UT2K3 DM-Icetomb performance feedback needed ASAP!

[saulin]

sucky IQ eh?

Well I'm happy with the IQ and so is fivefeet8 and many FX owners and yet our cards beat the crap out the Radeon 9800 Pro in current games. Wanna talk about tomorrow games?

well for tomorrow games there are new cards in the making already since I doubt that the 9800 pro will cut it. Is like the GF4. It kicked ass but then with newer games it just could not keep up with the highest settings.

I remember Serious Sam 2 would kill my GF4. Now that game is not so tough anymore.

So you still have your GF1 eh?

Well good go and play some Ut 2K3 with it. I'm sure with everything in lowest you can still kick as$ at the game.

If you think IQ is sooo much important than performance then I guess you are not a true gamer. Specially when the difference in quality even without trilinear filtering is that freaking small.

Oh BTW Behemoth I'm sure my card can beat your mighty 9800 Pro on just about any map in UT 2K3 using the same settings. It is a fact that UT 2K3 runs faster on the FX. so don't even go there. Of course if you wanna bring the new maps the dets 50s probably would still kick the 9800 pro butt. But I rather wait for the official release before switching Dets.

[saturnotaku]

Originally posted by saulin
If you think IQ is sooo much important than performance then I guess you are not a true gamer.

That's some 1997 thinking if I ever saw it. If this was everyone's stance, we'd all be playing games at 640x480, 16-bit color. FSAA would be a pipe dream and anisotropic filtering wouldn't be on anyone's radar. Hardware and game development would be stymied and far behind the curve that it's at today.

Wtih today's technology and what's on the horizon for next year, there is no reason why gamers can't have their cake and eat it, too.

[saulin]

saturnotaku you must be kidding.

As fast as todays video cards are. they are not fast enough for a lot of todays games.

Sure these cards can do 12x and 16X FSAA. But really not many games would be playable with those settings.


You could say heck I can run game "x" at 12X FSAA and 8X AF and get playable fps 20+.

Well good luck having fun with that game. Anything that lags is not playable for me.

It is nice to have nice IQ and speed. But having the best IQ possible and the speed is still a dream.

[mrsabidji]

Originally posted by saulin
Oh BTW Behemoth I'm sure my card can beat your mighty 9800 Pro on just about any map in UT 2K3 using the same settings. It is a fact that UT 2K3 runs faster on the FX. so don't even go there. Of course if you wanna bring the new maps the dets 50s probably would still kick the 9800 pro butt. But I rather wait for the official release before switching Dets.

Bah... you really have to convince yourself your card is beating everyone else's to enjoy it ? See, I have a 5900 too, and I know it actually gets smashed by a 9800 in Dx9 apps. So what ? It's still a good card anyway. It's not a contest, you know... If you're a real gamer, then you know that what really matters is play games at decent framerates, not necessarilly the highest. 5900s will run HL2 at about 40fps (or a bit more with the new 5x.xx) whereas 9800s will run it at about 65fps in the same conditions. That won't make the whole game unplayable so you should stop trying to fool yourself with those BS (no offense intended, but your 5900 beating Behemoth's 9800 looks like BS, unless you omitted something). Well, if you opted for the nv35 thinking it would be a Radeon killer, like I did, I can understand you're a bit disappointed, but you don't HAVE TO own the fastest card in the world to get a real good experience from a game. I will buy HL2 and enjoy it on my not-the-fastest-in-the-universe card, and so you should do. enough said.

mrsabidji

[Behemoth]

yes this is sucky IQ, in my defination:
http://www.nvnews.net/vbulletin/attachment.php?s=&postid=198620
also, i have helped my friend finish half life story mode on his 486/matrox millenium/voodoo2 machine, everything set to very low if not the lowest, we definately had a good time.
if you dont appreciate trilinear filtering and if you can stand sucky IQ, so be it, i cant care less. nv38 and nv40 are coming out, its unacceptable to have 5800/5900 users to play things with disabled fog, disabled trilinear(for peformance sake) in a just a dx8 game and still ask 5800/5900 to wait the oh-so-fix-everything-magical-working-driver, bloodly hell....

p.s. see you all in a few days

[Hellbinder]

Oh BTW Behemoth I'm sure my card can beat your mighty 9800 Pro on just about any map in UT 2K3 using the same settings. It is a fact that UT 2K3 runs faster on the FX. so don't even go there. Of course if you wanna bring the new maps the dets 50s probably would still kick the 9800 pro butt. But I rather wait for the official release before switching Dets.

I'm Sorry but that is simply not correct.

There are many a review out there that Show Nvidia Generally scores higher on their highly tweaked application detected common benchmark levels... But Tie or LOSE (by a lot) on any of the other levels.

This can be carried through virtually any commonly benchmarked Game. Whats more the cat 3.7's increased FSAA scores on the 9800pro's by a little bit. Which should put them a FPS or Two over the 5900U even on the common benchmarked levels. Leaving it much further in the dust when looking at the rest of the game.

Generally Speaking though.. depending on the game and settings (not counting Dx9 or any Heavy PS games) the 5900U and 9800pro perform about the same. With the main difference being the vastly superior AA on the 9800pro and the sometimes slightly better AF on the 5900U.

[saulin]

See ya dude :)

BTW give the beta drivers a break. I'm sure the fog will be back as for trilinear filtering in UT 2K3 I did not see the difference in textures from the shot the other fellow posted and mine.

I guess what I'll do is get me some old dets like 43.x where trilinear filtering is on and compare image IQ and performance there. The problem is that I think only dets 44.03 and up support the FX. However someone with a GF4 could do that test.

[saulin]

Hellbinder before getting my card I read like 30 reviews and many ATI vs Nvidia talks.

Back then there were no DX9 benchmarks. And in some games the scores are so close like 2-3 fps difference. And in some cases the FX won in high resolutions like 1280x1024 and up by a lot when using FSAA and Aniso. And it was the other way arround in other games.


But in the end almost all reviews would say. well as we can see the FX is the faster card however it is priced at $499.

And You are right the ATI seems to have less performance hit in AA as I did see that in UT 2K3 when I tested the 9800 pro on my system.

However at least on these maps that are the default for the benchmark

dm-antalus dom-suntemple dm-phobos2 dm-inferno ctf-face3 ctf-citadel dm-asbestos

The FX won both with FSAA and AF and without FSAA and AF. I did use Cats 3.7

The Radeon won in 3DMark 2001 by 200 3Dmarks and lost in 3DMark 03 by 400 3Dmarks.

In emulators the ATI card had a few issues. Could be driver ralated or OpenGL related or maybe the graphics plugins used on the emulators don't have that great Radeon support.

In the end the FX was the faster card. The better card to get if money was not a problem.

Now that we have DX9 benchmarks things are different. But the FX is not done yet simply because there are no DX9 games out yet. Plus people will be playing DX8 games for a big while still. However Nvidia does need a new card to beat the Radeon 9800 pro at DX9. That is a fact.

[hovz]

since saulin is so confident his card is faster then a 9800pro at dx8, how about someone with a 9800 pro bench v saulin. of course ud have to use performance antialiasing v nvidias quality because they both perform bi linear. what do u guys say?

[saturnotaku]

Originally posted by saulin
Sure these cards can do 12x and 16X FSAA. But really not many games would be playable with those settings.



Who says the be all end all of image quality is that games should be run with 12x or 16x AA?

4xAA (on an R300 card) looks fantastic and is more than playable at 1280x1024 in UT2003 and every game before it. The 4x on a Radeon is far better than 4x on an FX. I've owned both cards, I know.

[StealthHawk]

Originally posted by fivefeet8
I don't think the Fx cards have sucky Iq at all. THey may not be better than ATi IQ, but you can't say they suck either. All this time you've been trying to prove your points, I've been playing the game and the IQ is definately not sucky. Lack of Trilinear filtering does lose some detail from distant ground texture, but everything else looks fine.

Trilinear filtering has existed since like the Voodoo2! Having worse filtering than a 4 year old card is not "sucky IQ?"

[saulin]

Trilinear filtering has existed since like the Voodoo2! Having worse filtering than a 4 year old card is not "sucky IQ?"

Sorry but you are talking about one game here. Nvidia does use trilinear filtering dude. Just because UT 2K3 does not use it doesn't mean all games don't use it.

And if you call that sucky IQ then the Radeon is not far from being sucky as well since the IQ is almost identical.

BTW hovz read some reviews dude and you'll see what I mean. Newer reviews will include DX9 stuff and say well the Radeon is faster. But really the FX wins at most DX8 benchmarkks otherwise the reviwers would not have said it is the faster card for DX8.

This is getting boring you keep bringing the same old stuff guys. No trilinear, sucky IQ.

I personally like the performance and IQ I get out of my card for today's games. And I'm sure many FX 5900U owners do as well. Of course they are not happy with DX9 performance though.

Just give it up.

Yes the Radeon beats the FX in DX9 and yes the Radeon has better IQ. happy now?

[StealthHawk]

Originally posted by saulin
Sorry but you are talking about one game here. Nvidia does use trilinear filtering dude. Just because UT 2K3 does not use it doesn't mean all games don't use it.

We're talking about UT2003 :rolleyes: Look at the title of this thread. And as stated before, 51.75 doesn't use trilinear in all D3D games.

And if you call that sucky IQ then the Radeon is not far from being sucky as well since the IQ is almost identical.

No. Maybe if you use bilinear with your Radeon it looks identical. Trilinear does not exhibit the same artifacts as bilinear, sorry.

[Hellbinder]

They also need a card to beat the 9800pro in Dx8. Thats a Fact.

You need to look at some of the serious reviews from the major sites put out in the last 3 months. ATi wins 60-70% of all benchmarks put out in the last 90 days easy. That is before we even get to Dx9.

Ill get some links for you later.

[saulin]

mrsabidji I know that UT 2K3 runs faster on the FX. Don't even doubt it. I have tried both cards on my system and it doesn't just beat it. It beats it by a good percent in some maps.

Yes the Radeon gains ground with FSAA but it still gets beat.

I brought up that point cause he said that we should see which card gets better performance using the same AA and AF settings.

And no, I don't have to prove my card is faster than yours to enjoy my games. I just don't think the FX IQ is crap or the whole card itself is garbage like I heard along this discusion.

[saulin]

Dude here is a good review...

Is from July and it includes a bunch of DX8 stuff and don't tell me the FX 5900U does not win most benchmarks. Well not 3DMark but most of the gaming benchmarks. Also note that the FX wins the UT 2K3 benchmarks.


And "look" what the final vederict has to say...


When you couple this with eVGA’s e-GeForce FX 5900 Ultra, a card which supports the exact same hardware features but also boasts a 500/900MHz clock frequency and sells for essentially the same price as the FX5900-VTD256, you can see why many enthusiasts have turned to it already. After all, hardware enthusiasts want cutting-edge hardware.

http://firingsquad.gamers.com/hardware/msi_geforce_fx5900u_vtd256_review/page4.asp

http://www.hardocp.com/article.html?art=NDk2LDM=

http://www.hardocp.com/article.html?art=NDcy

[saturnotaku]

It's time I put up my UT2003 results on my 9500. 1280x1024, 4xAA, 4x aniso. And to make it as even as possible, I have my aniso set to performance in the ATI control panel. Also, my card is clocked to a 350 core/310 (620) RAM.

I rest my case in terms of how this card can compete. It's slower on the maps where raw power is required, but on maps that can take full advantage of FSAA and aniso, well...it speaks for itself.

And if your results, saulin, are with a 5900 Ultra...well, I rest my case.

[hovz]

can someone with a 9800 pro plz post hardcop benchmark utility results with 4x performance af and 4x aa

[saulin]

hovz I would do it from my brother's PC but his AMD CPU is crap. It would not stand a chance.

BTW read the ATI vs Nvidia article I just posted from HardOCP.

In Image quality. Well can't tell the difference aside from FSAA.

Is a good review it even has Doom3 benchmarks.

And I still think FX has the lead in DX8.

[saulin]

saturnotaku those are nice results man which proves ATI FSAA performance is good.

Plus the fact you run the settings in performance mode.

Oh yeah here are my scores without FSAA or aniso.

http://members.shaw.ca/navasa/12x10.html

[Flavius]

As the originator of this thread, and former owner and ACTUAL PLAYER of UT2K3 on the same system with both a 9800Pro and a 5900U, let me chime in here and say that performance and FPS were significantly lower ACROSS THE BOARD with the FX5900U. To hell with that damn flybys and fake benchmarks. In the real world game the 9800Pro simply crushed the FX card. The 9800 Pro, if you have read this whole thread, has gotten as much 4 times faster FPS than an FX card on some of the UT2K3 maps like Icetomb and Rustatorium. I cannot possibly recommend an FX card to ANYONE now, and yes, I am a former Nvidiot! The ATI card with 8XAF and 4XAA at 1024x768 is even and often times MUCH faster than the FX5900U while all the while having much better IQ in UT2K3. The difference in the quality of AA alone is mind blowing.

[saulin]

dude we are talking average fps here... that is what benchmarks do.

there is a low and there is a high.

Is like 3DMark it does whole bunch of tests and then takes the average score.

Yes ATI FSAA performance is nice compared to Nvidia's. As for 6X ATI being sooo much better than Nvidia's well 4XS or 8XS do look pretty sweet. Jaggies difference would probably be only that much noticiable on older games that have terrible graphics.

[Spotch]

I see that there is a big debate going on here about IQ comparisons. Thats nice but I am reminded of a recent accusation that a pretty big cheese in the gaming industry just made a couple weeks ago.


Something like: "NVIDIA Det 50's use screenshot detection to produce captures at higher quality than those rendered during actual gameplay."

I am not sure if anyone here can really verify this "optimization" but it should be addressed in this thread as a possible cause for the much improved framerate with little IQ difference in many of these screenshots.

[saulin]

Spotch if that was true I guess Nvidia would be in big trouble then...

That would be some major cheat. Plus the person playing the game would notice that right away and bitch to Nvidia. I don't think they want to do that really.

Anyways just like Behemoth I'm taking a break from this discussion. It keeps repeating and repeating and well for ATI users Nvidia sux no matter what and Nvidia IQ simply does not meet their standards hahaha.

People that switched to ATI may think they did the best thing but I think in a way it was foolish, switching from a FX 5900U to a Radeon 9800 pro since either they had to spend a few more hundreds and got an extra video card now or they had to sell their FX for cheap and still spend a few hundreds. Plus I don't really see a clear future on DX9 for either card.

Well unless you do what someone said to start playing with lower settings untill the speed is fine. But then again there is no point going for that if you are with ATI and since IQ is all that it is about lol.

If I switch it won't be to a 9800 pro for sure I rather wait and see what the future holds for both ATI and Nvidia and any new company that steps into the ring.

See ya!

[Hellbinder]

Plus the fact you run the settings in performance mode.

:rolleyes: He ran the tests in HQ mode. It says Right on it.

Secondly the 500mhz e-Vga card never worked right from day one. They released a bios for it to lower clock speeds and its not been heard from since.

Its late and I have to crash but Here is some stuff to chew on.

http://firingsquad.gamers.com/hardware/ati_radeon_9800_review/default.asp

Also note that I am Quite sure than Nvidia is Cheating in Splinter Cell by App detecting this specific level (always used for benchmarking). B3D and other places have Nvidia getting pummeled in Splinter cell. I also Question their Nascar 2004 (EA fix is in). I have a my doubts about IL2 as well.. But...

Those results from [H] are known to be marred by the AF filtering issue for the UT results.

Look at these....

http://www.viperlair.com/reviews/video_graph/ati/aiw/9800pro/p2.shtml

http://www.************/_preview.php?articleID=335

Ignoring the 256MB vs 128MB discussion, the performance of this card is absolutely stunning! The card beat the competition in almost every benchmark, and I could feel the difference in speed in several games when switching over. The image quality as expected was excellent.

Look at the numbers for yourself. They show a bunch of actual "games" tested with Fraps. The 9800 smokes the 5900U.

Wile your at it you may want to take a gander at these numbers...

http://www.ukgamer.com/article.php4?id=98&page=1

Here's some interesting numbers...

http://www.digital-daily.com/video/fx5900-vs-radeon9800pro/index04.htm

Look here is another sound thrashing...

http://www.hothardware.com/hh_files/S&V/gigabyte_radeon_9800pro(4).shtml

Oh and you For Sure want to read this one. Where they show a multitude of Dx8 games.

http://babelfish.altavista.com/babelfish/urltrurl?lp=de_en&url=http://www.computerbase.de/article.php?id=237

Guess who gets an ass whipping in most of those. Like War Craft 3. Big time Dx8 game.

See the pattern? Every single site that does not push standard well known Benchmark levels shows the 9800pro winning the majority of tests. There are some other examples i did not post.

It was also shown by the Guy who makes Nvtweak that by forcing APP detection off the Nvidia scores go into the tank. Ati Scores stay pretty much the same. Meaning game Performance is likely to suffer until nvidia gets the "Fix" into their next driver.

Exceptions are some EA games which have been shown to have a propensity for unfair coding practices. Like Device ID detection to change rendering paths which have a negative affect on the output. IL2 is puzzling but it may actually be just faster on the 5900U. I have posed the Question To Terry & the gang about what’s up with that. I’m sure ill get some kind of answer.

The Same patterns seen in some of these non standard test results can be seen in this Very Thread and a couple others here. Where The ATi card just Creams, i mean Creams the FX in some level where they have not had any opportunity to put an app detect in.

Quite often.. Nvidia is winning some of these tests because Fog is removed. This goes all the way back to the Nv30 at UT2003. It continues today. Especially the 50 series drivers which completely remove Fog for speed.