driver round up!

[legacy-metalhead74]

drivers i tested!

cats:
4.3
3.7
3.6

vs.

omegas
4.3
3.7

(i have pics but their pretty big and i'm to lazy to resize, but if want them then i was give them to you)

4.3 cats:
flyby: 151.645676
botmatch:75.402084

4.3 omegas:
flyby:151.824631
botmatch:76.555122

3.7 cats:
flyby: 150.410919
botmatch: 76.559151

3.7 omega's
flyby: 150.494476
botmatch: 77.191780

3.6 cats:
flyby: 150.528610
botmatch: 77.021759
(didn't test the 3.6 omega's, something about not compatable with my card, so didn't take the chance of screwing my system)

summary: well as you can see the 4.3 cats and omegas had the highest scores is flyby, and the 4.3 omegas did beat the 4.3 cats in only about 1 fps. the 3.7 cats scored low in the flyby but they did beat out both 4.3's i the botmatch by .3 fps. the 3.6 cats scored the highest in the botmatch section, so i praise them for that! the omega 3.7 cats did ok in the flyby, but bot match is what counts! the 3.7 omegas do the best ingame in ONS, in ONS i gained 5+ fps!

(note: all were done at 1024x768 resolution. my card is not overclocked and AA and AF were both turned off.)

have any questions post!

[legacy-EvilXenu]

Interesting info. Thanks for taking the time to post your results.

--evilxenu

[legacy-metalhead74]

update! i will test the 4.4's today, and hopfuly their omega counterparts will be out soon, and i went and played a game of ONS in the new map ascendancy and i can say the fps are very solid, on the omega 3.7's, i encourage everybody to try these!

[legacy-isidis]

well i am using the 4.3's so i want to know if it is worth going to the 4.4's

/me waits for your new tests

[legacy-metalhead74]

well test later tonight, and i know htat you guys are waiting! sorry for the wait

[Folk]

Not trying to rain on your parade dude, but there is less than 100th of a percent difference in the highest and lowest scores, so what's the point? Anything within 2% is within the margin of error for the benchmark.

[legacy-metalhead74]

dude, the ones that get good scores, tend to have a more solid fps, the 4.3 cats were to jumpy in ons, but the the 3.7 omega's are alot more stable! they gave me a 5+ fps bouns!

[legacy-metalhead74]

cat4.4
flyby: 151.721542
botmatch: 76.399117

and played a game of ons in asendancy, and their very jumpy! i turn the corner and boom down to 20fps and have to wait to they load up to 60fps and then shoot the tank boom back down to 20fps and so on, the 3.7 omegas didn't do that and their are not big jumps.

can't wait till the omegas come out!

[Folk]

Originally posted by metalhead74
dude, the ones that get good scores, tend to have a more solid fps, the 4.3 cats were to jumpy in ons, but the the 3.7 omega's are alot more stable! they gave me a 5+ fps bouns!

And where exactly in those scores you posted is that 5+ fps difference?

[legacy-metalhead74]

k, dude you notice the benchmark is a flyby in antalus and asbestos, and the are just DM, in that you get 70-80fps in dm.

but i went and play some ONS and notcie that i got a 5 fps diference in ONS and my fps stay higher throw out the game.

[legacy-metalhead74]

4.4 omegas coming soon!

[legacy-metalhead74]

4.4 omegas
flyby: 151.870102
botmatch: 76.141266

boot up a game of ONs and assualt and boy was i amazed these by far are the best drivers i have ever used, better the the omega 3.7's. in ONS now i get 70-80FPS and in big fights it only goes dow to 50FPS at the most.

my fps has improved by 15+ fps in every game type, and in DM is in the 80-90FPS constant, these are very good! go download them now!!!!!!!!!!

[legacy-isidis]

so overall omega 4.4's are best

[legacy-JaFO(JBE)]

Originally posted by Folk
Not trying to rain on your parade dude, but there is less than 100th of a percent difference in the highest and lowest scores, so what's the point? Anything within 2% is within the margin of error for the benchmark.

I'd say anything within 5-10% margin isn't worth worrying about, because you won't notice it anyway.
Plus these are averages. Without the min/max they're absolutely useless (you can drown in a river with an average depth of 1 centimeter ... ).

Plus ... bugfixes are more important than a standard benchmark in a default level.
Trying a user-made level with plenty of detail and lots of fighting will give you a better idea of the card/driver.

// metalhead74
Chech your driver-settings ... with a +/- 85 average you probably have vSync enabled in the drivers.

// --
If you need a benchmark to 'see' the improvements/problems then they're not much of an improvement/problem ...

[legacy-st_trooper]

That's nice for people using a 9600 **AND** **if** they only play UT2k4 and don't care about bugs. Two things you failed to take into consideration.

[therogue]

If you're implying the omega's are buggy, I've been running them all week, and they're fine for me

[legacy-^ngelus]

where to download? can't find them...

[Folk]

Originally posted by JaFO(JBE)
If you need a benchmark to 'see' the improvements/problems then they're not much of an improvement/problem ...

I would agree with that up to a point, but with the large number of tweakable variables in UT2004, you need to benchmark in order to see the small improvements. Let's say that you have 10 variables (2K4 actually has many more) and that each variable can either give or take away 1 FPS. As you pointed out, you won't "see" the improvement of a 1 FPS difference, but you would certainly see a 10 FPS difference. But without a benchmark to fine tune, you would not enjoy the aggregate of those small improvements and would just leave those 10 FPS laying there because you couldn't see or feel the individual improvements. Or even worse, if operating simply by feel you might positively change 5 variables, negatively change the other 5 and then walk away convinced that you can't improve performance because "nothing changed".

[legacy-^ngelus]

link?

[legacy-JaFO(JBE)]

// Folk
At 20 fps average you would notice 10 fps more.
Altough that would be a 50% increase in average fps.

At steady 80 fps 'average' (I get over 100 fps in the easier maps) .... I'm not even worrying about a mere +/- 10 fps ( < 10%), because I would not see them without the aid of a benchmark.

Hell ... I'd only worry if the framerate should consistently dip below the minimum fps, which for me is somewhere between 15 - 25 fps.

Eyecandy > fps pretty much anytime for me.

// ^ngelus
For the official drivers ? http://www.ati.com
Omega : http://www.omegacorner.com/

[legacy--Loric-]

Originally posted by JaFO(JBE)

At steady 80 fps 'average' (I get over 100 fps in the easier maps)

Many monitors have a max FPS around 80 Hz (eg 85 Hz @ 1280x1024 and 75 hZ @ 1600x1200). Trying to get higher FPSes is useless since the monitor will not allow to display them anyway.
And thus there is no visual difference between any number which is above the max FPS the monitor may output.

/ That said, my Sapphire 9800 XT gets me a nice 500+ FPS in UT2004... in the menus

[legacy-bman]

Juding from these scores, it looks like you're CPU limited anyway. Everybody is in botmatch. Why not run them with AA and AF on. I can't see what your card is, but with a 9700 or 9800, you'd never run in 1024 with no AA or AF.

[legacy-Buffy Summers]

I tested the 4.4 cat drivers, but they gave me a slight mouselag so I immediately discarded them. Using Omega for 4.3 now.

Btw, a better way to benchmark is to:
1. Start a practice game against 8 godlike bots on a semi-open or open map (I use deck17).
2. Type 'god', 'loaded' 'allammo' in the console
3. Type 'demorec bench' in the console
4. Frag away as fast as you can, 20 frags or so, on various places in the map.
5. Stop the recording with 'stopdemo'
6. Benchmark by typing 'stat fps' and 'demoplay bench?timedemo'
7. Watch the FPS and remember the lowest.
8. Get the average from the console once the demo has finished playing.

This way of benchmarking is much better than the flyby/bot benchmark, and will give results that are more like real playing.

Notice that the bots don't take any CPU time for their AI when replaying the demo, but instead take the same amout of CPU as a nomal human teammate or opponent does.

Another thing: don't use to high resolution even if the scores are the same. For example, I get almost the same framerate with 1600x1200 as in 1024x768, but the mouse lags in 1600x1200. And I hate mouselag worse than the plague.

[Folk]

Originally posted by JaFO(JBE)
Eyecandy > fps pretty much anytime for me.

Well there you go. There are two classes of FPS players... those who want the eyecandy and those who are willing to sacrifice visuals to get the best possible performance. I guess we know to which camp each of us belongs.

[legacy-UberBone]

hmmmm.... no difference to speak of then ...

Try running 'em all again and they'd swap and change places I bet ... so little dif ...

[legacy-metalhead74]

ok guys, yes the 4.4 omegas are the best on my system, and what are you talking about vsync off? in the ut2004 menus or in my ati control panel? cause it's turned off in my ati control panel.

[legacy-Master-Yodaa]

I get 101fps in the botmatch bench with my 3.9 drivers.
And 98fps with the 4.4 Omegas.
The 4.4s also bring a p|ss poor opengl gamma bug, what a waste of space new drivers are, they fix problems with crap games, and at the same time introduce new bugs for the good games, 3 f*cking driver revisions and still no fix for something that USED to work :sour: :down: