nvidia 45.23

[ThugsRook]

benchmarks scores are way down!
quality IQ is way up!

looks like the driver cheats have been at least partially removed.
i can notice a huge quality improvement w/o any effort at all.

its about time NV :na:

[pointreyes]

Yahoo! I have been needing more IQ. :p ;)

Seriously, I'm glad to hear that the IQ has improved. I will have to give these drivers a try now. Thanks for the info. :)

[Mikki]

Thugs, are these comparable to the quality of the 30.82 drivers? :)

[ThugsRook]

Mikki ~ better quality then 30.82 :chit:

[ThugsRook]

* 3.45g 191fsb 382ddr c2622-1
* nv gf4 ti4600
* image = quality
* d3d = quality
* no AA / AF


http://www.bleedinedge.com/images/thugs/ico-3dm2k.gif 3DMark2000 32bit
43.03 = 18234
45.23 = 17238


http://www.bleedinedge.com/images/thugs/ico-3dm2k1.gif 3DMark2001
43.03 = 16453
45.23 = 15214


http://www.bleedinedge.com/images/thugs/ico-q3a.gif Quake III Extreme 1152x864
43.03 = 141.7 fps
45.23 = 146.2 fps


http://www.bleedinedge.com/images/thugs/ico-ut2k3.gif UT2003 Demo 1024x768 fly / bot
43.03 = 199.6 / 92.2 fps
45.23 = 189.5 / 92.2 fps


http://www.bleedinedge.com/images/thugs/ico-xisle.gif X-Isle Tech Demo 1024x768
43.03 = 142.0 fps
45.23 = 147.1 fps



dispite some of the benchmarks scores i will be using the 45.23 drivers. IQ over the 43.03 will be obvious once seen.

the interesting part?
that x-isle and q3x liked these drivers better then the older "cheat" drivers.

:wave:

[Bobbyeyes]

Has anyone tried therse w/ the new intel chipset drivers?
intel link (http://downloadfinder.intel.com/scripts-df/Detail_Desc.asp?agr=N&ProductID=816&DwnldID=6413)
my scores have dropped (on the p4p800 and asus 9480vc)
even more, but the VQ is nicer

ps.. hope no one was flashing there bios in the northeast yesterday evening @4:12pm :rolleyes:

[ThugsRook]

II 502? i tried them ~ were very bad for me

[Bobbyeyes]

Abysmal for me
(but i love the digital vibrance setting-just one ore two knotches on the mouse wheel is all you need)

[ThugsRook]

yea aint that nice? i have mine set for 0.5 on that scale

(its in the color correction tab)

[DAPUNISHER]

Hey Thugs, How do I do the QIII xtreme tests? I ran QIII benchmark with demo1 and custom settings all on max E.G. trilinear, 32bit, max details, ect. and did a 3 run pass but my score looks way too high to be right :confused: BTW, My monitor doesn't support above 1024x768 and won't load the bench at resolutions above that.

[ThugsRook]

if you change your details (model and texture quality) it reverts Q3 Extreme back to the standard configuration and you have to re-impliment Q3X.

nothing special to run the benchmark ~ same as always....

~
/timedemo 1 <enter>
demo four <enter>

:-)

[DAPUNISHER]

Been so long since I played with that game! My disc is so old school I had to install the latest point release to get demo 4 :o I was also using the QIII bench utility before so I wasn't getting the Xtreme settings, yep, I'm a dumbass! :wave:

I ran demo 4 with xtreme settings using quality with 4xAA/4xAF and averaged 173FPS@1024x768

I ran some other benchies but I didn't bother switching drivers in and out *too lazy* so I just used scores I still had for other drivers when I could find one for comparison's sake.

3Dmock2001
44.03's = 18043
44.71's = *N/A
45.23's = 16433

3Dmock '03 build 330
44.03's = 5187
44.71's = 6070
45.23's = 6016

UT2k3
44.03's Flyby = 232.5 Bot Match = 78.7
44.71's *N/A
45.23's Flyby = 230.4 Bot Match = 78.3
*indicates benchmarks not run

Looks like whatever they used to boost the '03 performance with the FX cards after build 330 is still there, but they evidently don't give a damned about 2001 anymore judging by the hit my card showed there.

BTW, those extra shadows in QIII are trippy sometimes :wack:

[ThugsRook]

3Dmock
:lol: :rotflmao: :lol:

thx for the benchies on your fx5900 :wave:

[ThugsRook]

interesting cheat?

i recently discovered that by enabling "texture sharpening" (for AA) makes nonAA benchmarks go faster :wack:

Aquamark3 in perticular ~ cant anyone else verify this?

[Mikki]

:scratch: :?: :)

[Schneider]

I thought texture sharpening reduces scores coz it turns on 2xAF :scratch:

[ThugsRook]

obviously not ;)

http://home.comcast.net/~d3rook/graphics/ico-am3.gif Aquamark3 ~ AA off / default benchmark
TS off = 43.85 fps
TS on = 45.61 fps

bug or cheat? you decide ;)

[chuckbam]

I am using the BETA 51.75. When will the offical 50 something DET be out?

[ThugsRook]

i donno ~ but the 52.xx betas are out.

[chuckbam]

Originally posted by ThugsRook
i donno ~ but the 52.xx betas are out.

I have been looking? Have a link

[ThugsRook]

3dchipset.com (http://www.3dchipset.com)

[chuckbam]

Originally posted by ThugsRook
3dchipset.com (http://www.3dchipset.com)

Thanks ThugsRook :cool:

[ThugsRook]

Originally posted by Schneider
I thought texture sharpening reduces scores coz it turns on 2xAF :scratch:
you may be right.
by default AM3 calls for 4x AF ~ but the texture sharpening might be forcing 2x AF? :scratch:

[ThugsRook]

Originally posted by Schneider
I thought texture sharpening reduces scores coz it turns on 2xAF :scratch:
tested and confirmed!

enabling "texture sharpening" also enables 2xAF.
it also overrides AM3 benchmark's call for 4xAF and makes AM3 run with 2xAF insted ~ a cheat.

[Mikki]

Nice find guys...;):wave:

[Schneider]

Originally posted by ThugsRook
tested and confirmed!

enabling "texture sharpening" also enables 2xAF.
it also overrides AM3 benchmark's call for 4xAF and makes AM3 run with 2xAF insted ~ a cheat.

Well I don't think u could consider it a cheat because its always been part of nvidia's drivers since the Det 40's i think, plus its not like everyone uses texture sharpening
How'd u confirm wat i said, just by comparing scores??

[ThugsRook]

bug, cheat, doesnt really matter what ya call it these days ;)

yea, comparing TS on vs 2xAF vs 4xAF vs nothing on.
TS should not override a 4xAF application call tho, but it does.

:)