone (such as the pics from bjorn3d...sent's examples of return to proxy) and saying that nVidia does AA better then ATI (so to speak), is not what I would call fair or informative.
I call it deseption, and misleading.

I could go on and on, but yes, your missing alot of points I made as well.
no need for me to retell the story.

mica

[CandymanCan]

Well all my pics were using the same settings for both cards.

[micamica1217]

I was about to say this is kinda getting outa hand now, i mean Mica, really you dont have a 6800 so why are you so worried and eager to bash the card ? Personally when i first read thsi thread it ticked me off that i got the 6800gt but now that i relise the diff between the IQ on my ATI and NV is so minimal it doesnt bother me really. The only thing im mad about is that the VP doesnt work on my card.

Other then that i think you should lay off and consentrate youre efforts on ATI's "Optimised" drivers for 3dmark05, ya you know the ones that increased the scores by 1500-1800 points on the same day the bench was released

that investigation is still being done as we speak.

sofar the statements from ATI are true (memory fix), and there is nothing really wrong at this time from them.

I get an increase in almost all games too. (including Doom3)
I've been using the 8.07's now for quite a few days.

if you find something, let me know.

mica

[CandymanCan]

Well i have a rather old ATI card (9700pro) but ill let ya know if i find anything newo n both sides like a fix for this AA problem. I know chowdy has been hard as work trying diff tweaking programs and such lol.

Btw mica youre Avatar is that youre WC setup if it is, did you use some of that Yellow hi-liter stuff ? If you did can u tell me did you get any gunk in youre pump cuss i see in the pic thier is an Eheim pump i was thinking about trying the HI-LITER idea but i was worried about my pump clogging

[micamica1217]

Btw mica youre Avatar is that youre WC setup if it is, did you use some of that Yellow hi-liter stuff ? If you did can u tell me did you get any gunk in youre pump cuss i see in the pic thier is an Eheim pump i was thinking about trying the HI-LITER idea but i was worried about my pump clogging

I am using a mix that was blended from SSS.
this watercooled setup has been running with out flaw, since OCT 2002.

what you are seeing in my avitar is my original P4 rig...1.6a@2.4ghz, 512mb rambus, abit mobo, and GF3 vid card.

Well all my pics were using the same settings for both cards.

I know, and it was a fare comparison that clearly showed the problem with FSAA in the 6800 cards.

mica

[CandymanCan]

where can i get that mix linky ? cuss it looks awsome, How much did you have to use or get and how many of the Uv cathods ?

[micamica1217]

where can i get that mix linky ? cuss it looks awsome, How much did you have to use or get and how many of the Uv cathods ?

I got about 3 or 4 bottles of pre mixed military grade distilled water from SSS.
I have no idea what he used to fully mix in it...so no linky.

at the time I had only one CC UV light....I now have two.
pic was taken with low light, no flash.

mica

[Dragonprince]

You sure do seem to be a hardcore ATI Defender. I know you say non-biased but everytime something negative is metioned about ATI your quick to jump on the bandwagon. If your non-biased you should be camping on both sides. Just my observation.

I dont think personal comments about someone are needed. micamica1217 has raised a valid question around nVidia using proper full screen AA. I paid the same 400 bucks most others did to by a GT and don't appreciate nVidia selling me something that is not doing what they claim. I was a long time Ti fan, then a fan of the R300 cards, so I think im fairly open minded on using whatever hardware is the best at the time.

The facts seem to be that nVidia's IQ issues have not gone away and they are using a "speed is king" mentality to sell cards. Honestly the whole situation reminds me of the Radeon vs Ti arguments from a couple of years back. Most folks admitted ATi's IQ was visably better than nVidia's but they were willing to sacrific some IQ for the extra speed because games didnt play well, or at all, when you went the max details route.

Today things are slightly different since all of the high end cards are fast and most of the FPS tests really only come into play while benching cards. Under normal game play both the 6800 and X800 cards are wickedly fast running todays games so now IQ can, and should, take the focal point in discussions. Stating that "My card does 60 fps running Doom3000 or Quake2999" is no longer enough because more and more people will ask "At what detail settings?"

The point is IQ matters, and anyone thats ok with getting lousy IQ from any company selling 400 dollar cards is a fool to buy a high end card, why not just buy a cheap 9800 and run without any IQ turned on???? :-/

[micamica1217]

Dragonprince,

the rest of the pics were sent in two emails.
I'll be gone till latter tonight.

enjoy, for whatever they are worth.

mica

[Dragonprince]

Thanks for the pic's mica. clearly when comparing images from your X800 and my GT there are some substantial differences between the two. Your image is showing AA being applied in areas that my GT leaves untouched. Not good, im going to run some tests when I get a chance and see if any of nVidia's other drivers give better results.

[CandymanCan]

The game that has the worst AA probs iv seen so far is Bf1942, UT2K4 or anything else doesnt even compare. I almost get sick playing the game

[micamica1217]

Thanks for the pic's mica. clearly when comparing images from your X800 and my GT there are some substantial differences between the two. Your image is showing AA being applied in areas that my GT leaves untouched. Not good, im going to run some tests when I get a chance and see if any of nVidia's other drivers give better results.

my pics that I sent you had no AA in them.
if your looking at the shadows like in your pic, then yes, because the ATI cards are not running with DST then the shadows are sometimes better.

there are always some shadows that will have little jaggies in them, since 3dmark05 is not perfect and even ATI cards will show that too.
it's just softer on the ATI cards, but still not perfect.

again feel free to post up my pic of the #1697 high quality pic, as it's close to yours and peeps here want to see the differance with my card.
(again, that pic has no AA in it, it's just with the texture sliders turned up to max and vsync turned on, like yours.)

btw, did you try 61.77's as they seem to have nice IQ in most games/benchmarks???

mica

[Chowdy]

I'd like to confirm, if it hasn't been already, that it happens in CS:S. 4xAA, 8XAF, there's some aliasing on wires and the ground. On the ground at certain angles, and on wires at a distance.

:|

[CandymanCan]

Bf1942, is really bad i wish icould make a video and post it on here, because when you walk looking at the ground ahead of you its really sickning. This seems to only happen on 1942 for me tho cuss every other game its fine, i only have the AA probs on certain objects in other games like minor things like wires and stuff.

[Dragonprince]

my pics that I sent you had no AA in them.
if your looking at the shadows like in your pic, then yes, because the ATI cards are not running with DST then the shadows are sometimes better.

there are always some shadows that will have little jaggies in them, since 3dmark05 is not perfect and even ATI cards will show that too.
it's just softer on the ATI cards, but still not perfect.

again feel free to post up my pic of the #1697 high quality pic, as it's close to yours and peeps here want to see the differance with my card.
(again, that pic has no AA in it, it's just with the texture sliders turned up to max and vsync turned on, like yours.)

btw, did you try 61.77's as they seem to have nice IQ in most games/benchmarks???

mica


I will run the 61.77 in testing against my first pic. Can you supply an image at max details with AA and AF maxed out also ?

[micamica1217]

I will run the 61.77 in testing against my first pic. Can you supply an image at max details with AA and AF maxed out also ?

pic sent....btw, you do know I'm using the 8.07 beta drivers?

mica

[Chowdy]

Although it's off topic, the 8.07 "magical" drivers should be looked at closely, because i've been reading around and even users with 128MB cards posted increases.

Aside from that, i'd like to note that changing from quality to high quality has little effect on FPS and makes the moire effect go away on repetitive textures.

[micamica1217]

Although it's off topic, the 8.07 "magical" drivers should be looked at closely....

I think Dragonprince is doing just that with all the pics I sent him...as well as others.

mica

[Chowdy]

I think Dragonprince is doing just that with all the pics I sent him...as well as others.

mica

Good to hear, but i'm wondering if the rest of the community is. I hope nvidia releases a magical driver soon too. :beer:

[micamica1217]

Good to hear, but i'm wondering if the rest of the community is. I hope nvidia releases a magical driver soon too. :beer:

why? you think the rest of the community is (only) investigating nVidia's drivers?

and with all the cheating that nVidia has done over the past 2 years, you think nVidia hasn't put out "magical" drivers?

could you tell me why, when ATI is able to put out a new WHQL driver each month (as well as having futuremark update it's approved drivers each month), that nVidia can't?

while this is way off topic to this thread, I'm wondering what nVidia users are thinking here in this forum.

mica

[johan851]

It sounds a bit harsh, but I'd have to agree. People in the past have bought, for example, nVidia cards that perform worse than their ATi equivalents with the EXPECTATION of nVidia releasing some super driver that will fix all of the issues. This has been going on for quite some time...and I haven't seen a super driver.

[CandymanCan]

I got my 6800gt because of recomendations, iv been outa the video card race for a long time mainly stay with cpu's and mobo's. Now since i listend to people i got the card that i shouldnt have. Oh well i cant take back my $400, im happy with my 6800gt it plays all my games cept a few (old games) fine with the exseption of some games having bad AA. Other then that my card mods into a 6800Ultra overclocks great and is faster then any x800pro or Xt.


Btw guys ATI cards have the same exsact AA problem as the NV counterpart it just doesnt show as much. Looking at my pics from my 9700pro to 6800gt, the 9700 has all the jaggies in the same area as the 6800gt on poles wires and straight lines o nthe roads ect, it just isnt as bad. Nv 4xAA looks like Ati's 2xAA, NV 8xs on the other hand blows ATI'S 6xAA away

[JeffP]

while this is way off topic to this thread, I'm wondering what nVidia users are thinking here in this forum.

mica

LOL well since you asked.

I think I made the better buying decision getting my 6800 over the X800. Not to say I might not still get a X800XTPE for my other rig just to have both on hand to play with.

I could careless as could 99% of the population if the drivers are WHQL approved and I'd count on Nvidia as already knowing this. As long as they play the games when I want them to it's fine by me.

Only people hung up on the approved driver thing are ATI fanboys looking for something to complain about. IMHO.

Jeff

[PhobMX]

Only people hung up on the approved driver thing are ATI fanboys looking for something to complain about. IMHO.

Jeff

A better way to say it is people that like iq above raw speed, or get the goodies a $400 card MUST be worth. so quit your flameish (sp?) generalizations...

[micamica1217]

I could careless as could 99% of the population if the drivers are WHQL approved and I'd count on Nvidia as already knowing this. As long as they play the games when I want them to it's fine by me.

Only people hung up on the approved driver thing are ATI fanboys looking for something to complain about. IMHO.

Jeff

realy, noone is complaining here about drivers.
yet, not every nVidia owner thinks like you as well on this issue.

but you realy didn't anser my question....
why do you think that nVidia "can't put out drivers that are WHQL certifide each month?"
better yet, "why can't they put out drivers that pass the "approval" of futuremark?"
(it's not like futuremark is realy picky on the drivers)


seeing as how "some people" would like drivers that don't cheat the consumer out of IQ in some way....
I would think that WHQL would be at least a step in the right direction.

btw, WHQL 61.82's are out today. (edit: for quadros anyway.)

I'm all for using beta drivers, but when I last had a nVidia card, they put out offical drivers all the time.
now it's like once every 3 to 4 months.
why do you think that is?

mica

[JeffP]

but you realy didn't anser my question....
why do you think that nVidia "can't put out drivers that are WHQL certifide each month?"
better yet, "why can't they put out drivers that pass the "approval" of futuremark?"
(it's not like futuremark is realy picky on the drivers)

Good question. The same thing is being asked of ATI with Linux. Why can't they produce a decent set of drivers that perform? I think the answer is they both could accomplish the task at hand they just choose not to for whatever reason. As to the real reason I don't think we'll ever know, only speculate.


seeing as how "some people" would like drivers that don't cheat the consumer out of IQ in some way....
I would think that WHQL would be at least a step in the right direction.

btw, WHQL 61.82's are out today. yet not on nVidia's site yet.

I'm all for using beta drivers, but when I last had a nVidia card, they put out offical drivers all the time.
now it's like once every 3 to 4 months.
why do you think that is?

The 61.82 drivers i'm hearing are junk (at least for the newer cards). They don't take advantage of any of the new features in the 6800 series (i think it was ps3.0 and something else, didn't pay to much attention sorry). I would assume thats why they are still the 61.x drivers and not the 66.x drivers. If it was 66.82 as a WHQL i'd be willing to give them a shot.

I could take a million guesses at why drivers are lagging from Nvidia but it would be just that, a guess.

[micamica1217]

Good question. The same thing is being asked of ATI with Linux. Why can't they produce a decent set of drivers that perform? I think the answer is they both could accomplish the task at hand they just choose not to for whatever reason. As to the real reason I don't think we'll ever know, only speculate.


The 61.82 drivers i'm hearing are junk (at least for the newer cards). They don't take advantage of any of the new features in the 6800 series (i think it was ps3.0 and something else, didn't pay to much attention sorry).

they are quadro drivers, and have full suport of the features...
400 marks less then a beta driver is not that bad.
just my 2 cents.

I would assume thats why they are still the 61.x drivers and not the 66.x drivers. If it was 66.82 as a WHQL i'd be willing to give them a shot.

I could take a million guesses at why drivers are lagging from Nvidia but it would be just that, a guess.

I realy hear you bro....and I understand that it would only be a guess.

mica

[Paladine]

LOL well since you asked.

I think I made the better buying decision getting my 6800 over the X800. Not to say I might not still get a X800XTPE for my other rig just to have both on hand to play with.

I could careless as could 99% of the population if the drivers are WHQL approved and I'd count on Nvidia as already knowing this. As long as they play the games when I want them to it's fine by me.

Only people hung up on the approved driver thing are ATI fanboys looking for something to complain about. IMHO.

Jeff


I sure hope that I come to the same conclusion when I get my 6800GT, and I am pretty sure I will. (As long as I can play my games that I have now, and those new ones I am getting ready for, I should be happy. - Been out of the video card race for a while now!)

Why? Well even playing certain games with my stand-in Ti4200 I am having issues with certain drivers. 66.70 & 66.72 WHQL drivers to be exact. They cause strange problems ( a moire like effect) om Jedi Academy, and in Battle Field Vietnam I get extreme gittering. Occurs even more with 66.82 drivers as well. The only drivers that seem to work fine for me are 66.71. So does that mean I will have the same or worst issues with my 6800GT?? Sure hope not!

:bang head

[CandymanCan]

if you mean shimmering you'll have the same issue

[Sentential]

Well I hope this is enough reason for nV to increase their IQ.... ATi has a *BIG* problem on their hands now :eek:

New SLI benches from nV
http://www.hardocp.com/images/news/1097603215cFVsh2AA0F_1_1_l.gif

....wowza thats a hellova boost :eek:

[JeffP]

Well I hope this is enough reason for nV to increase their IQ.... ATi has a *BIG* problem on their hands now :eek:

New SLI benches from nV
http://www.hardocp.com/images/news/1097603215cFVsh2AA0F_1_1_l.gif

....wowza thats a hellova boost :eek:

Holly smolly. Look at those 3dmark05 scores in SLI mode. I assume all those are at stock speeds with no mods? Thats sick. Of course trying to watercool 2 GPU's could be a challenge but I'd be up for it :)

[Hellion]

bah, challenging? give me the money the parts and the girls...wait...no girls? fine. anyways, i could do it. but it would be hard as hell to get some decent tubing in there, considering how close the sli cards sit to each other, i can see were your coming from. can you say kinks? eh, it would be hard because anything that would get rid of the kinks. would probably lean the card, or put pressure on it.

[micamica1217]

Well I hope this is enough reason for nV to increase their IQ.... ATi has a *BIG* problem on their hands now :eek:

New SLI benches from nV
http://www.hardocp.com/images/news/1097603215cFVsh2AA0F_1_1_l.gif

....wowza thats a hellova boost :eek:

I agree, what a nice boost.

too bad we wont see dual PCI-E mobos for almost 6 months.
by then, both nVidia and ATI will be about to put out new cards.

mica

[Sentential]

I agree, what a nice boost.

too bad we wont see dual PCI-E mobos for almost 6 months.
by then, both nVidia and ATI will be about to put out new cards.

mica
Its the concept that matters.... all of their PCI-E cards will have SLI capability from now on. Not only that but from what i understand they plan to expand it. SLI can technically handle up to 16 GPUs I belive

[micamica1217]

Its the concept that matters....

really?
vaporware doesn't make my games any faster.
but at least I'm getting 11,560 in 3dmark05 with my x900pro...j/k bro. :D


the future is just that, the future.
I like the concept, but I'll wait till nVidia puts out a dual (16x) PCI-E mobo and some new vid cards first.
again, by the time this happens, we'll have new cards.

mica

[Dragonprince]

I wouldnt mind an SLi rig, except a couple of things keep nagging at me. :-/

First is obviously cost, lets face it, 2 PCI-E cards at 350-400 each = rediculous. Who is gonna front that kind of cash for cards from companies that cant make IQ good on one 400 dollar card let alone two.

Second and more importantly is do you really need 300-500 fps in a game? If Doom3 is the big boy for OpenGL and FarCry is the big boy for DX9 and current cards are running those games fine whats the incentive for SLi ? It's not like this is 1995 all over again and you actually need a VooDoo rig to play at decent framerates.

Call me unimpressed by over the top, over priced, and over hyped SLi rigs....no matter who makes it, im not buying.

Now dual chip CPU's have my interest, those might actually be worth the cash :D

[PhobMX]

I wouldnt mind an SLi rig, except a couple of things keep nagging at me. :-/

First is obviously cost, lets face it, 2 PCI-E cards at 350-400 each = rediculous. Who is gonna front that kind of cash for cards from companies that cant make IQ good on one 400 dollar card let alone two.

Second and more importantly is do you really need 300-500 fps in a game? If Doom3 is the big boy for OpenGL and FarCry is the big boy for DX9 and current cards are running those games fine whats the incentive for SLi ? It's not like this is 1995 all over again and you actually need a VooDoo rig to play at decent framerates.

Call me unimpressed by over the top, over priced, and over hyped SLi rigs....no matter who makes it, im not buying.

Now dual chip CPU's have my interest, those might actually be worth the cash :D

damn i also join the sli-hatters club, mostly for the price... heck even two 6600s would be the same price or little less as a single x800pro, and 2 gts or even ultras would be twice the price of the rest of the system. imho the concept is too pricey

[Chowdy]

The concept is actually less pricey, if you think about it. Buy a SLI-capable 6800U now, use it and enjoy. Maybe almost a year comes and u want to upgrade, 6800U's come down in price, slap another in, and enjoy a 70%+ boost!

[PhobMX]

The concept is actually less pricey, if you think about it. Buy a SLI-capable 6800U now, use it and enjoy. Maybe almost a year comes and u want to upgrade, 6800U's come down in price, slap another in, and enjoy a 70%+ boost!

right, but when your sli setup ages enough, youll end up with a couple of old and slow once $700 video cards (a gt now and another gt in 1.5 years) in 3 years only worth $200 at best.

I going to stay with single video card and upgrade every 1.5 or two years, that way you wont have that much devaluated hardware and still play on top notch cards EFECTIVELY USING the "NEW features" some fanboys like to talk about (but never trully use).

[CandymanCan]

I had my 9700pro for 3 years and it still runs todays game just fine, i plan to keep this 6800gt for 3 years also

[Hellion]

hmm....i wonder if you can do ULTRA/GT SLI?

[Sentential]

hmm....i wonder if you can do ULTRA/GT SLI?
No, it must be the same vender, same chipset (ie same BIOS is my guess)

[PhobMX]

I had my 9700pro for 3 years and it still runs todays game just fine, i plan to keep this 6800gt for 3 years also

i dont agree, maybe its your cpu helping you out but my system wont cut for 1024x768 and descent detail ;(

[Hellion]

No, it must be the same vender, same chipset (ie same BIOS is my guess)


k, than what if its BOTH pny, and one is an ultra, and the other is a gt FLASHED to ultra?

[Sentential]

k, than what if its BOTH pny, and one is an ultra, and the other is a gt FLASHED to ultra?
Im almost positive it would work. I honestally belive the same vender BS nV is putting out is mainly because of BIOS constrictions. Its almost like Dual-channel RAM in that sense.

So long as the BIOS are exactally the same (voltage, ram timings, speed, ect) I see absolutly not reason for it not to run in SLI properly.

[Hellion]

Im almost positive it would work. I honestally belive the same vender BS nV is putting out is mainly because of BIOS constrictions. Its almost like Dual-channel RAM in that sense.

So long as the BIOS are exactally the same (voltage, ram timings, speed, ect) I see absolutly not reason for it not to run in SLI properly.

SWEET. lol
hmmm.....i wonder, if the ultra would perform better than the gt, at same clock speeds? cause of it being a different gpu? no its the same right? just crippled? im asking, cause i might go that setup for my comp soon, and i bought a gt now. and if its worth it, ill get an ultra later, IF it performs better at same speeds. otherwise ill just get another GT and brning it up to ultra speeds.

[Paladine]

I was curious if anyone has found this problem with the New Omega Drivers that have been released (based on the 61.77 drivers)??? :confused:

Hmmm....
Well I will report back on my results with this tonight as I finally got my video card :attn: .....my system is now COMPLETE (well it will be when I get off work tonight). :thumb: My Gainward [/b] 6800GT 256MB Golden Sample video card came in. And BytewiseComputers sent me my case as requested. YAHOO! :thumb:

Let you know how it works after I install it and play some games. :clap:

[micamica1217]

The concept is actually less pricey, if you think about it. Buy a SLI-capable 6800U now, use it and enjoy. Maybe almost a year comes and u want to upgrade, 6800U's come down in price, slap another in, and enjoy a 70%+ boost!

realy? less pricy for who? you?

this is what you'll need and the costs...take a gander.

new mobo:
even if you have a PCI-E mobo (and you don't), you'll now need a dual PCI-E mobo at a price of $150-$200
you also have a 4-6 month wait for the dual PCI-E mobos to come out.

new CPU:
the dual PCI-E mobos won't come in all form factors, so most of us will need to by a new CPU as well.
cost, about $200 or more.

DDR2 memory:
many people, including myself will not by a new mobo without DDR2 support.
we might not get that choice with AMD at this time, but intel seems to want DDR2.
if you need DDR2, then add $200-$400 to the cost.

new PSU:
many of us are just getting by with our PSUs at this time.
when you think of adding a seconed vid card to our systems, then think about a better PSU too.
cost, $50-$100 (think closer to the $100 mark, if you don't feel like sending back a cheap PSU your hoping will work on a dual system.)

new vid card:
WHAT? you don't have a PCI-E card now?
then make that two new vid cards.
cost, at about $300-$400 each, that's $600-$800

you do the math.

yes, the concept is nice, and I may go for it sometime down the road.
but by that time, we will have new cards, and more choices to look at.

mica

[kouch]

i know what you mean mica. I mean by that time we will have dual core gpus maybe and that may be better than sli?

Also no one ever mentioned features. Yeah 2x 6800's will be nice but when are you gonna really need them? Like in 2 yrs and by then dx10 will be out with a newer feature set and then you probably won't want to invest in a dx9 card. So SLI is nice concept but it has failed in the past (remember vodoo)

[Hellion]

lol, dude. mica, you should work for ati. that speech could stop ANYONE from doing it. lol. very true.

[JeffP]

I think SLI will come around. It just takes time. It took how long for PCI to become popular over ISA/EISA. When AGP first broke it took a while for people to convert as well. I think this will be the same way. A slow migration as it is a huge advancement in technology for PC's.

As with anything else in the electronics world you pay the price to be a early adopter. If you don't have to be on the very bleeding edge you can wait a while and the prices will surely fall.

As for the concept it's well over due. We've been doing parallel processing and scaling horizontally for how long now because it's more cost effective?

[micamica1217]

lol, dude. mica, you should work for ati. that speech could stop ANYONE from doing it. lol. very true.

the last thing I want is to work for ATI, but the truth is the truth.

maybe in 1 or 2 years things will change, and SLI will be more then realistic.
it might be practical for most of us then.

but right now PCI-E has to take off first, before we look into dual PCI-E.
that means AMD is going to need to make things worth it to upgrade....
since intel isn't even pushing me into an upgrade.

so why should I ditch my nice AGP based system?
no reason I can think of yet.
the mobos aren't even out yet...lol

mica

[a c i d.f l y]

This all sounds like the same debate when ATI and nVidia released their first second generation 3D graphics cards, and everyone who dropped $400-500 on their Voodoo2 SLI setups were getting pissed off when everyone started saying their speed sucked because it didn't look as good. However, in that case, the details were BLATANTLY obvious that 3dfx cards blurred the every living **** out of the image to create the speed. In the case of the current ATI vs. nVidia... if we're pointing out small little squares lacking detail that will all go away in the next frame.. you can all eat my ass.

This is an over-tweaked conversation that is better left alone. Much as the whole ATI vs. nVidia conversation is. Meaning -- If you're going to be that damn picky... I don't know. Quit the cock fighting.. its retarded.

I'm an nVidia supporter... and I can be satified with a faster card, cheaping me out in such a small portion it is hardly noticeable without it being directly pointed out in a screenshot. What's ironic though, I own a 9800Pro.. but only because it cost $120. If the X800 were faster in all respects then the 6800, and looked better.. well. Then we wouldn't have something to fight over. 6800 is faster, X800 looks better. You pick which you would prefer -- oh, isn't the X800 more expensive?

-Frank

[micamica1217]

This all sounds like the same debate when ATI and nVidia released their first second generation 3D graphics cards, and everyone who dropped $400-500 on their Voodoo2 SLI setups were getting pissed off when everyone started saying their speed sucked because it didn't look as good. However, in that case, the details were BLATANTLY obvious that 3dfx cards blurred the every living **** out of the image to create the speed. In the case of the current ATI vs. nVidia... if we're pointing out small little squares lacking detail that will all go away in the next frame.. you can all eat my ass.

This is an over-tweaked conversation that is better left alone. Much as the whole ATI vs. nVidia conversation is. Meaning -- If you're going to be that damn picky... I don't know. Quit the cock fighting.. its retarded.

I'm an nVidia supporter... and I can be satified with a faster card, cheaping me out in such a small portion it is hardly noticeable without it being directly pointed out in a screenshot. What's ironic though, I own a 9800Pro.. but only because it cost $120. If the X800 were faster in all respects then the 6800, and looked better.. well. Then we wouldn't have something to fight over. 6800 is faster, X800 looks better. You pick which you would prefer -- oh, isn't the X800 more expensive?

-Frank

ya know, I basicly agree with you.

only, we are talking about a feature that is promessed (FSAA), and once again nVidia seems not to be delivering.

I won't go into beta drivers that lower IQ in order to get more frames...
since if you want to use them, then that's your choice.
but should you want to use FSAA, then with the 6800's you should get it.

mica

[Hellion]

sigh, by far the most wisdom pointed toward the subject of IQ ati vs nvidia, its not a MY DAD CAN BEAT UP YOUR DAD, its. WHO THE **** cares! ill kick you *** myself. So instead of saying, this is better, your saying. who cares? there both better than these past cards. thats what matters.

[Dragonprince]

I think SLI will come around. It just takes time. It took how long for PCI to become popular over ISA/EISA. When AGP first broke it took a while for people to convert as well. I think this will be the same way. A slow migration as it is a huge advancement in technology for PC's.

As with anything else in the electronics world you pay the price to be a early adopter. If you don't have to be on the very bleeding edge you can wait a while and the prices will surely fall.

As for the concept it's well over due. We've been doing parallel processing and scaling horizontally for how long now because it's more cost effective?

Sounds good but your forgetting the fact that single slot solutions wont be at a stand still in developement. 2 years down the road we are likely to have single chip single slot GPU's that will stomp on the current 6800's running SLi, and that includes cards from nVidia. If the GPU makers jump on the CPU dual core bandwagon then all bets are off...we could have quad core SLi rigs.. :drool:

[micamica1217]

Sounds good but your forgetting the fact that single slot solutions wont be at a stand still in developement. 2 years down the road we are likely to have single chip single slot GPU's that will stomp on the current 6800's running SLi, and that includes cards from nVidia. If the GPU makers jump on the CPU dual core bandwagon then all bets are off...we could have quad core SLi rigs.. :drool:

with GPUs getting smaller each day...most are 120m-90m now.
heat will be a major issue if we goto dual chip cards....can you say SLK900 heatsinks for stock cooling. lol
I just don't see this happening any time soon.

mica

[bobmanfoo]

i don't think building an SLI rig is going to be that much more expensive if you plan on building a new rig now or in the near future, an sli and non-sli would cost about the same minus the extra video card. either way you go you would have to upgrade every component of your pc anyway, cpu, mobo, ram, video card, and probably hdd, except maybe ur psu. i don't plan on jumping straight into buying 2 video cards, but i definitely plan on having a capable mobo.

[@/2ct!<]

If I could just relate to original topic... how noticeable is AA @ 1280x1024, it would seem to me that it would be less important than @ 1024x768, but my current card isnt fast enough so I cant tell what the differences might be.

[PhobMX]

If I could just relate to original topic... how noticeable is AA @ 1280x1024, it would seem to me that it would be less important than @ 1024x768, but my current card isnt fast enough so I cant tell what the differences might be.

some peeps like 4xaa at 1280x1024, but myself im perfectly good with 2xaa at that resolution. 1280x1024 without aa is almost unbeareable for me, 2x is good and enough but 4x is perfect

[JeffP]

Not sure if someone can test this in the real world that has the games mentioned but.....Playing around with benchmarking today I threw the Nivida settings over on High Quality and ran 3dmark05 (this is with the 66.81's). Strangely enough the shimmering was gone, the textures were much improved and overall IQ was much more improved (frame rates dropped about 3-4fps). The part that did amaze me is why is shimmering gone with the slide of a rule from performance to quality?

While it did pull the end score down a bit I'd say the real world trade-off was worth it. Maybe this wasn't as noticable before as it is now with the OC i'm running. It would be nice to see if ayone else can support this finding.

[micamica1217]

Not sure if someone can test this in the real world that has the games mentioned but.....Playing around with benchmarking today I threw the Nivida settings over on High Quality and ran 3dmark05 (this is with the 66.81's). Strangely enough the shimmering was gone, the textures were much improved and overall IQ was much more improved (frame rates dropped about 3-4fps). The part that did amaze me is why is shimmering gone with the slide of a rule from performance to quality?

While it did pull the end score down a bit I'd say the real world trade-off was worth it. Maybe this wasn't as noticable before as it is now with the OC i'm running. It would be nice to see if ayone else can support this finding.

yes, in real world testing, playing on high quality is always better then the default opts as on....
but each game's framerate is differant between the 2 settings.

some like farcry, is only a small loss in framerate...
some like BF:V, it can be as much as a 50% loss in frames.

as for the shimmering in many games such as UT2004, BF:1942, and painkiller, it seems that they are still shimmering at high quality.

to use this video, right click, then play with any player you want, such as windows media player.

http://www.abu-el-mot.de/rs/bf.avi

that's BF:1942 whith high quality settings.

mica

[JeffP]

mica what I found strange was that it was so obvious now where as it wasn't before. Last time I tried these settings was with the older drivers and I was only running at 400Mhz or so.

Now with the newer drivers and a 455Mhz speed the difference is well night and day. I'm a little stumped as to if it's the drivers or is it the additional GPU speed that allows the better IQ at useable rates.

I know this doesn't answer the question at hand but rather raises another.

The shimmering problem has been a issue for a lot of Nvidia owners in 3d05 especially on the serpent textures. Under high quality I get none under performance i get a ton regardless of GPU speed, unfortantly without a major OC the HQ mode is about worthless.

[Dragonprince]

I dont know why you'd be stumped JeffP, obviously the nVidia drivers crank the details way down when set to performance. Thats what you would expect right ? All sorts of graphic anomolies are possible when your intentionaly not rendering a scene as the developer intended. Its possible that nVidia might even be locking out some features of the core when set to performance on the CP sliders. That is pure speculation and admittedly I have no knowledge that they are in fact doing that, its just possible. This applies to both nVidia and ATi; we all know ATi has suffered thier share of issues when they have gone for all out speed in the past.

As far as an overclock causing IQ to improve, im sceptical. I can understand an OC allowing you to run higher IQ settings due to more available "horsepower" but I dont see that, in and of itself, how the OC could create better IQ. In fact the opposite should be the case more often than not due to the usual OC'er pushing to the limit and possibly creating artifacts after extended gaming or benching.

[JeffP]

As far as an overclock causing IQ to improve, im sceptical. I can understand an OC allowing you to run higher IQ settings due to more available "horsepower" but I dont see that, in and of itself, how the OC could create better IQ. In fact the opposite should be the case more often than not due to the usual OC'er pushing to the limit and possibly creating artifacts after extended gaming or benching.

I didn't mean to say it improved IQ if thats how you interpited it (but the overall scenes do look much better then previous runs). I'm thinking it's the fact that with the extra GPU speed i'm able to run at the high quality setting and still have aceptable frame rates. Where as before FPS dropped enough that it became a little choppy. With that lower fps and when things are less fluid maybe you notice the anomolies more then you would otherwise? Or the 66.81's really improved IQ on the High quality setting. Like I said before I'm not sure.

I can say for certain last time I set it to high quality and ran 3dmark05 with older drivers (I want to say 61 or 65.x series, could be wrong) and had lower GPU speed the difference was not so drastic as what I see now between performance and HQ. My question is why?

[micamica1217]

I didn't mean to say it improved IQ if thats how you interpited it (but the overall scenes do look much better then previous runs). I'm thinking it's the fact that with the extra GPU speed i'm able to run at the high quality setting and still have aceptable frame rates. Where as before FPS dropped enough that it became a little choppy. With that lower fps and when things are less fluid maybe you notice the anomolies more then you would otherwise? Or the 66.81's really improved IQ on the High quality setting. Like I said before I'm not sure.

I can say for certain last time I set it to high quality and ran 3dmark05 with older drivers (I want to say 61 or 65.x series, could be wrong) and had lower GPU speed the difference was not so drastic as what I see now between performance and HQ. My question is why?

there are a few things going on that I'll try to touch on...

from driver to driver, yes, 3dmarks IQ could be better or worse...mostly at the cost of performance (even if it is tiny).
but the thing is, nVidia for quite some time has worked on a per game/benchmark mode of optimising it's drivers.

what I'm trying to get at is that just because a benchmark might have slightly better IQ, doesn't mean that in all games your IQ will now improve.
some might get better, or some might stink even more then the last driver you used.
as a whole, what we should hope for is some new official WHQL drivers that are in the intrest of the consumer...
slightly less performance, but vastly improves the IQ.
this way, we the consumer could lower the IQ ourselfs when we so chose to get the added performance we want.

for all we know, your right though...maybe the high quality settings got better. (but as a whole as far as the rest of our games of benchmarks? only you can let me know.)

mica

[Matt McFaul]

I too have noticed all of these problems however being an online player I am used to sacrificing looks for higher fps.

I do though have a really odd problem that I would like to sneak in to see if anyone else is having these problems.

First, I have had 2 bfg 6800gt oc's....one revision 1 and one revision 2. Both worked flawlessly at stock speeds with the exception of the known image quality issues. I tested both cards using Joint ops, CS:source, commanche 4, doom 3 , far cry, cod, and cod day of defeat. Never had a single glitch there.
Once I flash the bios to any other bios I can no longer play commanche 4 or joint ops for longer then a few seconds without the screen going nuts with tiny bars of color which ultimately leads to a freeze or auto restart.
On the other hand I can play any other current title flawlessly and even while substantially overclocked to 440/1200. Heat is definately not an issue since my current version 2 is cooled by the danger den nv68 block. The versiojn was cooled by a moded mcw50-t with dedicated meanwell 320 for the peltier. Even at 15c I experience this problem.
One little bit of info I forgot to add: no artifacts are present at all and then wham instant fubar screen....restart and run 3dmark2005 in loops for a few hours, try a few other games, then go back to JO and it happens again.
I reflash to my backup bios and it works fine all the way around though.
system specs: P4 3.2c@3.84ghz (runs fine at 4ghz though) prommie cooled
abit ic7max3
1gb hyperX pc4000 @ 240mhz 1:1
dual antec true control 550's (1for cooling, 1 for system)
bfg 6800gt oc (lots of flashes and oc's) water cooled
dual raptors raid zero
audigy 2zs platinum
I have also tried stock system with flashed card to no avail. tried the following as well: fastwrites,agp aperture, various in game settings, various driver optimizations, various drivers..etc. In the past an overly aggressive oc led to minor artifacting as you work your way up in clock speed however for those 2 games from the same manufacturer (hmm) it would seem to be an all or nothing battle. If JO wasnt my favorite game i'de be content and just toss it out the window but hey...it was my main reason for getting the card in the first place.

SORRY FOR THE BOOK!-and thanks if you read it~ :bang head

[techbinge.com]

bobmanfoo and to everyone else,

after looking at your CoD screenshots w/ some pretty bad jaggies (bad for being at 4x AA) I decided to run a quick test myselt. I'm not sure why the jaggies are less aparent in my tests but the images speak for themselves. I think one other thing to note as well is that the jaggies do look WORSE in .jpg form - they are better, slightly, but better in engine.

Settings:
4x forced AA, max quality, 1600x1200.
BFG 6800 GT.
66.81 Drivers.

I also wanted to note that full screen AA didn't at 4x and 8x greatly effect my FPS. No shimmering whatsoever was seen in game.

Nvidia Image Quality Comparison
(http://www.techbinge.com/cod/index.htm)

-Tyler

PS. I will run some more tests w/ other games and post them here.

[12am]

welp... I too have this poor aa issue, sure wish I woulda read this thread before I bought this 6800GT :( don't get me wrong it's a very nice card, however the lack of proper aa is really an eyesore especially coming from a 9800pro, heck I might have to beg my g/f to trade me her 9800pro for this 6800Gt.

[CandymanCan]

Use the 67.02 or 67.03 drivers it fixes it.

[12am]

Using the 67.03 drivers as we speak, and indeed the problem is not fixed.

[CandymanCan]

Oh well

[micamica1217]

welp... I too have this poor aa issue, sure wish I woulda read this thread before I bought this 6800GT :( don't get me wrong it's a very nice card, however the lack of proper aa is really an eyesore especially coming from a 9800pro, heck I might have to beg my g/f to trade me her 9800pro for this 6800Gt.

do you mind posting the "kizoar wings" pics from both the 9800pro and 6800gt here?

listing the resolutions and AA settings for each pic

mica

[12am]

Not at all, afterall it's for a good cause :)

This is a 1600x1200 screenie taken from saga of ryzom on the 6800GT when 4xAA 8xAF are applied in the control panel (note the kizoar's wings and body also the player charachter)
http://img.photobucket.com/albums/v621/cruze8/6800gt1.jpg

Now here is a 1600x1200 screenie taken from saga of ryzom on the 9800Pro when 2xAA 8xAF are applied in the control panel (again look at the kizoar's wings and body, also the player charachter)
http://img.photobucket.com/albums/v621/cruze8/9800pro1.jpg

[JeffP]

Doesn't appear AA is working at all in that game with the 6800. Not sure if it's a driver issue or a programing issue in the game. Defiantly not normal for the 6800.

[CandymanCan]

I have a similerp roblem with Ut2004, sometiems my AA works and sometimes 4xAA looks like no AA, i found that in order to fix it you need to change the resolution to 800x600 i nthe game, then change it back to whatever you want. I dunno why it does this but it only happens in ut2k4 which IMO is a buggy game.

[ryanmartini]

that first screenshot is from city of heroes, but in the options, there are no settins for AA....

[comfortablynumb]

12am's screens are the worst I've seen, even my 5900XT does not do that badly.
I'm truly surprised, as I've read all reviews/threads about the 6800's I could possibly find,
and never really got the sense that AA was so poor. I hope Nvidia will fix this soon, as I will buy a 6800 series card within the next week....

CN

[CandymanCan]

Yea they are pretty bad

[micamica1217]

12am's screens are the worst I've seen, even my 5900XT does not do that badly.
I'm truly surprised, as I've read all reviews/threads about the 6800's I could possibly find,
and never really got the sense that AA was so poor. I hope Nvidia will fix this soon, as I will buy a 6800 series card within the next week....

CN

please understand, that not every game is as bad as that.

as I've said in the past on this, it seems that some games are far more inclined to show this lack of FSAA then others.
in some games it's so hard to notice, that it's almost not a problem at all.

yet what I feel is a far bigger problem is the lack of qualifide reviewers from review sites.
using just this thread as an example, IQ pics were taken from bjourn3d and NVnews....
yet the examples had flaws in it, and were not truly investigated.
in the case of bjourn3d, 2xAA was used for the ATI card, and 4xAA was used for the nVidia card.
that's not just a slip up, that's a clear mistake that should have been caught way before it was posted.
I don't even think that today, the pic was changed, and replased with a 4xAA one for ATI.
can you tell me how I can trust a review from them now?

and they are not alone.
I've spoke in the past about xbit labs, and how they clearly say that there farcry results show same IQ pics for both cards...
they even posted pics from both cards.
yet when looking at the nVidia outside pics of the mountins, you'll notice missing light sources...clearly an oversite.
yet how can I trust that persons eyes now on IQ?
I could go on about what was said about HL2 too, but I think you get my point about xbit labs.

almost each site has made at least one mistake in benchmark results or IQ comparisons for thier reviews...even Anandtech.
it's a sad state that user reviews are far more important today then ever when wanting to know what is really going on.

mica

[12am]

ryzom is not alone, I have some hl2 screenies too that show nv's lack of proper antialiasing almost as badly. Want to know the funny thing? saga of ryzom is a "The way it's meant to be played" title.... so I guess this is how nv wishes us to see the game ;).

[micamica1217]

ryzom is not alone, I have some hl2 screenies too that show nv's lack of proper antialiasing almost as badly. Want to know the funny thing? saga of ryzom is a "The way it's meant to be played" title.... so I guess this is how nv wishes us to see the game ;).

I've tested the 6800gt and my x800pro in HL2 for more hours then I'd like to admit.

clearly the trees and wires, as well as a few other areas, don't seem to be helped by using AA on both cards.
yet look at the image with both AA at 4x and with it turned off...
you'll see that it is working (just not that well).

yet I could have missed something with all my testing...so please explain what your seeing.

mica

[CandymanCan]

Blah blah blah ok guys now look at the diff between the older drivers that dont have the LOD clamp and the newer ones that do look much better the newer drivers look.


Nice the diff between the 2 pics the bottom one looks so much better.. All the straight lines like in the tunnles and on the walls which everyone says NV doesnt AA, well once the LOD clamp is on its fully AA the jaggies are almost gone.

Make sure you expand the pic when you click on it otherwise itll look like crap. To expand the picture first click on it then hold the mouse over the pic for a second and 2 things will pop up clock on the X on the bottom to expand the picture to full size

The first pic is from the 61.82's @ 4xAA 4xAF @ 1280x960.

http://img.photobucket.com/albums/v466/Candymancan/shot00005.jpg

Second pic is from the 67.02 drivers same AA/AF and reso but with the lod clamp on.

http://img.photobucket.com/albums/v466/Candymancan/Shot00001.jpg

[comfortablynumb]

Blah blah blah ok guys now look at the diff between the older drivers that dont have the LOD clamp and the newer ones that do look much better the newer drivers look.


Nice the diff between the 2 pics the bottom one looks so much better.. All the straight lines like in the tunnles and on the walls which everyone says NV doesnt AA, well once the LOD clamp is on its fully AA the jaggies are almost gone.



Ok...MUCH better.

Can this be demonstrated from any other games?
Maybe 3dmark05?

It gives me hope, as I really like to play UT2004 natively in 64bit Linux, major reason for my upcoming 6800NU purchase.

Thanks

CN

[CandymanCan]

3dmark05, i dont have installed but i can do some other games

[Sentential]

Another reason in the 6800's defence. HDR rendering in Farcry <this is also with LOD clamp>:

http://img.photobucket.com/albums/v180/Sentential/FarCry0000.jpg
http://img.photobucket.com/albums/v180/Sentential/FarCry0005.jpg
http://img.photobucket.com/albums/v180/Sentential/FarCry0006.jpg
http://img.photobucket.com/albums/v180/Sentential/FarCry0004.jpg
http://img.photobucket.com/albums/v180/Sentential/FarCry0003.jpg

HDRrendering in farcry is only capable when using an nVidia card. Ill take a couple pix without HDR

[CandymanCan]

Those pics look so sweet, id like to see an X800 do that

[Hellion]

watch it candyman, or well be defending ourselves against ati fans, :P
but i agress....COMPLETELY stunning....kinda makes you wanna count each frame, and make it into a slide show screen saver :P

[CandymanCan]

Oh im sorry i just cant stop, and be Awd, now that my lock up problems are solved "DAMN SP2" i can sit and watch the awsome graphics from the 67.02's and above with the LOD clamp.

[micamica1217]

Blah blah blah ok guys now look at the diff between the older drivers that dont have the LOD clamp and the newer ones that do look much better the newer drivers look.


Nice the diff between the 2 pics the bottom one looks so much better.. All the straight lines like in the tunnles and on the walls which everyone says NV doesnt AA, well once the LOD clamp is on its fully AA the jaggies are almost gone.

Make sure you expand the pic when you click on it otherwise itll look like crap. To expand the picture first click on it then hold the mouse over the pic for a second and 2 things will pop up clock on the X on the bottom to expand the picture to full size

The first pic is from the 61.82's @ 4xAA 4xAF @ 1280x960.

Second pic is from the 67.02 drivers same AA/AF and reso but with the lod clamp on.


is there any reason that the pics are not at the same resolution...the 2nd pic is saved at a higher res (800x600 vs 960x720)????

better yet, why not at full size like you did in the past?

mica

[comfortablynumb]

is there any reason that the pics are not at the same resolution...the 2nd pic is saved at a higher res (800x600 vs 960x720)????

better yet, why not at full size like you did in the past?

mica

I would like to see them full size as well.

And the FarCry images from Sent. are absolutely gorgeous.

CN

[CandymanCan]

Photo bucket limits me.

Ill post some more

[Sentential]

watch it candyman, or well be defending ourselves against ati fans, :P
but i agress....COMPLETELY stunning....kinda makes you wanna count each frame, and make it into a slide show screen saver :P
Zomg you have no idea....if only i could post a WMV of it...its more stunning in person than it is with individual pix.

It is absolutly amzing. Ive NEVER had to squint when firing a gun in any game, but i have to do it with HDR enabled....take a look:

http://img.photobucket.com/albums/v180/Sentential/FarCry0002.jpg

This is completely playable at 50+ FPS <anyone know how to turn a FPS counter on in F-cry>

[micamica1217]

Photo bucket limits me.

Ill post some more

I understand this.

that's why I just don't understand why you bothered with the farcry shots...
the game pics came from a resolution that was originaly 1024x768 with NO AA at all.
in other words, the pics were totaly unnessasary and unrelated to this thread.

mica

edit: oooops the farcry pics were from sent

[micamica1217]

This is completely playable at 50+ FPS <anyone know how to turn a FPS counter on in F-cry>

I forgot too, but you could use FRAPS, and make an AVI of your game too.

mica

[CandymanCan]

Yea t he pics werent full size because Photobucket rezises them when the image uses to much space in kb or whatever. Im using freeservers right now so the picsm ight go away saying remote linking forbiddon

I took some more shots. As you can see the first one is being partial AA'd like what is being discused you can see some things without jagges and the one that d have jaggies are the lines like the Poles and Wires and stuf. The so called 4xAA almost looks like no AA or 2xAA

http://residentevil.mods.freeservers.com/images/shot00006_no_lod.jpg

The second picture its all gone. You still dont belive i could careless, go use youre 6800GT and find out foryousefl im telling the truth. This looks like it should

http://residentevil.mods.freeservers.com/images/lod_clamp.jpg


The second pic is a littlem ore blurry due to the clamping of the LOD but the jaggies are like almost gone.

[micamica1217]

The second picture its all gone. You still dont belive i could careless, go use youre 6800GT and find out foryousefl im telling the truth. This looks like it should


The second pic is a littlem ore blurry due to the clamping of the LOD but the jaggies are like almost gone.

bro, chill out, noone here is calling you a lier.
and thanks for showing that using the "clamp" helps sometimes.

I tryed clamp too with the 67.02 drivers, and I didn't notice this fixed in UT2004...
yet I did notice that it blurs the textures alot like in your pic.

above all, thanks for your help in this matter. :attn:

mica

[CandymanCan]

Maybe its just the 61.82 driver making it look bad, i use that because before last night i couldnt only use those drivers and other driver and my pc kept rebooting, after countless weeks trying to figure it out sen said to get rid of SP2 in XP and now i can use other drivers. All i have right now is the 61.82 and the 67.02. I thought you were implying i was a lier cuss you started to question why my pics were diff sizes. Anyway does ur Ut2k4 look as bad as the pic ?? Cuss mine always looked like that it was discusting now that i can actually try diff drivers with out the rebooting problems the LOD ones look awsome.

Try using these 67.02 drivers.., hey improove the IQ alot .

http://www.valancestudios.com/index.php

[Sentential]

I forgot too, but you could use FRAPS, and make an AVI of your game too.

mica
Already done. If you wanna see em I can send them to you. Sadly they are like 100MB a piece, so i cant host em :(

<directed at mica>
Bro in order to use HDR you have to disable AA. Hence the pix

[Sentential]

Couple more:

http://img.photobucket.com/albums/v180/Sentential/FarCry0009.jpg
http://img.photobucket.com/albums/v180/Sentential/FarCry0004.jpg
http://img.photobucket.com/albums/v180/Sentential/FarCry0007.jpg

[micamica1217]

<directed at mica>
Bro in order to use HDR you have to disable AA. Hence the pix

I know and understand this, that's why it doesn't pertain to this thread.

mica