PDA

View Full Version : Cityintro.unr benchmarks here (updated)


Pages : 1 [2] 3

Dazz
28th May 2001, 06:54 AM
1024x768x32 Max settings VS off

Max 202fps
Avg 87fps
Min 56fps

AMD Duron 800@1Ghz (125Mhz DDR FSB)
ASUS A7M266/ AMD 761 Super bypass
128MB Crucial PC2100 Cas 2-2-2 DDR Ram
32MB Creative Geforce2 GTS/Fast writes/Sideband addressing/ 4X AGP
SoundBlaster Live Player 5.1
Windows ME 4.90.3000
10.1GB IBM Deskstar ATA100 7200rpms
DX8a

IEatShrew
29th May 2001, 08:51 PM
After every tweak I could think of


Min 22.62
Max 73.61
Avg 36.86

450mhz Micron
128 MB ram
Diamond Viper 550
Voodoo 2
800x600x16
Highest detail on everything
Windows 98
And latest drivers for everything

Although I still have to try that tweaker mod by Usaar23... :)

SadaraK
29th May 2001, 10:09 PM
Ok from now on will everyone include data on the following in there posts for fps in intro (casue it makes alot of difference and without the detail there is absolutally no point to this thread).

1/ Full system specs
2/ Resolution
3/ world/model detail level
4/ What mode you run in (d3d/open gl/glide)
5/ NoFractalAnim on or off
6/ Dynamic lighting on or off
7/ High detail actors on or off
8/ Volumetric lighting on or off
9/ Hardware audio on or off
10/ 3D surround on or off


I can add about 20 fps to my score by fiddeling with these settings (not even touching the world detail or resolution) so please include this info from now on, it may be a hassle but there isnt anypoint in posting if it isnt going to be usefull to people....

JayX
29th May 2001, 10:13 PM
Ok from now on will everyone include data on the following in there posts for fps in intro (casue it makes alot of difference and without the detail there is absolutally no point to this thread).
1/ Full system specs
2/ Resolution
3/ world/model detail level
4/ What mode you run in (d3d/open gl/glide)
5/ NoFractalAnim on or off
6/ Dynamic lighting on or off
7/ High detail actors on or off
8/ Volumetric lighting on or off
9/ Hardware audio on or off
10/ 3D surround on or off


I can add about 20 fps to my score by fiddeling with these settings (not even touching the world detail or resolution) so please include this info from now on, it may be a hassle but there isnt anypoint in posting if it isnt going to be usefull to people....

This is true,,and some of the numbers stated look a little outrageous..

bootybandit
30th May 2001, 12:15 AM
45 FPS

ASUS P2B
PIII 800, 100 Mhz FSB (not o/c'ed)
256 mb PC100 CAS2
GeForce 256 DDR (Asus)
Win2k

1024x768
D3D
32-bit color
High world/high texture
Dynamic lighting enabled
Hardware audio off

Straight from my UT.ini file...

[D3DDrv.D3DRenderDevice]
Translucency=True
VolumetricLighting=False
ShinySurfaces=True
Coronas=True
HighDetailActors=False
UseMipmapping=True
UseTrilinear=False
UseMultitexture=True
UsePageFlipping=True
UsePalettes=True
UseFullscreen=True
UseGammaCorrection=True
DetailTextures=True
Use3dfx=False
UseTripleBuffering=True
UsePrecache=False
Use32BitTextures=False
DescFlags=1
dwDeviceId=257
dwVendorId=4318
UseVSync=True
Description=NVIDIA GEFORCE DDR
UseAGPTextures=False
UseVideoMemoryVB=False
Use32BitZBuffer=False
UseVertexFog=False

Degroij
30th May 2001, 10:02 PM
bootybandit set it to 16-bit color instead of 32 bit, and I will bet that your FPS will be 60-65 :)

bootybandit
31st May 2001, 10:34 AM
Degroij, 16-bit color increases my FPS about 8 - 10. 32-bit just looks so good I'm always switching back and forth.

Athlon_2o0o
31st May 2001, 03:32 PM
AMD Athlon (what else hehe) 650 MHz
Diamond Viper II using S3 savage 2000
256 MB SDRAM
18 GB harddrive
soundblaster live!
windows millennium edition

running on S3 MeTaL
average 60 fps with s3 texture compression textures from second CD
resolution 1152x1024.
all other features at the highest possible (high textures, high models etc)
running pretty well! but sometimes my computer freezes. what can be the problem?

Hatecrime69
31st May 2001, 04:43 PM
I'll soon have my new benches with my gf2mx that i'm getting tommrow, good bye savage 4! :)

Odie|BuF
2nd Jun 2001, 08:20 PM
:p amd 1.2ghz
raydeon 64 ddr tv out
all high,sound blaster live surround sound
played 2 hrs, system resources at 72%
min 41.56
max 101.07
average 66.28
on monitor

min 40.62
max 98.57
average 62.35
on 52" big screen tv

Scootercuder
2nd Jun 2001, 08:31 PM
Here izz mine

p3 600MHZ
256 RAMMM
Radeon 32MB
All goodies on high
This is with web browser on Battle Com on and have been playing for about 2 hrs.
1024x768x32 min=23.99Max=74.29Av=42.01
800x600x32 Min=25.22Max=84.91Av=43.56

Went and bought a geforce 3 today and got home and found out my cheap ass HP dont have any AGP slots.SUCKED

JayX
2nd Jun 2001, 09:38 PM
Went and bought a geforce 3 today and got home and found out my cheap ass HP dont have any AGP slots.SUCKED


:D HEHEH Should have checked first..hehe:D

Odie|BuF
2nd Jun 2001, 10:05 PM
:confused: So How Important Do Ya All Think The Frames Per Sec Are ? Ping An Lag Play A Big Part In The Art Of Fragging Dont Ya Think?

Extermin8r
2nd Jun 2001, 10:27 PM
ok here is what i got:

ran @:
800x600
16 Bit (menu will not work in 32 bit w/ my TNT2)
High world
High skin
no decals
dynamic lighting on
no enhanced sound crap (spks suck anyway)
d3d

min 53.18 max 60.39 avg 59.75

my comp specs are listed w/ my signature... and i see i am getting whooped up by some of you. i didn't expect to be at the top, but i thought w/ a 1.3 i would do a little better. any suggestions on how to improve it would be really appreciated. i think one thing is that my system resources are at 75% :mad: wtf! i close all the programs that come on at start up and the best i can get is like 85%... and that goes down really fast. when i ran it it was at 75% grrr. also, wouldn't my video card have a lot to do w/ it? i heard that TNT2's aren't that good, but I didn't have a huge budget when I bought my comp. I just got it a few months ago, and it replaced my POS 75 mhz, so i'm not very up to date on all the new stuff. ok, you've heard my life story, give me some suggestions if you want. also, read my post titled "I can't play UT online" & help me out... thanks.
oh yea one final thing, my monitor cable came w/ one pin missing (haven't got around to calling the m) it still works most of the time. but do you think this can greatly degrade monitor performance? later

Athlon_2o0o
3rd Jun 2001, 10:56 AM
Originally posted by (SOB)Odie
:confused: So How Important Do Ya All Think The Frames Per Sec Are ? Ping An Lag Play A Big Part In The Art Of Fragging Dont Ya Think?

Yep, but Frames Per Second is very important. It's your reaction time. But ping is also very important for the same reason.

Odie|BuF
3rd Jun 2001, 02:45 PM
:D okay thanx athalon i was just wondering i ran time demo through a couple of games on our server i had a ping of over 300 :mad: (wife constantly downloading movies non-freaking stop) an had a average of 74.37 an 74.45 through both games :p

Athlon_2o0o
3rd Jun 2001, 05:37 PM
Originally posted by (SOB)Odie
:D okay thanx athalon i was just wondering i ran time demo through a couple of games on our server i had a ping of over 300 :mad: (wife constantly downloading movies non-freaking stop) LOL DivX and Kazaa.com rule hehe an had a average of 74.37 an 74.45 through both games :p
hmm, that's pretty good !!!!!

Odie|BuF
3rd Jun 2001, 09:19 PM
amd athlon 1.2ghz
raydeon64 tv out ddr
soundblaster live surround sound on
details an dynamic high an on
b.c.on icq avalible
played 3 hrs system resources at 63%
800 by 600 (1024by 760 too small cant see that good)
cable, 20 gig hard drive
mim 40.37
max207.09
average 140.70

JayX
3rd Jun 2001, 09:32 PM
amd athlon 1.2ghz
raydeon64 tv out ddr
soundblaster live surround sound on
details an dynamic high an on
b.c.on icq avalible
played 3 hrs system resources at 63%
800 by 600 (1024by 760 too small cant see that good)
cable, 20 gig hard drive
mim 40.37
max207.09
average 140.70


You arent suppose to put how many FPS you get IN-GAME but your suppose to put what you get in the cityintro.unr. If everyone posted what they had IN-GAME then all scores would be different because there would be different levels, different number of people playing in the game,etc.

SadaraK
3rd Jun 2001, 09:51 PM
Hold god to get an average of 140 ud need a 3 gig processor and the GF3 Ultra! (that isnt even out yet). Ud think people would actually read the thread b4 posting......

Athlon_2o0o
4th Jun 2001, 03:51 AM
Yep, post the fps from your cityintro.unr.

GuruJockStrap
4th Jun 2001, 06:56 AM
AMD 1.2GHz
Abit KT7A
256MB PC133 Cas2
Elsa Gladiac Geforce 2 Ultra 64mb
Videologic Sonic Fury
IBM 30gb Deskstar UDMA
Pioneer Black Slot DVD
Teac Black Floppy
3Com US Robotics 56k Internal PCI Modem
Black FongKai Case
Iilyma 19" Monitor
MS Keyboard and Intellimouse Optical
Cambridge Four Point Surround Sound Speakers
Windows 2000 SP2

I play in 1600x1200 in 32 bit colour with all the graphic detail settings on and maxed, even online. I love my eye candy :D The only thing I don't enable is 3d audio.
Direct 3D is my renderer.
Here are my benchmark results from city intro(V-sync disable but on during gameplay):

1600x1200@32 bit Colour
Low= 41 fps
Highest= 127 fps
Average= 70 fps

Only thing is due to my 56k modem, netspeed caps my fps to 39-48 fps! :( Mind you it hardly ever drops..oh well :rolleyes: I want cable!!!

SadaraK
4th Jun 2001, 09:07 AM
Guru change your net speed to 6000 and (if you dont start getting lot of packet loss) your fps will be caped at about 150 instead of what it is.

Odie|BuF
4th Jun 2001, 09:10 AM
:( sorry cityintro:
min43.97
max100.27
average70.03

Odie|BuF
4th Jun 2001, 03:42 PM
:confused: just seems to me that frames per sec were more valuable while in game playing than just going thruogh the intro where there isant the reaction that you'd need while playing in a game if f.p. s. are part of your reaction time wouldn't it be more important in a match where ya need the reaction time?

Pabu
4th Jun 2001, 06:11 PM
Originally posted by Extermin8r
AMD Thunderbird 1.3 Ghz
256 MB PC-200 mhz RDRAM

I thought RDRAM only worked with the motherboards for P4's. Am I wrong?

Rooster
4th Jun 2001, 06:45 PM
Odie - the intro is actually a pretty accurate representation of the high, low and avg scores you'll see in a normal game.

Pabu - you're right. He may have meant PC1600 (which is PC100 DDR vs PC2100 which is PC133 DDR). Some people don't quite have the grasp on terminology others do. Just a matter of figuring out who's honestly ignorant (not in a bad way) and who's pulling sht out of their ss.

Odie|BuF
4th Jun 2001, 07:27 PM
:D thanx rooster im new to this sorry all thanx for your paticence just tried time demo in faces an couldnt belive the frames per sec thanx again for berring with me an ill try not to be so dumb lol

You-Suk|BuF
6th Jun 2001, 05:59 AM
Originally posted by You-Suk
specs.

cel2 566@850
512 mb 133 mhz
Matrox G400
wd 9.1 (7200 rpm)
Ibm 40 (7200 rpm)
17" Liyama

1024-768 (every set to minimum :( )

high 61
low 34
aver. 50

next week another update, i'm gettin' a
Geforce II pro 64 Mb :D can't wait

hehehehe,

got it

1024-768 (settings high) :p

high. 78
low. 32
avg. 59


new card roxxxx

Extermin8r
6th Jun 2001, 11:28 PM
Originally posted by Rooster

Pabu - you're right. He may have meant PC1600 (which is PC100 DDR vs PC2100 which is PC133 DDR). Some people don't quite have the grasp on terminology others do. Just a matter of figuring out who's honestly ignorant (not in a bad way) and who's pulling sht out of their ss.

I was mistaken. I typed up my profile without looking at the specs of my computer. Like I said, I'm not as good w/ newer technology b/c my previous computer was 5 years old. :o it's actually 266 mhz ddr-ram. but yea, i'm not pulling shlt out of my a$$. if i was gonna lie, i'd say i have a 1.5 P4 hehe. the only reason i even have a 1.3 is cuz i got a good deal. (1466 w tax) later.

badmaninheaven
8th Jun 2001, 03:24 AM
my fps are a measly 19 avg. !!!
but hey i'm yet to upgrade my PC..?
could you tell me all the tweaks in the world to overclock??

FragFest2001
8th Jun 2001, 07:49 PM
I have set every setting and advanced setting to the best looking possible. Here is what I got on the demo in 800x600 resolution.

Min. - 24.99

Max. - 86.14

Average - 54.84

Also, I have a TNT2 overclocked to 180 mhz on it's core.

Extermin8r
8th Jun 2001, 11:49 PM
hey i just noticed something that could be the reason i'm getting sh*tty fps. i went and ran the cityintro demo at 320x200 just to see what i could get and it never got higher than 60 fps. i'm thinking it has to do w/ the monitor. the refresh rate is at 60 hz and since hz means cycles per second wouldn't this have a direct correlation with frames per second (60hz=60fps)? am i just way off here or am i kind of right? and also how do i figure out the max refresh rate i can set my monitor at? lastly will this imrpove my framerate? thanks.

CrackerJack
9th Jun 2001, 11:28 AM
Cele 600@600
64mb ram (buying a stick of 128 this weekend)
TNT2 32mb - Dets 5.31
SBLIVE!
Windows 98SE

800x600
D3D
16-bit colour
World- Medium/Skins- Low

Min 50.96 Max 91.87 Avg 76.92
Min 73.97 Max 85.67 Avg 78.15

[D3DDrv.D3DRenderDevice]
Translucency=True
VolumetricLighting=False
ShinySurfaces=True
Coronas=True
HighDetailActors=True
UseMipmapping=True
UseTrilinear=False
UseMultitexture=True
UsePageFlipping=True
UsePalettes=True
UseFullscreen=True
UseGammaCorrection=True
DetailTextures=False
Use3dfx=False
UseTripleBuffering=True
UsePrecache=True
Use32BitTextures=False
DescFlags=9
dwDeviceId=45
dwVendorId=4318
Use32BitZBuffer=False
UseVertexFog=False
UseAGPTextures=False
UseVideoMemoryVB=False
UseVSync=False

However when I play online it seems my fps is capped at around 50fps ... anyone know why? I have a cable modem.

FragFest2001
9th Jun 2001, 06:19 PM
Come on people. Put DetalTextures=True on! It looks sooooooo much better. I have it on and the game still runs good enough.

Danny_Diplo
11th Jun 2001, 09:57 AM
As I'm at work I can't remember the exact details, but I get over 70fps in CityIntro at 1024x768x32 using OpenGL with high detail world textures and skins and all graphics options on, except decals and screenflash.

However, UT Bench (http://www.utclans.org/spirit/files/UTbench.zip), contorary to what some people say, is by far the best demo for benchmarking. Why? Well, it simulates intense fire-fights with 16 bots on a large map (Gothic). This is a like a worst-case scenario, and really shows how your machine will cope in real games when there is a lot of action going on, with gibs flying everywhere and lots of flak flying around. City Intro is not at all equivalent to what you get in an actual game - there are no player models in it, no weapons shooting and no tracking of gibs, bullets etc.

Below are my UT Bench scores - again at 1024x768x32 OpenGL with S3TC textures :
http://www.diplo.co.uk/images/ut/utbench.gif

My system :
Athlon T-Bird 1Ghz @ 1.2Ghz (100x12)
Abit KT7-RAID mobo
GlobalWin FOP-38 HSF
256MB PC-133 Crucial Cas2 RAM
Nvidia GeForce256 32MB DDR
Nvidia 10.41 Detonator drivers
UT Version 436 with Vogel OpenGL driver and S3TC textures enabled

Danny_Diplo
11th Jun 2001, 10:11 AM
Originally posted by Extermin8r
hey i just noticed something that could be the reason i'm getting sh*tty fps. i went and ran the cityintro demo at 320x200 just to see what i could get and it never got higher than 60 fps. i'm thinking it has to do w/ the monitor. the refresh rate is at 60 hz and since hz means cycles per second wouldn't this have a direct correlation with frames per second (60hz=60fps)? am i just way off here or am i kind of right? and also how do i figure out the max refresh rate i can set my monitor at? lastly will this imrpove my framerate? thanks.

You're right - if V-Sync is enabled (Vertical Sync) then your frame-rate will only ever be as high as your monitor refresh rate. Basically, V-Sync means that your video-card will only render a frame to the screen at the same time your monitor display is being refreshed, hence it can never be higher than your maximum refresh rate. Personally I don't know how you can stand a refresh rate of 60hz because you get awful flicker at such a low rate. Save your eyes and up it!

You need to do one of these things to solve this :

1) Disable V-Sync in your video card drivers. This will stop your frame-rate being tied to your refresh rate and give you a massive frame-rate boost. Sometimes, however, you may experience 'tearing' of images because of this, but it's rare. I'm not sure what card you have, so check your advanced properties for you drivers to see if there is an option there.

2) Disable V-Sync in your UnrealTournament.ini file, found in your UT System folder - there should be an entry saying something like UseVerticalSync=True - change it to False

3) Change the refresh rate of your monitor. To do this right click on your desktop and choose Properties>Settings>Advanced Properties>Adapter and choose Optimal from the Refresh Rate drop down. This should use the highest refresh rate your card and monitor support for that resolution. For this to work properly you need to have the correct monitor profile for your monitor under the Monitor tab.

4) Update your video drivers to the latest ones for your card.

ASMDstick
11th Jun 2001, 06:03 PM
If anyone interested in Kyro2 , i got one, cant be happier with how it peforms on UT
10b7b32 , d3d, everythings maxed out including vol light and decals on cityintro min 10.4 max 98.7 avg 37
not too bad for p3-450 , 128m , 3d prophet 4500 kyro

Rooster
12th Jun 2001, 12:52 AM
Diplo, problem with UT Bench is no one plays like that for extended periods of time.

And any heavy fire-fight UT demo will not show the differences very well when you change settings. It will also likely cripple any computer on the market (P4 1.7GHz w/V5 5500 or GeForce3 excepted)

That's what benchmarking is for - not to see who's the fastest - but also HOW'S the fastest.

If all you run is UT Bench, and you want to see how your FPS changes without volumetric lighting - and you run it before and after and get a .2fps increase. OMG! WOO HOO!.. uhm.. not.

If you run the Cityintro - which is MUCH more similar to a real game - and you get a 5fps increase - then you can figure out if it's worth it or not.

ASMDstick - that 10fps minimum is a killer. :(

Extermin8or - remember to change your desktop to the resolution you play in when you go to choose what refresh rate you want (85Hz is good, but I like 100+).

Danny_Diplo
12th Jun 2001, 08:05 AM
Originally posted by Rooster
Diplo, problem with UT Bench is no one plays like that for extended periods of time.
Guess you ain't played on any heavily populated public servers then?

And any heavy fire-fight UT demo will not show the differences very well when you change settings. It will also likely cripple any computer on the market (P4 1.7GHz w/V5 5500 or GeForce3 excepted)
Well, I don't have either, and get around 53 fps in UT Bench, with a very acceptable minimum of 38 fps. And the minimum frame-rate is the most important aspect of benchmarking your system - it's when your framerate dips in heavy fire-fights that jerkiness creeps in and you die because you cannot respond easily. UT Bench simulates these fire-fights, that happen all the time in real games, very well and lets you see what this minimum is. If you can get a reasonable minimum in UT Bench then you know you're going to be all right in real games.

That's what benchmarking is for - not to see who's the fastest - but also HOW'S the fastest.

If all you run is UT Bench, and you want to see how your FPS changes without volumetric lighting - and you run it before and after and get a .2fps increase. OMG! WOO HOO!.. uhm.. not.
Oh, you can see the differences. Turning off decals earned me 5 fps in UT Bench. Also, as I mentioned above, you're average frame rate is not what matters, it's what your minimum one is that is important. In your example you might only get an average increase of .2fps, but you may well see your minumum rise from 25 to 30 fps - that is what is important.

If you run the Cityintro - which is MUCH more similar to a real game - and you get a 5fps increase - then you can figure out if it's worth it or not.

I'm sorry, but how can you say Cityintro is in any form like a real game? There's no HUD, there's no weapon on screen, no player models, no fighting, no flak, rockets or ripper blades bouncing around - in fact none of the things you encounter in a real game. Which is why hardware sites such as Tom's Hardware etc. use UT Bench when benchmarking machines, not Cityintro.

So now you know :)

Axel
12th Jun 2001, 01:22 PM
Hi All
I just got my G3 a week ago and thought i would bench for you guys. City intro marks and settings are as follows.

Athlon 1gig
256mb pc133 ram
Elsa geforce 3

UT Settings are ALL HIGH, res is 1024x768 (new monitor aint here yet but i dont expect it to drop to much at higher res) 32 bit colour, 32 tap Anisotropic filtering and 4 times FSAA (really makes UT look SOOOOO good, but i get too distracted when im playing cos i keep looking at the pretty pictures)

MIN 42, MAX 128, AVERAGE 70

Axel

Rooster
12th Jun 2001, 08:45 PM
I'm sorry, but how can you say Cityintro is in any form like a real game? There's no HUD, there's no weapon on screen, no player models, no fighting, no flak, rockets or ripper blades bouncing around - in fact none of the things you encounter in a real game. Which is why hardware sites such as Tom's Hardware etc. use UT Bench when benchmarking machines, not Cityintro.

So now you know :)

Because on average, with highs & lows, it's almost exactly what you would see during a real game - with the exact same highs, lows and average. So yes, it's the best judge of how a normal game will go. UT Bench is like blowing up cars with 1 stick of TNT - and seeing which make/model has the most left. It's so totally not indicitive of how a real game goes.

Guess you ain't played on any heavily populated public servers then?
:rolleyes: Nope. I haven't played in any 32 player DM's or 10vs10 CTF games or played just about every style of play with the game's best clans. Nope. I'm a n00b. :)

Btw, my numbers about volumetric lighting were totally out of the blue. My point is, if you hit your minimum for 5 seconds out of a 20 minute game - it's not worth losing the graphic quality the other 19 minutes and 55 seconds to fix it.

Kray
13th Jun 2001, 02:02 AM
Guess you ain't played on any heavily populated public servers then?

You mean like All-Face server with 24 max players? Hehe, no thanks.

iolair
13th Jun 2001, 06:45 AM
AMD K6-2-500
64MB RAM @ 100 MHz
Voodoo 5 5500 PCI
Windows 98
VIA MVP3 motherboard

800x600x16
skin textures high, world textures medium,
VSync off, Volumetric lighting on, dynamic lighting on

Shameful Benchmark results are:

Min 13.35 :(
Max 57.07
Avg 23.43

Neil.

iolair
13th Jun 2001, 06:47 AM
(sorry ... I know I posted this in a seperate thread ... please don't hurt me)

OK, I've created version 1 of a form for submitting Benchmarks ... please take a look and let me have any feedback via this forum, there's a seperate thread just called "Benchmarks"; then I'll update the form as necessary and throw together the PerlScripts to collect and view the data.

For example ... is there any information that definitely is needed that I'm not asking for; or am I asking for any information that's really not relevant? In what different ways would you want to be able to interrogate the data?

How many of you would be prepared to submit your BMs like this, or be interested in viewing the data?

The form is at:
http://www.iolair.f9.co.uk/UT/UT_BM.htm

THE SUBMIT CURRENTLY DOESN'T GO ANYWHERE SO PLEASE DON"T WASTE YOUR TIME FILLING OUT THE FORM YET, JUST LET ME KNOW WHAT YOU THINK OF IT!

I also need to update the Help file with more information on checking your settings (and of course how to set desired framerate to 0 )

Thanks,

Neil

[RAPE]Kronchev
13th Jun 2001, 04:52 PM
on:
Asus A7V-E
T-bird 850@875
128 mb PC133 RAM
Voodoo 4 4500 @166 mhz (default)

in:
800x600
All sound
All detail on High
Decals on
Dynamic lighting on

Min: 35.87
Max: 123.08
Avg: 59.21


W00000000T!

-[BHS]-Firewall
13th Jun 2001, 11:18 PM
466 Celeron @ 466
192 MB ( 128 MB PC 100 and 64 MB PC 66 )
7.84 gig hdd @ 5400 rpm
V3 2000 PCI @ 143
SB Live Value (3d sound & Music Disabled)

Using Windows 98
Glide Rendering @ 800x600 w/ Med World & Med Skin
Min Desired Framerate = 999 :)
Netspeed 10000

[GlideDrv.GlideRenderDevice]
Translucency=True
VolumetricLighting=False
ShinySurfaces=False
Coronas=False
HighDetailActors=False
DetailBias=-1.500000
RefreshRate=85Hz
DetailTextures=False
FastUglyRefresh=False
ScreenSmoothing=False
Resolution=Default
DescFlags=0
DisableVSync=True
Description=

Curved Surfaces = False
Decals = False
Dynamic Lighting = True

Min 15.46 Max 75.28 Avg 42.19
Min 25.30 Max 85.02 Avg 41.55
Min 25.09 Max 86.24 Avg 41.45
Min 25.40 Max 84.77 Avg 42.16

Boogey
14th Jun 2001, 06:50 AM
For what it's worth I'm with Diplo all the way in this argument.
I'm also interested in knowing the MIN fps my system will give me in online infights. And utbench gives u a much more accurate reading of that.

The cityintro is very light on the cpu as opposed to utbench. Many aspects of the engine aren't displayed in the cityintro at all. Judging a gfx card on utbench makes less sense than the cityintro I guess, but generally I'm just interested in seeing what fps my SYSTEM call pull.

I still think cityintro is useable for testing how much impact many of the gfx settings have on your gfx cards performance. For instance texture detail settings should only affect your card - and the impact it has on fps by changing it would be less significant in utbench than cityintro.

ANYWAYS,

If I want to know how tweaking settings will influence my fps, I'll go online and record matches which represent the games I typically play.

If I wanna know how a system upgrade will affect my fps, I either have to compare scores from cityintro or utbench (both commonly available - and I'd always prefer utbench) and post them here, or give a nice PUF'er my demo and ask him to test it on his system.

[RAPE]Kronchev
14th Jun 2001, 10:06 AM
Originally posted by Rooster



Btw, my numbers about volumetric lighting were totally out of the blue. My point is, if you hit your minimum for 5 seconds out of a 20 minute game - it's not worth losing the graphic quality the other 19 minutes and 55 seconds to fix it.

but i never get that when i turn on vol lighting, it drags down my FPS 10-15 frames. its just not worth it, i dont even notice a loss in visual quality. of course, after turning that off, and putting world detail to med (which i can tell a difference between high and med, but i dont really notice when i play), i get 100 fps. yes, thats right, one hundred frames per second. and it looks kickarse also.

68minimag
16th Jun 2001, 04:38 PM
min 50
max 121
avg 75
second pass cityintro.unr timedemo

68minimag
16th Jun 2001, 04:42 PM
sorry, forgot to add, all goodies on and 32 bit color @ 800*600

kelemvor
19th Jun 2001, 03:40 PM
I'm brand new to UT. How do I get/run the demo and record my info. I have no idea how good my system works but I'd love to find out.

THanks,
--Kel

iolair
19th Jun 2001, 06:36 PM
1. How to run the CityIntro movie as a Benchmark

a. Start Unreal Tournament
b. Open the Console (press ~)
c. Type "timedemo 1"
d. Type "start cityintro.unr"
e. Press ~ again to close the console

Let the demo run through twice

f. Open the console again to view the results

The bottom entry in the console will tell you the Minimum, Average and Maximum framerates

Danny_Diplo
20th Jun 2001, 12:44 PM
A quick way of doing this without resorting to using the console is :

Go to Tools menu (I think) and select Timedemo Statistics, then go to Multiplayer and choose Reconnect To Server - presuming you have just started UT, then this will replay the cityintro for you.

Probotector
25th Jun 2001, 04:43 PM
Does anyone run with FSAA 2x/4x on?
Cus nobody seems to mention it...

SamD
26th Jun 2001, 05:27 AM
OS: windows XP beta 2 (whistler)
CPU: 500mhz P3
Mobo: Some old Abit I got a year ago
Video: Leadtek winfast TNT2 Ultra 32MB AGP (XP drivers)
RAM: 256MB @ 133mhz

1024x768, 16bit, all texture details on high, default options enabled

45FPS average (after 1 run)

most sevices as well as start up programs in XP are disabled

Deathmaker
28th Jun 2001, 09:00 AM
1 gig T-Bird @ 1 gig
Abit KT7A
256MB 133 SDRAM
Radeon 64MB VIVO (retail)
SB Live! 1024
40 GB HDD
CS 4.1 speakers yada, yada, yada.

1152 x 864 x 32bit
All eye candy, vol lighting on, detail textures on etc.
No compressed textures.

min = 38
max = 126
avg = 73

Bobafett
3rd Jul 2001, 03:49 PM
Ok attached Excel benchmark file of my two rigs and I am wondering why one is slower but on paper looks faster. Check the stats. Also why can quake 3 render higher frame rates than both counter strike and UT ?????

Bobafett

Rooster
3rd Jul 2001, 04:05 PM
I'd say it's the combination of a BX board (faster than older VIA chipsets) & the video card - Savage 4 is about 1.5 generations newer than that TNT2 M64 (which is a crippled TNT2 Ultra - I believe).

The Q3 graphics engine is much more streamlined but more heavily dependent on video card.

By the way - you ran most of your Rig2 benchmarks in 800x600, but Rig1 was in 640x480.

Rooster
3rd Jul 2001, 06:53 PM
Wow.. this thread is really going to show it's age soon.

3 upgrades for me.

First post was when I went from my Celeron 500 & Pentium3 500 to my P3-650/800. Then I posted one where I got my GeForce2 GTS (wish I had a PCI Voodoo3 to play UT on)..

Next Tuesday (stupid UPS Ground - 6 day shipping) - I'll be getting
EPoX AMD761 DDR motherboard
1.2GHz (266 bus) T-Bird
256MB PC2100 CAS2.5
to go with my
GeForce2 GTS 32MB (5.5ns)

You can bet I'll be posting before & after frame rates once it's up!

Bobafett
4th Jul 2001, 03:51 AM
so thanks rooster what would u recomend getting as an upgrade new board and CPU or just a CPU as my MB supports upto 1ghz p3 coppermine chip ? Or is it my old graphics card ?? I was looking at a p3 700 @ 933 will it shift my framerate closer to 80 ??

Bobafett

Rooster
4th Jul 2001, 11:21 AM
Check your e-mail Boba.

68minimag
6th Jul 2001, 01:10 AM
all goodies on
12 background tasks running
vsync off
no fsaa
1024*768*32:
32.63;104.64;62.98
1280*1024*32
21.05;79.85;44.23

drunkenmaster2000
8th Jul 2001, 08:20 AM
I thought I'd try this too

High textures 1024 - 65 (low 40 high 100)
Low textures 1024 - 67 (low 40 high 100)

High textures 800 - 69 (low 50 high 110)
Low textures 800 - 74 (low 55 high 120)

And I was just online, on Caves, and was getting 210 fps average @ Low tex 800 !!!1 :eek:

meh has got the following:
Athlon 1G@1.26
256Mb PC150 Mem
Abit KT7
TNT2 Ultra 32
soundblaster etc..

Imagine what it'll be like if I bother to get a decent gfx card.. Or the proper mobo for my chip..

This system made a huge leap when I installed the 12.41 detonators with Athlon support. (huge=2x)

Clayeth
8th Jul 2001, 12:09 PM
Ok, I'm probably in this thread somewhere, but I don't feel like looking for it. I don't know what's different now, but I know I was getting better framerates before I took my little break from UT. My DMA-66 controller died and I'm running my HD on regular IDE, don't know if that could do it or not.

Abit BE6-II rev-1
Celeron 566 @ 850
192mb CAS-2 PC133 ram
Elsa Erazer X (GeForce SDR)
SB LIVE!
56K :(

All goodies on, except the usual (volumetric lighting, AGP Textures, etc.)

1024x768x16
avg: 58.62
max: 108.09
min: 35.52

1024x768x32
avg: 50.83
max: 95.80
min: 23.76 :con:

I seem to remember getting in the 60's in 16-bit before, but I think I just used some lower details. I've been playing in 32-bit lately, but I think i'll drop it down, I don't like that minimum being under 30.

HuFlungDung
8th Jul 2001, 04:02 PM
Hi you guys, long time no postee!

I was so ashamed of my old athlon 550/V3-3000 specs that I went into self-imposed exile ;)

Anyway, I built a new system lately, using a Microstar K7 Master m/b, Tbird 1.2 o'ced to 1.33 (with a copper heatsink, runs at about 41 degrees C), 256 meg ddr2100, and a V5-5500, which I bought a long time ago at a good price in an auction.

I'm seeing about 95 fps average in city intro, 1024 x 768, all detail high, vol lighting off. With 2x FFSA, this drops to about 75. Frankly though, I think FFSA is wasted in UT anyway. I appreciate the option much more in my flight sims or any game with really big outdoors, like Operation flashpoint. The "horizon jaggies" are painful to look at when you have the option to turn them off, so to speak.

Rooster
11th Jul 2001, 01:00 AM
P3-800 (650@124Mhz bus)
384MB PC133
V3 3000 AGP
Windows ME

1024x768x16 - GLide
High Everything including all lighting & coronas
Sound @ 44.1k sample

Minimum 41
Max 112
Avg 61P3-800E (650 o/c to 800)
384MB PC133 CAS2 RAM
Windows Millenium
1024x768x32 OpenGL
S3TC textures (CD2)
Detonator 6.50
GeForce2 GTS @ 230Mhz Core, 376Mhz Memory
High Everything including volumetric lighting & Coronas
Sound @ 44.1k sample

Min: 39
Max: 127
Avg: 69P3-800E (650 o/c to 800)
384MB PC133 CAS2 RAM
1024x768x32 D3D
Detonator 6.50
GeForce2 GTS @ 235MHz Core, 370MHz Memory
High World, Med Skin - Vol lighting off, coronas off, VertexFog off
Min: 30
Max: 96
Avg: 551.2GHz/266 AMD Thunderbird
256MB PC2100 CAS2.5 (Crucial)
1024x768x32 D3D
Detonator 12.41
GeForce2 GTS @ 200MHz Core, 333MHz Memory
High World, High Skin - Vol lighting off, coronas off, VertexFog on, Trilinear Filtering on, Mipmapping set to High Quality (drivers)
Min: 44 (46 @ 230/380)
Max: 102 (102 @ 230/380)
Avg: 81 (83 @ 230/380)

I'll be testing out the newer d3ddrv.dll's from Epic as well as o/c my video card (not sure if it can stand that much more).

update: Don't use the newest beta d3ddrv.dll from Epic (unless you like 1/2 the FPS)! Use the one from 436 patch.

I may even try to o/c the T-bird, but we'll see. It's a retail CPU and the OEM heatsink aint all that great.

the~oreca
12th Jul 2001, 12:19 PM
I just decided to test it out on my computer...
Pentium Celeron 366
96MB PC133
Voodoo 3 3000 PCI
Windows ME

800*600*16 - Glide
Medium Everything.

Minimum 17.79
Maximum 66.71
Average 31.35 fps

Thrakhath
12th Jul 2001, 03:09 PM
OK, I have a CUSL2 motherboard with an 800 MHz P3 and a GeForce2 Ultra (oh Yeah! :D).
Settings were 1600 X 1200 X 32 with all eyecandy on maximum. I got:
Min 31.27
Max 96.43
Avg 55.19

Clayeth
12th Jul 2001, 07:23 PM
Damn, i thought rooster was dead... must try again :con:
You just lurk now, or really show up that rarely?

Rooster
12th Jul 2001, 07:55 PM
Thrakhath - You must note that you're running in OpenGL - my tests were done with D3D as noted.

Clayeth - I been around a bit lately - do a search on my name, last 5 days. :D

Thrakhath
12th Jul 2001, 08:01 PM
Why do I have to note that?

I think 55 FPS is pretty good at 16x12x32 with all the goodies. Am I wrong?

Rooster
12th Jul 2001, 08:39 PM
Yes, but you can't do 1600x1200x32 in D3D.

OpenGL scores are typically higher. So to be accurate, you should tell people what you're using.

Rooster
16th Jul 2001, 06:31 PM
1.4GHz/266 AMD Thunderbird (1.2GHz CPU x 10.5)
512MB PC2100 CAS2.5 (Crucial)
1024x768x32 OpenGL
Detonator 12.41
GeForce2 GTS @ 200MHz Core, 333MHz Memory
High World, High Skin - Vol lighting off, coronas off
Min: 59
Max: 144
Avg: 91

I noted the changes in bold.

BIG differences.

I'll likely try and bump up the core speed & memory speed a little on the GTS next.

the~oreca
16th Jul 2001, 09:35 PM
Wow nice, my new PC will have about the same thing but a little cheaper since I don't have enough money :D

Rooster
16th Jul 2001, 11:27 PM
1.4GHz/266 AMD Thunderbird (1.2GHz CPU x 10.5)
512MB PC2100 CAS2.5 (Crucial)
1024x768x32 OpenGL
Detonator 12.41
GeForce2 GTS @ 200MHz Core, 333MHz Memory
High World, High Skin - Vol lighting off, coronas off
Min: 59
Max: 144
Avg: 91

at 1280x960x32 OpenGL
Min: 38
Max: 133
Avg: 67

at 1600x1200x32 OpenGL
Min: 20
Max: 55
Avg: 32

Thrakath, could you please enlighten me as to how you got a high of 90+ and avg of 55.. whereas mine are 55 high and 32 avg? GeForce Ultra doesn't help that much - and I know UT doesn't care which you use, P3 or T-Bird - just whichever gives it the most juice.

Thrakhath
17th Jul 2001, 11:35 PM
As far as I know, the GTS and Ultra perform VERY similarly, at lower resolutions. The Ultra has more bandwidth than the GTS, and is able to push very high resolutions through much quicker than the GTS, which is bottle necked. To the best of my knowledge, if we were to put a GTS in a system at 1024 x 746 x 32 and an Ultra on the SAME system, the FPS would be almost identical (with the Ultra taking a 2-3 FPS lead). However, crank the resolution up to 16x12x32, and the narrow pipeline of the GTS causes it to fall way behind the Ultra.

Other than this, I do not know.

Rooster
17th Jul 2001, 11:48 PM
Hrm... I have a GTS w/5.5ns ram - I can try to o/c to near Ultra speeds... actually would close in on Pro speeds. I'll see what I can find out.

But that's a HUGE difference. Could be the 64MB of ram... I bet that's it.

the~oreca
17th Jul 2001, 11:48 PM
This thread is to long to go through all at once so maybe you can guess what kind of FPS would I get with this:
AMD Athlon Thunderbird 1.20 GHz
Asus A7A-266
256MB PC-2100
32MB Creative Labs GeForce256 DDR

I'm guessing I'm going to play at either 1600*1200 or 1280*1024... So tell me basicly around what should I look forward to.

Thrakhath
17th Jul 2001, 11:53 PM
Originally posted by Rooster
Hrm... I have a GTS w/5.5ns ram - I can try to o/c to near Ultra speeds... actually would close in on Pro speeds. I'll see what I can find out.

But that's a HUGE difference. Could be the 64MB of ram... I bet that's it.

It's not the speed, it's bandwidth. Sure, being fast lets you push it fast, but the extra bandwidth of the Ultra lets me push it AL AT ONCE. But you are also right, with 64MB, the whole surface can go in vram. That would help a lot. You didn't say how much you had (as far as I could see), but I'm guessing that you have 32. The reason would be that at 16x12x32 you are using 61.4 MB JUST for the surface.

Rooster
18th Jul 2001, 12:00 AM
Oreca, I'd say, 30 min, 90 high, 60 avg.

I could be guessing low though.

Not sure how big a difference the GeForce1 vs GeForce2 is (for UT).

the~oreca
18th Jul 2001, 12:09 AM
Well as far as I know the GeForce256 DDR out performs the GeForce2 MX... But It drags away from the Ultra and the GTS, not sure about Kyro's though.
Those are good scores, I guess I won't need to buy another video card right now after all :)

Rooster
18th Jul 2001, 12:17 AM
Oh yeah, those would be for 1024x768x32.

Deathmaker
18th Jul 2001, 07:13 AM
Originally posted by Thrakhath
It's not the speed, it's bandwidth. Sure, being fast lets you push it fast, but the extra bandwidth of the Ultra lets me push it AL AT ONCE.
Hi Thrak, the speed is the bandwidth: faster RAM = more MHz = wider bandwidth.
Look at SDRAM for example:
PC100 = 100MHz x 64bit bus = 800MB/s bandwidth.
PC133 = 133MHz x 64bit bus = 1064MB/s bandwidth.

The same thing applies to the RAM on the GFX card i.e. the faster you get it to run the wider the bandwidth. :)

Thrakhath
18th Jul 2001, 12:24 PM
Ok, you're right. I was refering to memory pipelines. I think. I consider myself to be fairly knowledgable about computers in general, but am just learning about the various differences in Graphics Cards, so forgive any mis-information here, please. Thanks.

Hav0c
18th Jul 2001, 03:01 PM
CPU: PIII 866 MHz
RAM: 128 Mb SDRAM
Video: Riva TNT2 M64
Sound: Sound Blaster Live! Value
OS: Windows ME ( :mad: )
Resolution: 1152 x 864
Colour Depth: 16 bit
Detail: High/High

Min: 17.09 fps
Max: 59.99 fps
Avg: 32.90 fps

WebSlinger
19th Jul 2001, 09:38 AM
I will be running benchmarks this week. My system

1.33GHz Athlon
256 DDR RAM
ASUS V8200 GeForce 3
OS Win2k

UT everything set to high

I ran unofficial fps tests (1024x768x32 OpenGL) in Liandri (140 avg) and LavaG (72 avg), but before I do the cityintro thing, if anyone has any tweak tips or suggestions please email me (webslinger@planetunreal.com), I wanna see what this thing (GF3) can do.

EL_NINO
19th Jul 2001, 05:03 PM
1.2@1.4 WBK38 and it is an AXIAR Y stepping should go to 1.5+
Iwill kk266-r
v5 5500 with 1.18bios and 1.04 drivers @166 I think! might be OCed don't remember now.
win2k server sp2
512 pc133 cas 2-2-2 crucial

1024x768x16 on high/high

61.5 min, 147.2max and 96.9average

WebSlinger
20th Jul 2001, 12:12 AM
1.33GHz Athlon
256 DDR RAM
ASUS V8200 GeForce 3
OS Win2k

UT everything set to high

1024x768x32 OpenGL

I expected higher, maybe I need to tweak something?

Min 40.77
Max 60.54
Avg 56.94

just seems a tad low

Rooster
20th Jul 2001, 01:15 AM
Yeah, your VSYNC is on.

See how close your avg comes to your max? And your max is at 60 - that means your vysnc is on and your refresh is set to 60Hz.. YUCK! <southern accent> Change that, boy! </southernaccent>

:D

Razer_Zero
20th Jul 2001, 03:45 AM
It really is Futile you know?
We all go for these huge frame rates, but the truth is you will never see most of those frames. You monitor is most likley refreashing at 60hz, and therefore a 100fps is a waist of clock time. On top of that, you will not really notice after about 40fps anyway.

there, ive said my piece :)

<*ASUS A7V 266 mobo*>
<*AMD T-BIRD 900mhz PROC*>
<*512 megs MICRON DDR*>
<*30 gig Maxtor 7200 hd*>
<*PROMISE UDMA100*>
<*ASUS V7700 ULTRA GEFORCE 2*>

I wont bother posting my frame rate.

Razer.

WebSlinger
20th Jul 2001, 09:34 AM
thanks Roo, I'll turn it off. :D

Razer, yeah, it's more of a "let's see what this thing can do." When I was studying 2D TV broadcast computer animation they never let us do anything over 30fps.

Danny_Diplo
20th Jul 2001, 11:01 AM
Originally posted by Razer_Zero
It really is Futile you know?
We all go for these huge frame rates, but the truth is you will never see most of those frames. You monitor is most likley refreashing at 60hz, and therefore a 100fps is a waist of clock time. On top of that, you will not really notice after about 40fps anyway.

Huge frame rates aren't futile at all. For a start, most monitors allow a much higher refresh rate than 60hz - I don't know anyone who runs a monitor that low, it seriously flickers at that rate. My monitor has a refresh rate of 100hz at 1024x768x32, and so do most modern ones.

Second, if you disable v-synce then your frame-rate ain't tied to your refresh rate anyway, allowing you to get the maximum out of your card.

As for not noticing the difference after 40fps, this is bull. At 40fps average you are probably dipping to 20 fps or lower in heavy fire-fights, and when that happens you are at a serious disadvantage over someone with a higher frame-rate. Also, aiming with hitscan weapons like the rifle becomes much smoother at higher frame-rates. Sure, there is a point where extra FPS becomes negligible, but it's much higher than 40.

SadaraK
20th Jul 2001, 11:11 AM
I agree danny, also dont forget that if your getting 100 fps on ur own in a game then when the maximum amount of action is taking place you can cut that frame rate by 3/4 so have a high a farme rate as possible to start with is essential.

The Gravedigger
20th Jul 2001, 11:35 AM
Originally posted by Matsumoto
FACT: after 56fps your eyes can no longer tell the difference. Most movie theaters run the film at 60fps. Any card that pushes 60fps is a great card, regardless of the manufacture, again for the reason stated before.

WRONG.

A Cinema shows films at 24fps (they can do it this low cos you sit in the dark and you get burn-in between frames making it seem much smoother).

TV is produced at 30fps, but TV's run at 60hz so each frame is shown twice - tricking you into thinking it is more.


How many FPS you need before it appears "smooth" (ie. the optic nerve is saturated) varies between 60-70, depending on the length of your optic nerve, but once you have reached 72fps, then (theoretically) NOBODY can tell the difference between that and say 1000fps.

Thrakhath
20th Jul 2001, 12:47 PM
What you say is true, but please note that movies and TV are predefined. They now what's going to happen and are able to keep their FPS stable. With a computer however, it as to calculate everything every frame so there is no way to keep it constant. That is why very high frame rates are good, because even if a lot of extra stuff is piled on the video card, it's still not enough to knock it down below a sutible frame rate.

WebSlinger
20th Jul 2001, 02:28 PM
Roo, thanks, that did th trick. I will post again tonight

The Gravedigger
20th Jul 2001, 03:26 PM
Originally posted by Thrakhath
What you say is true, but please note that movies and TV are predefined. They now what's going to happen and are able to keep their FPS stable. With a computer however, it as to calculate everything every frame so there is no way to keep it constant. That is why very high frame rates are good, because even if a lot of extra stuff is piled on the video card, it's still not enough to knock it down below a sutible frame rate.
Oh please, your'e trying to teach your daddy to f*ck here - I'm well aware that what matters is minimum FPS etc, I was just being a geek :)

Thrakhath
20th Jul 2001, 03:30 PM
Originally posted by The Gravedigger
I was just being a geek :)

Hey me too, don't harass ME.

The Gravedigger
20th Jul 2001, 03:38 PM
Originally posted by Thrakhath


Hey me too, don't harass ME.
:)

Rooster
20th Jul 2001, 11:59 PM
Do I need to make a separate "Visible FPS" thread for you guys? Cut it out. :D

LordKhaine
21st Jul 2001, 01:15 AM
fps varies for different people, I can tell the difference between 40 and 50 no problem. I'd say around 60-70 odd it doesnt matter anymore.

As for v-sych, it only effectively caps your frame rate at a high fps. So while your avg fps will be lower, it would appear exactly the same ingame as with v-sych off. So turn it off for fps tests, and put it on for playing

NiftyBoy
21st Jul 2001, 04:24 AM
My specs:
Pentium III
866mhz
384mb 133mhz SDRAM
Ati Radeon VIVO 64mb DDR
Soundblaster Live!
Windows ME

Game Setup:
1152x864
32-bit
D3D
High world detail, etc.

I ran the intro 3 times just to make sure I was getting a level FPS.

1. Average: 65 fps
2. Average: 65 fps
3. Average: 66 fps

Highest: 122 fps
Lowest: 38 fps

WebSlinger
21st Jul 2001, 12:34 PM
1.33GHz Athlon
256 DDR RAM
ASUS V8200 GeForce 3
OS Win2k

UT everything set to high

1024x768x32 OpenGL

Min 49.94
Max 156.78
Avg 82.90

NiftyBoy
21st Jul 2001, 03:08 PM
i know this isn't exactly on topic (but neither was that fps debate :)) but why can't i use OpenGL with my vid card? it just doesn't show up in the available video devices, but 3dfx glide does for some reason.

i had this same problem in unreal and had to get smirftsch's patch...

Thrakhath
21st Jul 2001, 03:10 PM
Tell it to show "uncertified" drivers, since GL is uncertified. Then you can pick it after reading a disclaimer.

NiftyBoy
21st Jul 2001, 04:29 PM
ah ty. strange though...d3d is faster in UT than opengl, but its the other way around in unreal...

Rooster
21st Jul 2001, 07:34 PM
Negative Nifty, OpenGL is faster in UT than D3D - if you have the system to push it (ie., it can attain higher rates than D3D, but you have to have a little more juice).

Check my previous posts (page 10 I believe). I compare D3D & OpenGL rates with the same system.

NiftyBoy
21st Jul 2001, 08:51 PM
thats weird... in d3d i get 65 fps average on the intro, but in opengl i get 40-50 fps... maybe it's cos im using detail textures =\

Rooster
21st Jul 2001, 09:37 PM
As am I, but not 2nd CD textures... which last I checked actually ran faster than the normal textures.

LordKhaine
22nd Jul 2001, 08:51 PM
P3 500, 256ram (One 128 stick at 133, but the other 128 is only 100 :()
Win98 SE
Full install of UT:Gotye (4.36)
32mb GF2 GTS with latest official drivers (6.5 if I recall)

Intro benchmark @ 1024*768*32 in opengl
medium detail, dynamic lighting turned on, decals on, reflective surfaces on, fog turned on etc. etc.

Max 77.50, Min 15.87, Avg 31.99

Same as above, but in D3D

Max 76.22, Min 11.81, Avg 36.91


I normally play UT in D3D (ingame it seems smoother I think) with dynamic lighting, fogging and decals off, and at 800*600*32.

NiftyBoy
23rd Jul 2001, 10:21 AM
but dynamic lighting is one of UT's specialties... i say good day, khaine!!!!

anterobs
23rd Jul 2001, 04:00 PM
Specs:
Pentium III-800 MHz :)
64 MB PC100 Ram :(
Windows ME :mad:
Slot AGP 2x :(
ATI Rage Fury MAXX 64MB/AGP 2x/4x :)
ATI D3D driver set to not wait vsync http://forums.planetunreal.com/images/icons/icon4.gif

Conditions:
UT 432
Details low/low :( (too much HD swap with 64MB)
Fog/coronas/reflective surfaces/decals off
Sound low quality/11kHz
Min. desired framerate 20
Direct 3D
640x480x16

Results:
Average: 63.89
Minimum: 42.97
Maximum: 83.99
MAX/MIN: 1.96

:hmm:
Well, my conditions are not that favorable. The amount of ram and the damned 2x slot may be limiting the overall performance. It should beat a Geforce 256, at least in Opengl, according to ATI.

http://forums.planetunreal.com/images/icons/icon4.gif Altough the higher framerate, this lead to some flickering textures, so I play with vsynch, and it still looks fine. And the 640x480x16 resolution is the one I actually play. I could test it at 800x600 or 1024x768 (the maximum My monitor seems to handle). :hmm:

Quake 2: just for reference.
640x480, opengl, gl_mipmap_nearest, maximum texture detail
Timedemo demo1: 107.6 fps

Rooster
23rd Jul 2001, 08:54 PM
Dude, Ram is cheap - grab 256MB, drop it in there.

AGP 2x isn't really an issue - specially not with a 64MB video card.

What's wrong with WinME? It runs great here (on all 4 of my computers)

Rage Fury's are typically sub-par 3D performance. Sorry my friend.

Upgrade UT to 436

At least put world at medium or high. That won't make much difference on swapping (as it's all loaded in the beginning of the game).

Gotta turn on those reflective surfaces (and decals!)


I'm willing to put money it's the Rage Fury holding you back (My V3 3000 out ran your specs (on my P3-800 - see my first post).

anterobs
23rd Jul 2001, 11:41 PM
Now @800x600x16, details medium/medium, waiting vsync:
Average: 43.43
Min: 20.94
Max: 72.19
Max/min: 3.57

I've noticed the HD led blinking sometimes. Actually, with this memory, I have to play some large maps for a few minutes until the swappage drop to an "acceptable" level. I'm grabbing more memory tomorrow. I realized that this board will be useless with 64MB. :hmm:

My concern about Windows ME is not exactly perfomance, but the amount of free memory I get. :hmm:

I found this benchmark at http://www.sharkyextreme.com , comparing it to the early Geforces. I'll try to find in this thread somebody with a Geforce, to see how it compare to my performance, and try to find out if I'm having some problem or this is all that I should expect. :hmm:

Maybe I should simply get the 3dmark 2000 and se the score I get... :hmm:

But anyway, comparing it to my previous 8MB board, I'm pretty satisfied. :)

Deathmaker
24th Jul 2001, 05:49 AM
Hey, Anterobs, don't believe a thing ATI tells you about the MAXX! The box says that it's AGP4x & a 64MB card when it's NOT, it is 4x compatible (in other words it will fit in a 4x slot :rolleyes: ) but it only runs @ 2x, and it's two 32MB cards on 1 pcb that use AFR technology.
Apart from the desperate & blatant breech of the trades description act, it's still quite a good GFX card, you just need a couple of things to get the most out of it. Firstly, you've gotta get more RAM m8 (256MB for ~30 in the UK) and secondly don't bother with 16bit, 32bit will only loose a couple of FPS & it looks sooooo much better! Also I found that UT runs terrible in OGL with the MAXX (although QIIIA & Halflife look sweet) but in D3D it runs just fine. :)

Oh yeah, another of ATI's fallacies is Texture Compression so don't bother trying to get disk 2 to work in UT.

BulletRain
24th Jul 2001, 02:02 PM
ive clocked my PIII from 733 to 760 (further it will not go)
is this dangerous for the processor?:stick:

FireSlash
24th Jul 2001, 02:53 PM
<non tweaked UT.ini>

800x600, all detail settings on/high
Glide rendering system.

CPU: 500 Mhz Pentuim 3
Memory: 256 meg Kingston PC100
Video: 3DFX Voodoo 3 3000
Mobo: Generic (yeah, i know :() ATX
Sound: Creative Sound Blaster AWE 32 (Old version)
HDD: Westren Digital 72000 RPM UDMA 100 45 GIG
Other hardware
10/100 Realtec LAN card
DVD Decoder card (mainly used for routing)

Stats: 3704 frames rendered. 85.52 seconds. Min. 61.17 Max. 69.73. Avg. 63.23

FireSlash
24th Jul 2001, 03:09 PM
this is WEIRD!!!
i changed my res mode to 1024x768 and ran 3 time demos (like last time) and on average it was 3 frames HIGHER than it was on the 3 times i ran it on 800x600

:confused:

LP
24th Jul 2001, 03:23 PM
once i find some condictive paint im going to clock my duron 800 to 1 gig

Thrakhath
24th Jul 2001, 03:48 PM
Originally posted by FireSlash
this is WEIRD!!!
i changed my res mode to 1024x768 and ran 3 time demos (like last time) and on average it was 3 frames HIGHER than it was on the 3 times i ran it on 800x600

:confused:

The reason is in the math involved in the graphics calulations. Computers calulate thing which are a mutiple of 2 much much much faster than other number because the whole math system of a computer is based on the binary (two digit) system.

In 800x600, your computer probably uses the following math to calculate screen coordanites:
(2^9 + 2^8 + 2^5) -> 512+256+32=800
600, reqiures too many ^'s, so it is used as straight 600. It would be (2^9 + 2^6 + 2^4 + 2^3) -> 512+64+16+8=600

1024 x 768 on the other hand:
(2^10) -> 1024
(2^9 + 2^8) -> 512+256=768

As you can see, your processor has to only calculate 3 eqautions for 1024x768 vs 7 for 800x600.

This may seem like a small thing, but these formulas are used many times per frame to address video memory, and of course many frames per second means these are used a LOT.

I'm a game programmer and have used this kind of math in my little games, so I KNOW this one for sure.

FireSlash
24th Jul 2001, 04:47 PM
i wasnt so lucky witt 1280x1024 :( anyway, what you said leads me to believe my processor is the next thing i need to replace, and i have needed a new one for a long time anyway (i hate slot processors)

Rooster
24th Jul 2001, 04:55 PM
Fireslash, run the entire demo - your min, max & average are too close together to be a true run. As well, your frames rendered seems low (this I could be wrong on).

I've had near your exact system (see my 1st post).

FireSlash
24th Jul 2001, 06:52 PM
Originally posted by Rooster
Fireslash, run the entire demo - your min, max & average are too close together to be a true run. As well, your frames rendered seems low (this I could be wrong on).

I've had near your exact system (see my 1st post).

no, thats right. i ran it many times, and all the values ran around the same, but there alot farther apart in 1280x1024.

key differnce: sound card
sound cards eat alot of CPU cycles, lucily for me, im using a fairly old card, and due to that it has 2 propertys the SBLive! does not
1) it monterious
2) its very easy on the CPU, as back when 486 80s ruled the world there wernt too many cpu cycles to spare. might be a few other differnces i havent worked out, but im fairly sure most of it is being buffered, as i ran the demo 3 times to make sure the HD wasnt large factor, a small HD access can be the heart of you minimum value. as for the max, its limited my WinME.

Rooster
24th Jul 2001, 08:12 PM
Then why does everyone else's demo render over 6000 frames?

As for your #1 & #2 -

Monterius? What is that?

and #2 -

I guarantee the SBLive! uses FAR fewer CPU cycles than an ISA Soundblaster AWE32 or AWE64 (which I have owned, both).

And it's not WinME limiting you to 70fps - it's your VSYNC is on - and your monitor is set to 70Hz @ 800x600.

FYI, mine is 143Hz @ 800x600 and 100Hz @ 1024x768.

You can turn your VYSNC off to get a true test of FPS.

FireSlash
24th Jul 2001, 08:44 PM
i let it run all the way through, i think. i just typed this sequence and left to get somthin to eat.

timedemo 1
open cityintro1

and monterious was suppose to be Monsterious :p

Rooster
24th Jul 2001, 08:45 PM
No clue if it would be different, try

start cityintro

instead

FireSlash
24th Jul 2001, 08:51 PM
<non tweaked UT.ini>

1024x768, all detail settings on/high
Glide rendering system.

CPU: 500 Mhz Pentuim 3
Memory: 256 meg Kingston PC100
Video: 3DFX Voodoo 3 3000
Mobo: Generic (yeah, i know ) ATX
Sound: Creative Sound Blaster AWE 32 (Old version)
HDD: Westren Digital 72000 RPM UDMA 100 45 GIG
Other hardware
10/100 Realtec LAN card
DVD Decoder card (mainly used for routing)

Stats: 5784 frames rendered. 86.00 seconds. Min. 42.30 Max. 75.18. Avg. 67.25

that sound a bit better? i watched it, but it just stayed on the menu screen and played the sound, is that normal??

LordKhaine
24th Jul 2001, 09:10 PM
Damn, looking at this thread my system really does suck.

Rooster
24th Jul 2001, 09:29 PM
You really should be "seeing" the intro - not just hearing it. I have no idea how your setup got broke, but it is. :(

Sorry man.

PiKe
27th Jul 2001, 10:49 AM
AMD 900
Geforce 2 GTS
128mb 133 Ram
Abit KT7-RAID
7200rpm IBM 30gig HDD
Latest Via 4in1's
AGP 4x
SBlive!
Windows Millenium

12.41 nvidia drivers, directx8a, All detail settings high, vsync on (100hz) 16bit colour. (32bit b0rks detail textures)

Renderer D3D

Resolution 1152*864

detailtextures=true

timedemo 1

start cityintro.unr

Min 40.72
Max 99.91
Avg 66.44

:D

littleboynb
27th Jul 2001, 04:41 PM
i am averaging 42.73 FPS on a 400mhz, Voodoo3 3000 PCI, 128mb memory, with high skin/world details

is this good?

:minigun:

Rooster
27th Jul 2001, 10:09 PM
Little, you'd really have to go through the posts. That's what this thread is for - for you to compare.

I'll probably go through this thread and get rid of the "conversation" stuff.

Sir Questor=UTK=
29th Jul 2001, 10:59 AM
Athalon 1.1
256 MB Ram
Geforce II GTS


49.23 Min

134.92 Max

78.14 Average

PiKe
29th Jul 2001, 12:55 PM
AMD 900
Geforce 2 GTS 64mb
128mb 133 Ram
Abit KT7-RAID
7200rpm IBM 30gig HDD
Latest Via 4in1's
AGP 4x
SBlive!
Windows Millenium

12.41 nvidia drivers, directx8a, All detail settings high, vsync on (100hz) 32bit colour.

Renderer D3D

Resolution 1152*864

DetailTextures=False

5540 frames rendered in 84.94 seconds. Min 41.51 Max 99.80 Avg 65.21 fps.

skrill
29th Jul 2001, 09:39 PM
AMD Athlon 1.3 ghz
200 fsb
384 MB DDR PC1600
Visiontek Geforce 3 (not O/C'ed)

3dMark2001 score 5384

1600x1200x32 (FSAA Off -- who needs it at this res)
UT city into 42.42 min, 61.25 max, 56.87 avg
Regular Online Deathmatch 90-100 FPS

Is this reading right? If so, I am pretty happy.

Rooster
29th Jul 2001, 11:18 PM
Looks good to me.

That GeForce3 is pretty damn tempting -

check my rig and my stats (probably the closest to yours).

the 64MB on video card is probably the biggest difference. (I am running PC2100 RAM though (1.2GHz x 10.5 @ 133x2 BUS)

my 3DMark2001 score is like 3400... considering the 2001 is made for DX8 cards - I did pretty good (gained TWO points going from 1200MHz to 1400MHz - so I know the card is my choke point).

Not sure you'll actually get 90-100fps average at 1600x1200x32 OpenGL (gotta be OpenGL) in real game.. but if you are.. damn, I gotta get me a GeForce3 :D

skrill
29th Jul 2001, 11:25 PM
Thanks for the input Rooster.

According to the Preferences page I am running Direct3D

Here is why I am questioning it. When I go into the preferences the gui does not have 1600x1200 as an option.

To get 1600x1200 I change to OpenGL -- which lets me select 1600x1200, set it. Then switch back to Direct3d, (OpenGL does not seem to work right).

Even though it says 1600x1200 in my preferences, I am wondering if its not defaulting back to 1280x920. Most Geforce3 reviews I have seen have actually not performed well in UT because of DirectX 7. Any ideas?

Rooster
30th Jul 2001, 12:02 AM
Everything I understand says UT does not do 32 bit at 1600x1200 in D3D.

I suppose they could have changed that.

Make sure you use the .dll's from the 436, the AGP miniport driver, & 12.41 nVidia drivers.

skrill
30th Jul 2001, 12:06 AM
I am using 436.

I don't know what the AGP miniport driver is? Can you link that for me, or tell me where to find it.

Finally, I am using a beta driver 12.60 (it increased my 3d mark by couple hunded points and automatically unlocks clock freq settings).

Rooster
30th Jul 2001, 01:17 AM
It's the AMD AGP Miniport driver, dont use VIA's.

AMD Download page (http://www.amd.com/products/cpg/bin/)

Where'd you get the beta for 12.60? I wouldn't mind access to o/c my GeForce2 (i know it can do 240 clock, 380 memory vs. 200/333)

skrill
30th Jul 2001, 03:12 AM
Check out nvmax.com (http://www.nvmax.com/main/news/load.shtml)

They have the beta 12.90 -- this one might be a little shaky. I've heard its good, but it may not be as good as 12.60.

You also get them here
guru3d.com (http://guru3d.com/)

Zaphrod
30th Jul 2001, 04:13 AM
Originally posted by Clayeth054
The way the flyby intro is (just a video) the min. framerate doesn't make a difference. If you set it really high, in the actual game the quality will lower to keep that speed, that doesn't happen in the intro.

I know this is kind of late to be correcting someone but the flyby intro is not just a video. It is a map with keypointed camera movement. Setting the min desired framerate does have an effect.

I am surprised no one said anything before or maybe they have and I just missed it.

PiKe
30th Jul 2001, 07:35 AM
I think you will find that the via apg gart driver works better than the amd one, the amd agp drivers is a generic one that'll work with all amd boards but the via agp gart driver will take advantages of the mobo chipset more. (or am I talking crap?) :)

Rooster
30th Jul 2001, 09:28 AM
"I know this is kind of late to be correcting someone but the flyby intro is not just a video. It is a map with keypointed camera movement. Setting the min desired framerate does have an effect. "

Actually, I've tested it myself - it makes no difference (for whatever reason).

As for the miniport driver - I've been all over OS/Motherboard forums doing all kinds of research on what I should use, and everyone, to a Tee, said to use the AMD miniport. I think I'll stick with the knowledable majority (hell, they picked me the right motherboard! - see my rig for more questions) :D

PiKe
30th Jul 2001, 09:41 AM
so do you install any via drivers at all? Or just uncheck the agp gart in via 4in1 setup?

Rooster
30th Jul 2001, 08:50 PM
Just uncheck the AGP portion - yes, I loaded the rest.

PiKe
30th Jul 2001, 08:59 PM
I'll try it next reinstall... bout 3 months time :)

Danny_Diplo
31st Jul 2001, 11:52 AM
Originally posted by Rooster
As for the miniport driver - I've been all over OS/Motherboard forums doing all kinds of research on what I should use, and everyone, to a Tee, said to use the AMD miniport. I think I'll stick with the knowledable majority (hell, they picked me the right motherboard! - see my rig for more questions) :D

Well, I don't know what forums you've been reading, but I've done a hell of a lot of research into my Abit KT7-RAID mobo (which uses the VIA KT-133 chipset) and everything that I've read, both in FAQs and forums has said NOT to use the generic AMD AGP miniport driver and use VIA's AGP miniport driver (as found in the Via 4in1 drivers).

I'll quote you from http://www.viahardware.com/faq/kt7/faqdrivers.html#Should I install AMD's AGP miniport driver?

Should I install AMD's AGP miniport driver?

This question refers to AMD's AGP driver, not the VIA AGP driver!!

No!!! AMD's web site recommend installing their AGP miniport driver for all Athlon users - see http://www.amd.com/products/cpg/bin/, but this
probably predates the release of VIA's Apollo KX133 and KT133 chipsets. The readme file for this clearly states "REQUIREMENTS:
Motherboard with AMD-751 System Controller (Northbridge), Windows 98/98 SE (not for use on NT 4.0 or Windows 2000), AGP video card".
This driver should not be used with KX133 chipsets, and this has been confirmed by AMD support. If it is accidentally installed the only
guaranteed way of removing it is to re-install Windows as there is no uninstall facility.

Zaphrod
31st Jul 2001, 01:00 PM
Originally posted by Rooster
Actually, I've tested it myself - it makes no difference (for whatever reason).

It can make a difference in certain circumstances. Here is how you can see it working.

Set your resolution to the highest setting (32bit if your card supports it), turn on all the goodies like dynamic lights and decals and set the minimum desired framerate to 0. Run the city intro 3 or 4 times. With my computer the minimum framerate at 1600x1200x32 with everything on was just over 29FPS 4 times in a row.

Then once you have done that set your minimum desired framerate to 3 FPS above your previous minimum, for me that was 32. Now run the cityintro again 3 or 4 times. The minimum FPS for my second set of 4 runs was 31.52.

If you set the minimum desired FPS too high above what you would get with it set at zero then the engine knows that turning off the dynamic lights and other goodies is not going to bring the FPS up to your request so it doesnt try. It really doesnt change the final result or the average FPS much but it can affect it.

Ming the Merciless
3rd Aug 2001, 03:25 PM
My contribution: I have upgraded my PC, from a good PC to a kick arse PC.

Original specs: Athlon 700Mhz, 256Mb pc100, Gigabyte mobo, GeForce 2MX 3d Prophet with 32mb SDRam. Graphics card overclocked to 190Mhz.
Frame rates: Min=36.6 Max=88.2 Avg 57.8

New Specs Athlon 1.4Ghz(266Mhz FSB), 256Mb DDR pc2100, Epox mobo,
GeForce 2MX 3d Prophet with 32mb SDRam. Graphics card overclocked to 190Mhz.
Frame rates: Min=60.2 Max=88.2 Avg= 81.7

All at 800x600 @16bit and all the bells and whistles

There you go for anyone interested in upgrading. This does show that the graphics card is the limiting factor at the upper end and the processor really does help push things along when the going gets tough.

You may ask why not upgrade graphics card as well. To be honest I think the GeForce3 is a little overpriced at the moment and the gains don't seem to be that significant. Anyway my current card is responding well to overclocking. I haven't even put any extra cooling on it. Although there are fans in the case helping things along. I'll wait until the end of the year/ beginning of next year before I make any changes there.
If your wandering whats happening with my old kit, its going into my server which I might run as a dedicated server for UT.

Rooster
3rd Aug 2001, 07:01 PM
Ming, would you run it at 1024x768x32?

Our systems are VERY similar - but I have a GF GTS2 32MB & 512MB of RAM.

That's about it.

Perhaps it would be nice to see how that pans out for people looking to find out if it's really worth upgrading to an MX or GTS.

TwoHardCore
4th Aug 2001, 01:16 AM
a buddy of mine didn't run the cityintro while I was at his house, but here's his spec's and what he averaged during a Domination match with 12 players:

AMD T-Bird 1.2 Ghz
512 PC133 Cas2
AsusA7V133
ATi Radeon64VIVO (retail, stock speed @183/183)
WinME
1024 x 768, 32bit, all details maxed

DOM-Cryptic, 20 minute match
Min FPS: 78
Max FPS: 254
Avg FPS: 139!!!!

OMG, that totally made me drool. And as soon as my ASUS A7V266 comes in, I'll have a very similar rig except the video card...I've only got a RadeonLE :( (it still Rox0rs!)

Ming the Merciless
4th Aug 2001, 05:03 AM
Here you go rooster. The new figures are:
Min = 33.3fps Max = 85.8fps Avg = 68.2

It definitely shows that the weakness in the system lies with the graphics card. DDR Ram and fsb are showing their metal.

Ming the Merciless
4th Aug 2001, 05:08 AM
Just looked at your rig and OMG they are almost identical. My mobo is the raid version and i'm running Windows Me. Im going to stripe my discs when I get a mo' for faster loading.
Also to overlcock my card I've used Powerstrip, get it from overclockers.com

Rooster
4th Aug 2001, 09:55 AM
MIng: Try OpenGL - it runs a little smoother.

THC - yeah, that's about what I got on a 20 minute match on Hydro (or wherever).

I turned VYSNC on to really play though - prevents tearing.

TwoHardCore
7th Aug 2001, 04:45 PM
heh, well I'm glad that I'll be able to join you guys with those massive Frame Rates!!! I just got my mobo ordered today (will be in by Friday) and I'll pick up 512MB PC2100 on Friday too. Spend the weekend taking parts off this machine to fill the voids for the time being.

This is my 3rd upgrade since I bought UT! Me is almost ready for U2!!!w00000t!

Dead_Hood
8th Aug 2001, 12:49 PM
Specs:

AMD T-Bird 800 MHz
ASUS A7V KT133
512 Mb pc 133
SoundBlaster Live! Platinum 5.1
GeForce 3 64Mb DDR
12.90 Detonator Drivers
800x600x32

I ran 2 timedemos on cityintro.unr, one with D3D and one with OpenGL. Everything, including EAX is on, except for vsync:

D3D: Min:17.94 Max:118.84, Average: 57.38
OpenGL: Min: 29.84 Max: 75.20 Average: 55.04

Also, I know that OpenGL is capped at 75 fps as of the latest driver. After haveing read through this post, I have come to the conclusion that my minimum framrate should be at least 10-20 frames higher. Personally, I don't think that a GeForce 3 should be running at anything less than 30 fps. I know that my processor is a bottleneck, but it shouldn't be cutting my fps down that much. Any help?

Danny_Diplo
8th Aug 2001, 04:34 PM
Originally posted by Dead_Hood
Also, I know that OpenGL is capped at 75 fps as of the latest driver.
Erm, no it's not - I can get far higher than 75fps using the latest OpenGL renderer. The only thing that caps it is if you have the following line in your UnrealTournament.ini file under the [OpenGLDrv.OpenGLRenderDevice] section :

RefreshRate=75

If you delete that line it won't be capped.

Dead_Hood
8th Aug 2001, 05:32 PM
After deleting that line:

OpenGL: Min. 30.24 Max: 112.89 Average: 58.88

Better, but I would still like to figure out how to get D3D to not suck so much on my comp. :(

Rooster
8th Aug 2001, 06:57 PM
As I've posted before - make sure you turn off volumetric lighting.

SamD
11th Aug 2001, 12:20 AM
P3 500 mhz, 256mb ram, Voodoo 3 3000 AGP

1024x768x16, High, High: 98 frames per second

800x600x16, Low, Low: 110 frames per second

Both lows were in the mid 60's
I play with the second setup, it's easier.
Dumped my TNT2Ultra-based card for a V3. Getting about 60 fps more now.

Ming the Merciless
11th Aug 2001, 12:05 PM
Umm rooster... I've tried running it in Open GL and all it does is crash after a few minutes. If you need the error message let me know.

Rooster
11th Aug 2001, 12:15 PM
Perhaps because you're overclocking your video card?

Try the 12.90 drivers from nVidia (beta) & the OpenGL drivers from Epic

Try the AMD AGP mini-port... maybe it's more stable?

Probotector
11th Aug 2001, 05:01 PM
a buddy of mine didn't run the cityintro while I was at his house, but here's his spec's and what he averaged during a Domination match with 12 players:

AMD T-Bird 1.2 Ghz
512 PC133 Cas2
AsusA7V133
ATi Radeon64VIVO (retail, stock speed @183/183)
WinME
1024 x 768, 32bit, all details maxed

DOM-Cryptic, 20 minute match
Min FPS: 78
Max FPS: 254
Avg FPS: 139!!!!

OMG, that totally made me drool. And as soon as my ASUS A7V266 comes in, I'll have a very similar rig except the video card...I've only got a RadeonLE (it still Rox0rs!)

w00t!! That's some badass fps there. Im getting my new system in less than 2 weeks and these are the specs:

19" Monitor
AMD T-Bird 1.3Ghz
256Mb DDR PC2100
AsusA7M (266Mhz FSB)
GeForce2MX 32Mb DDR AGP
WinME

Jummy! I'm expecting some nice FPS from it.
(pardon my spam, just bragin about my new sys :D )

SadaraK
11th Aug 2001, 06:26 PM
Thats the exact reason why we use the intro to judge fps. Its because there are far to many variables in a normal game that can effect fps.

Rooster
12th Aug 2001, 02:32 AM
Yeah, cause in a match, before it starts, I can go outside Orbital, WAY outside in the skybox, look at the sky and hit 320FPS... and if I stay that way long enough, it makes my average like 200... so yeah, it's not really accurate unless it's a given demo.

WebSlinger
14th Aug 2001, 02:28 PM
updated marks, I am running @ 1.33 GHz now; gained a couple of FPS

UT Everything high
1024x768x32 OpenGL
ASUS A7A266 Mobo
AMD Athlon 1.33 Ghz
Thermalright SK6 (Temp 54c)
ASUS v8200 GeForce3
256 mb Micron pc2100 DDR
Wes Digi 30 GB 7200 rpm
SB Live! Value
Win2k pro sp2

Cityintro:
51.96 min
158.44 max
85.54 avg

I thought the avg would be higher, oh well, I'm happy with the thing and ready for next gen games.

Rooster
14th Aug 2001, 06:55 PM
Turn off Vol.Lighting! (cause you gotta get better than mine.. cept I'm running at 1.4GHz)

WebSlinger
15th Aug 2001, 05:25 PM
Originally posted by Rooster
Turn off Vol.Lighting! (cause you gotta get better than mine.. cept I'm running at 1.4GHz) dang Roo, first vsync now this?! OK how do I turn that off? thx

NiftyBoy
16th Aug 2001, 04:00 PM
so the faster your processor is, the faster the ati radeon 64mb vivo will be?

how would i upgrade my processor? (i have an 866mhz one right now, but im on winblowz millenium so that slows everything down)

|ll|l|evilrob|l|ll|
16th Aug 2001, 08:11 PM
OK. before all the bashing begins. This is probably a n00b question. But that is what these forums are for.

I am trying to fine tune my UT. I want to show you guys my benchmarks for the cityinro so that we can compare numbers and such.

My question is....??? How I get the FPS for the cityinro?

Thanks.

NiftyBoy
16th Aug 2001, 11:41 PM
nm @ my previous question. whoever had the 100+ fps was using a godawful resolution with a godawful color-depth. c'mon! do 1024x768 32-bit! the radeon can handle it. most people cant tell the difference from 50fps on up, and the radeon stays above that.

to get your fps, fire up UT, and after the city intro has started, hit ~ and type "timedemo 1" in the console.

wait until it's done, hit escape, hit ~ again and check out yer FPS :)

btw, these forums aren't very newbie friendly. at least in my experience. *glares at certain people*

|ll|l|evilrob|l|ll|
17th Aug 2001, 11:23 AM
OK. here goes:

SYSTEM:

Pentium III 700MHz
384MB PC100
GeForce 2 MX 400 DualView 32MB
Sony 19" Trinitron Monitor (primary)
Logitech Optical Mouse
Keyboard :p
SoundBlaster Live

First test:

@ 1024 16bit everything on.. high detail

Min 28.21 Max 60.05 Avg 46.07 fps

Second Test:

This one I let it run twice @ 800 x 600.

1st Pass> Min 29.34 Max 60.59 Avg 47.95 fps
2nd Pass> Min 30.27 Max 60.61 Avg 50.22 fps

It seems the second time the intro runs it runs way faster.

What is a more accurate read? First take or the second time around for the intro?.

Are these respectable FPS for my system?

Deathmaker
17th Aug 2001, 12:32 PM
^^^^^^^ Hehe, another one with VSync on. ;)
If you take a look thro' this thread you'll see that you can get a much higher max fps by turning off VSync both in UT and your GFX card's properties box. :tup:

NiftyBoy, sorry you feel that way. I'm sure that I've helped you out a couple of times in other threads. :hmm:

NiftyBoy
17th Aug 2001, 12:49 PM
naw... you're still one of the good ones in my book, death :)

btw whats this garbage i've heard about vsync looking better? in games like Alice this is just pure bs.

|ll|l|evilrob|l|ll|
17th Aug 2001, 02:39 PM
vsync= no good?

NiftyBoy
17th Aug 2001, 04:21 PM
well in the experiences i've had, vsync doesn't do much good. it just syncs the frame refresh to be the same as you monitor refresh, but for some reason, it chops the fps down big time.

IE in UT,

before: average 60fps
after: average 30fps

and my refresh rate was 100hz.

|ll|l|evilrob|l|ll|
17th Aug 2001, 05:34 PM
That is a HUGE difference. I don't understand why they even have that as an option. CrAZZZzy

NiftyBoy: what do you think about my FPS?

NiftyBoy
17th Aug 2001, 08:13 PM
just a general thought to everyone: there's no point to this if we don't use the same resolution and color depth.

i got 3 completely different frame rates by changing those options.

1600x1200 32-bit: ~40fps
1024x768 32-bit: ~60fps
800x600 32-bit: ~100fps
800x600 16-bit (the horror): ~130fps

so why dont we make it a standard for people to test it at 1024x768 32-bit, since that's the most common one it seems.

|ll|l|evilrob|l|ll|
17th Aug 2001, 08:57 PM
OK.. here it is @ 1024 32Bit

1st Pass: Min 27.92 Max 60.38 Avg 44.51 fps
2nd Pass: Min 28.74 Max 60.08 Avg 45.83 fps.

VSync=False (UT.ini)

I can't find vsync settings for my video card [GeForce 2MX 400]
Anyone got a clue where this is?

Are these good frame rates? or should I remove Dynamic Lighting and Turn off decals?

Thanks.

NiftyBoy
17th Aug 2001, 11:59 PM
never EVER turn off dynamic lighting and decals. dont even joke about it!

i'd suggest getting the high def textures and turning detail textures on too.

for me, vsync is under display properties-> settings -> advanced

ill hafta run another test now that im no longer using d3d like i did before...

Rooster
18th Aug 2001, 03:01 AM
EvilRob, if you're using OpenGL (like I do), go into your Display control panel, dig a little and find on the OpenGL page where it says something about Sync Refresh, say Always Off.

Once you're done testing, turn it back on.

When playing, it only affects the top end, not your normal FPS (unless your normal FPS is higher than your refresh).

Tip #2:

Step 1: Change your desktop to the exact resolution that you play UT in.

Step 2: Go in to Display Properties and change the Refresh rate of your display to something like 75 or 85 (your monitor should be able to handle 75Hz @ 1024x768x32).

Then start up UT. If you can't eliminate the VSYNC, at least this will cap you at 75 or 85 instead of 60. 60 is god awful on the eyes. Do yourself a favor and increase your refresh rate.

|ll|l|evilrob|l|ll|
18th Aug 2001, 10:02 PM
with vsync off i made it up to 50 fps

1028 @ 32bit vsync off on graphics card

ScriptLog: 4073 frames rendered in 80.37 seconds. Min 27.98 Max 93.59 Avg 50.67 fps.

Jedi Knight
20th Aug 2001, 10:32 AM
OMG these are poor. someone tell me why? is it to do with that Vsync thing and how do I sort for mine.

UT 4.36v with new D3D driver off Unreal Technology to solve problem with 32bit graphics.
Not much installed (JB, skin and voicepack)

Athlon 700mhz
384mb RAM
Nvidia TNT Riva 2 32mb

1024 @ 16bit everything on (think)
Min 16.80
Max 68.20
Avg 36.91

1024 @ 32bit it just laughed at me really
Min 8.56
Max 34.93
Avg 19.83

or should I forget about having the option of 32bit graphics and go back to original renderer?

|ll|l|evilrob|l|ll|
20th Aug 2001, 10:40 AM
i'm using OpenGL.. i don't know if that would make the difference.. but i had problems with Direct3d..

Sapphire Nights
20th Aug 2001, 04:13 PM
Jedi, you can turn off your vSync in your video tools.

But your numbers are consistant with a TNT card. You will see a big increase with even the slowest GForce card. I had a GForce 256 and it made a huge difference from the TNT I replaced.

sKorpion
20th Aug 2001, 10:54 PM
Grab the GeForce tweak util from guru3d.com. You can turn VSync and much more off and on with it. IT'S EASY TOO.

I used it and it didn't look bad at:

Mine:1024x768x32, High/High, dynamic lights and decals on, min framerate 28. No AA, medium anisotropy, no VSnyc (EVIL!!):

Min: 48.76
Max: 170.70
Avg: 85.69

Oh yeah: GeForce3 w/Detonator 3 v.14.10

The GeForce3 can really wail if you turn AA off and change colors to 16-bit. LOOK AT A WALL AND YOU GET A STEADY 610 FPS!!!!!!!!!!!!!! Yay.

http://www.redravenaudio.com/sig2.jpg aka
Avatar(of)might

(BTW, I have a lot of alias screenames. If you're desperate for somthin new, let me know and I'll cook up somthin kewl ;)

Rooster
21st Aug 2001, 12:57 AM
Skorp, post what system you have.

That GeForce3 is makin' my 1.4GHz T-Bird drool.

Jedi Knight
21st Aug 2001, 08:21 AM
what u mean video tools? The only options I can find are under display, settings, advanced but I cant find V sync under them?

megatron
22nd Aug 2001, 10:49 AM
While I haven't had UT installed for a while now, you'll see why when you read my sig. I am definiately going to blow my hardsaved cash soon on a brand new system exc. monitor. I used to get 8fps when scenery was constant and no one got too close. I used to make up for my pitifully inadequate by looking at the floor (mostly) to run about and then 'pop' heads up to look and shoot. Those were the days. Beware though because I will soon be invincible with my new super computer and will have no worries about racking up kills online since the bill is now monthly and unmetered.

Wow this is a long thread and no I haven't read it all but I will bet no one else has a dafter system than I!

jkitchar
22nd Aug 2001, 01:15 PM
Originally posted by NiftyBoy
well in the experiences i've had, vsync doesn't do much good. it just syncs the frame refresh to be the same as you monitor refresh, but for some reason, it chops the fps down big time.

IE in UT,

before: average 60fps
after: average 30fps

and my refresh rate was 100hz.

I did a vsync test recently. In one particular map and location, UT displayed exactly 65fps with vsync off. In the exact same location with vsync on, UT displayed exactly 65 fps again. My refresh rate is 85Hz BTW.

Conclusion: vsync does NOT lower framerates below your refresh rate--UT slows down exactly the same regardless of vsync. The benefit of turning vsync on is no poly tearing when you strafe=better graphics.

Sure, if a higher (inflated) cityintro demo average makes you *feel* better, turn vsync off. If you want better graphics with no loss in framerate below your refresh rate (you can't see the difference at 85Hz), leave vsync on.

Try this--turn vsync off and choose no bots in a practice session to make your fps higher than your refresh rate--like over 100 fps. Now strafe while looking at a box or wall. Notice the tearing and warping effect with vsync off. Now go back and repeat with vsync on.

VSYNC=GOOD

JMK

megatron
23rd Aug 2001, 09:14 AM
This may have been covered already in this thread but looking at the difference between average framerate and minimum framerate you can see it is quite drastic sometimes. Since no-one mentions the minimum framerate setting below which the unreal tournament engine starts 'agressively' cutting out the special effects I thought I'd bring it up. No not in that way! People must ensure that they uncheck the minimum framrate option since this will give an unrealistic minimum rate and possibly affect the avg and max frame rates.

Rooster
23rd Aug 2001, 06:59 PM
I tested it myself earlier, setting Minimum Frame rate to 0 or 50 had no effect on my tests (when my minimum was under 50).

NiftyBoy
23rd Aug 2001, 07:08 PM
i know what i experienced. turning vsync on, for me, reduced framerates horribly.

i haven't noticed any poly-tearing with it off, only better performance.

Rooster
23rd Aug 2001, 10:48 PM
Here's the thing... it will reduce your framerates so that they CAN NOT go above your refresh.

Therefore, it will lower your average because your HIGH FPS will not be as high.

Technically, it does not hurt your framerate - just the numbers.

Ichi
24th Aug 2001, 01:07 AM
1024x768 @ 16
min 40.78
max 120.09
avg 78.75

system P4 1.7
256mb
geforce2 mx 32mb

NiftyBoy
24th Aug 2001, 02:51 AM
i repeat: the point of this thread is nullified if everyone doesnt run at the same resolution, bit depth, driver, etc.

Deathmaker
24th Aug 2001, 05:35 AM
I think that it's in the works Nifty, take a look at this (http://forums.unreal2.net/showthread.php?s=&threadid=76139) thread.

Any progress with the standardisation proposals Rooster?

Layziepop
25th Aug 2001, 10:31 AM
Aight....

PIII 800, 512MB Ram, Geforce 2 MX400 64MB (OC):

1024x768x16 - 72.81 fps
1024x768x32 - 55.83 fps


That's nothin, though. Believe it or not, my brother actually gts over 200 fps... I have NO idea how he does it...

Rooster
25th Aug 2001, 11:45 AM
Not on the cityintro he doesn't.

Not at 1024x768 he doesn't.

This, I know.

It don't get much faster than mine, and I don't come close to 150, much less 200.

Layziepop
25th Aug 2001, 01:33 PM
Your right, not in the city intro. But in game he does

Jedi Knight
26th Aug 2001, 11:18 AM
whats he got like?

Zaphrod
26th Aug 2001, 02:28 PM
If you are gong to standardize the test then you should not make it 1024x768x32, it should be 1024x768x16 as some people still use Voodoo 3's and below which doesn't have a 32bit option. Just my opinion.

Machismo
26th Aug 2001, 02:33 PM
:mad:

PIII 500Mhz
TNT2 Ultra with Det3 drivers
SBlive with Liveware3
192MB Ram
Seagate 13gb HDD
UT - 1024x768x 3bitcolor x high world tex detail x low skin deatil

MIN - 9.88fps Max 36.9 fps Avg - 20.03 fps.

/Me pukes repeatedly all over system again and again and again............................

Searching thread for similar systems its 15 pages looooooooong yawn............ :)

Rooster
26th Aug 2001, 03:08 PM
Actually, there are a lot more users here with non-Voodoo based cards - and 1024x768x32 is the default gaming resolution now-days.

Yes, there are going to be some people that say that's too high! Well, look at the typical PC's now - 800MHz CPU & a GeForce2 MX is probaly average. That's not really too shabby for UT (considering what we all started with when it came out!).

Voodoo may as well run in 1024x768x16 - speed will be about the same as other video cards (when they are) in 32bit color.

Zaphrod
26th Aug 2001, 03:33 PM
OK, I was just suggesting it for complete conformity(is that a word?) for everyone. I myself have a Geforce 2 PRO so I have no problems running at any resolution or colordepth available. Just my slow 700Mhz CPU lets me down in UT because the game is much more CPU dependant than most.

|ll|l|evilrob|l|ll|
29th Aug 2001, 11:48 AM
any news on the cityintro forum cleanup?

NiftyBoy
29th Aug 2001, 11:49 AM
your garfield sig kicks ass :)

|ll|l|evilrob|l|ll|
29th Aug 2001, 11:54 AM
:D THANKS

|Robot|NeXus6
29th Aug 2001, 04:04 PM
Finally upgraded the rig:

AMD Athlon 266/1.4Ghz
ASUS A7M mobo
VooDoo 5 5500
SBlive! Platinum
512MB DDR PC2100
RAID 0 (3xWD UATA100 Drives)
UT - 1024x768x 16bitcolor x high world tex detail x Med skin deatil

MIN - 74.2FPS Max 146.3 FPS Avg 101.5 FPS

Totally rox, I highly recommend that processor as it gives great bang for the buck. The comperable P4 cost over twice as much. One aside tho, get proper cooling 'casue you could fry a dozen eggs on this bad boy.

NiftyBoy
29th Aug 2001, 05:23 PM
and in 32-bit?

16-bit blows horribly compaired to 32-bit. i dont know why anyone would even consider playing in it.

Rooster
29th Aug 2001, 07:33 PM
Not in GLide it doesn't.

I still think GLide (always 16bit) looks better than 32bit OpenGL or D3D.

Actually, I got busy with a few things, and then I thought about the system/cityintro database - and things seemed to spiral out of control.

I'll get realigned soon and see if I can tackle it later. Right now, MotorCityOnline Beta is keeping me busy - as well, baby is due in 1 month, so things are getting busy. Once baby is born next month I'm sure I'll have a lot of down time during my week off (I regularly take off a week when baby is born - yeah, this is #3 for me, #4 for her) so I should be able to do some of it then.

|Robot|NeXus6
30th Aug 2001, 12:15 AM
Personally my crappy eyes can't tell a major difference in 16bit Glide versus 32 bit in OpenGL/D3D...

NiftyBoy
30th Aug 2001, 12:30 AM
how could having half of the colors make it look better?

i havent used glide since last christmas, so i dont really remember :\

Rooster
30th Aug 2001, 12:45 AM
Well, it has to do with the way it's rendered. you don't need so many colors if it's a better algorithm.

And it's more like 65536 colors (for 16bit) and 16 million (16,777,216) for 32bit - so it's closer to 1/256 the number of colors...

IF!! IF!! They used them all - which they probably don't even use 20% of them (of the 32bit) so it's closer to 3 million colors.

But when you're moving at speeds over 70FPS - the number of colors isn't so important, it's how it appears on screen.

NiftyBoy
30th Aug 2001, 01:08 AM
jeesh! thats...jeesh...

uh...dont get me wrong here, but arent the number of colors available directly related to how good it looks?

UT seems to use every color it can get :)

NiftyBoy
30th Aug 2001, 01:14 AM
btw i was referring mostly to single player. like in WoT, it looks horrible in 16-bit, but good in 32-bit. especially the particle effects...

*sniff* aah the olden days...

iolair
31st Aug 2001, 06:35 PM
Those of you that read the "shall I clean up the benchmarks" thread will know that I've been working on a data capture site for UT Benchmarks ... well in it's initial form it's now available at http://cgi.iolair.force9.co.uk/cgi-bin/unreal.cgi

I'll start a new thread relating to this, please let me have your feedback there...

Oh, and one other thing. As I'm writing this there's only 1 benchmark in there (a cityintro.unr from my secondary system), so please re-run your benchmarks, and make your submissions via the "Add a Benchmark" link and form you reach from there!

Rooster
1st Sep 2001, 01:08 AM
nice iolair - what about Video cards though?

Deathmaker
1st Sep 2001, 09:01 AM
Update: after a bit of overclocking:-

1gig (200MHz fsb) T-bird @ 1.4gig (266MHz fsb) (51C @ 1.8V under full load yeah baby :tup: )
Radeon 64 @ 200/200

Stock results-------> O/C'ed results
min. 38-------------> 42
max. 126-----------> 135
av. 73--------------> 81

I get similar results if I clock the CPU to 1.2gig and take the Radeon upto 210/210 so it would seem (and 3d Mark 2k1 confirms this) that the Radeon is the bottleneck. :)

Lodev
1st Sep 2001, 03:20 PM
Athlon 1.3 Ghz
512 MB RAM
GF2 GTS
Win2K
1024*768@32bit

avg: 47.6 FPS


Hmm, dunno, isn't that a little low for what the system is?

Rooster
2nd Sep 2001, 12:27 AM
That does seem low - have you tried OpenGL and/or turning off VSync?

Lodev
2nd Sep 2001, 09:36 AM
I use OpenGL because Direct3D causes mouse lag, and there is no VSinc setting for OpenGL

Rooster
2nd Sep 2001, 10:20 AM
In your display control panel there is - least with 6.50, 12.41, 12.90 there is.

Set it to Off by Default.

http://rooster.eicomm.net/pics/vsync.jpg

Lodev
2nd Sep 2001, 02:34 PM
52.1 now
Oh well, maybe it's because I had left on Winamp...

Rooster
2nd Sep 2001, 04:42 PM
Turn off Volumetric Lighting (it's harsh).
Turn off Curved Surfaces (it CAN be bad).

phillys
17th Sep 2001, 09:49 AM
Pentium MMX 233Mhz with 96MB RAM, 8GB HDD, 32X Max CD-ROM and S3 ViRGE DX Graphic Card

The results:
Min. 2.39 Max. 15.35 Avg. 4.16

Game ran on Software Render. Overall result = sucky

Rukee
17th Sep 2001, 10:45 AM
900 Thunderbird @965
348mb PC133
Elsa Gladiac Ultra 64mb card
45-65 FSP @ 1920x1440 @ 32 bit color all options on!
http://www.justwebit.com/members/12651/fps.jpg:cool:

Rukee
17th Sep 2001, 11:18 AM
@1600x1200 32 bit and all options on I get 100 fps......
More then enough!!
Gotta love that Gladiac card!!!
I don`t know of anybody else that can do 1920x1440.

|ll|l|evilrob|l|ll|
17th Sep 2001, 06:28 PM
how much did you spend $$$ on the card?

Rukee
17th Sep 2001, 07:00 PM
Originally posted by |ll|l|evilrob|l|ll|
how much did you spend $$$ on the card?
You can get them for about $260 at www.pricewatch.com (the best web site for computer stuff{IMO})
I`m just about ready to upgrade the pross to mabee a 1.4ghz. They`re only $113!!!! $104 if I had the 266fsb. I don`t think it`s gonna get any cheaper in the weeks to come......

Rooster
17th Sep 2001, 09:08 PM
Sounds to me like you've seriously modified your ini or renderer settings

And again, this thread is based on the cityintro - not your own match.

Rukee
17th Sep 2001, 09:18 PM
I haven`t modified any settings! I haven`t even played with the card settings at all. Just whatever default setting are. I have the FSB of the MB up to 105mhz, but that`s it. And I think taking the test while running the game gives far more accurate readings then the intro movie. Don`t you?

Rooster
17th Sep 2001, 09:33 PM
Yes & no.

The virtues of an ingame demo vs the cityintro have been discussed at length.

Currently, only 2 benchmarks are readily accepted:

cityintro

and

utbench

You can do a forum search to find UTbench. It is made to really push the limits of the card. I think you'll find it pretty rough.

Yeah, I could run a game that would show I average 130fps at 1024x768 (I only have a 32MB video card).

But that doesn't REALLY matter if - when push comes to shove (as in the UTBench demo) you only get 40fps.

WebSlinger
18th Sep 2001, 05:26 PM
w00t! Det 4 drivers pumped me out an aditional 3 FPS! now my Avg is 88.48

I think I need to reinstall my system, too many driver install/uninstall/installs

eNio
24th Sep 2001, 11:02 PM
min 35
max 109
avg 63
WinXP Pro final, duron 800, 384mb mem(133), gf2 gts,dets 1290,abit KT7...I was wandering why my min and max are so different(i benched on oblivion full action,blood gibs n stuff 4 bots)
oh yes res. 800x600x32.:)

Rooster
25th Sep 2001, 11:51 AM
For this thread, please use Cityintro... otherwise it's difficult to compare stats.

There is a site:

http://cgi.iolair.force9.co.uk/cgi-bin/unreal.cgi

That you can upload your stats to - read carefully before you enter a number in.

Correrott
25th Sep 2001, 11:26 PM
Ok, slightly older system.

Dell Optiplex Gx1 P2 400
192 meg ram
Voodoo 3 2000
Resolution 800 x 600 16 bit
3Dfx glide
volumetric lighting off
refresh 80 khz
world detail high
skin detail med

Dynamic lighting on:
min 23.09 Max 81.46 avg 42.42

Dynamic lighting off
min 26.10 max 90.38 avg 46.87

Stitch_that
26th Sep 2001, 06:39 AM
Hi all , I'm kinda new to UT and know diddly about all this stuff , however I ran the test and these were the two results
avg 62.39fps
avg61.23 fps

Athlon 600
128mb
Nvidia TNT 2
UT set at 640 X 480

Are these the kind of fps I should be seeing ?

Thanks

Rooster
26th Sep 2001, 09:01 AM
Hard to tell.. most are tested at 1024x768 (whether or not you use 32 bit will be dependent on your video card really).

If it's a 32MB card, it can usually do 32bit fairly well.

eNio
26th Sep 2001, 08:45 PM
Ok i've got it
min 45
max 125
avg 70

Rooster
26th Sep 2001, 11:21 PM
eNio, with that system, you should run at 1024x768.

Then your numbers will look more in line.

eNio
27th Sep 2001, 08:21 PM
Yeah but i like smoothness, eye candy is not everything,what did you mean by ''your numbers will look more in line''???(my gf2 has 64mb)

Alucard
27th Sep 2001, 11:13 PM
How do you like THEM APPLES

(all of my junk, is heavily modified)
PIII(I, im not sure) 2.0ghz
1GB DDR
Nvidia G-Force 3 MX II 128
(3) 40Gig HD's


1024x768=160(min) 180(max) 170(avg)
800x600=200(max)180(min)190(avg)


Im Running Windows 98 too.
Im also benchmarking Max Payne, the damn thing runs good.

Rooster
27th Sep 2001, 11:57 PM
Where's my bullsht flag? :D

I won't bother to point ALL the easy tell-tale signs. ;)

eNio - your CPU & Video card should be able to do 1024x768x32, smoothly. UT has enough jaggies... make UT a more beautiful world - play at higher res. :D

eNio
28th Sep 2001, 06:40 PM
I know but in sticky(heavy) situations(when i play there are alot of those)fps drops, and i dont like to see lower than 50 fps(just alergic to low fps). Playing in 32 or 16bit colors doesn't make almost any difference???(also dynamic lights are off, hardly ever notice them so bye bye dynamic lights). Rooster will i be able to pull off good fps playing in 1024x768x32(yeah ill try it but nonetheless, what say you)????????:) :) :) :)

Rooster
28th Sep 2001, 06:53 PM
In OpenGL you may dip to upper 30's. when I had my P3-800, that's about what I was running. In D3D, it may be a little rougher.. depends on your settings.

Gotta have dynamic lighting. It totally rocks and barely slows the game down at all. Volumetric lighting you'll hardly notice, fog, you can lose, I even turn down the gore.

With the GeForce2, you should not see a hit going from 16 to 32 (maybe 4-8fps). 32bit (in OpenGL & D3D) looks MUCH better than 16 bit. It's fairly comparable to 16bit Glide (which looks great).

eNio
30th Sep 2001, 05:15 AM
My openGL is so dark i cant see almost anything, try to adjust brightness but no good?????:(

Rooster
30th Sep 2001, 10:26 AM
Make sure you have the latest openGL renderer. I believe that fixes it. (search for opengldrv.dll) or something similar.

If you can't find it, I've got it. Just e-mail me.