Artifaqct Tester included with Omegas, question?

Da`Nacho · Jan 27, 2004, 02:27 PM

With my Radeon 9800 Pro I get a varying number of artifactmarks when running the software test, while the DirectX test returns zero artifactmarks. This happens wether the card is overclocked or not though the artifactmarks number is slightly higher overclocked, so I'm wondering if it's normal behavior and/or if 'Artifact Tester 5' is just crap.

If you have the omega drivers could you run artifact tester and do the software test, then share your results here? Along with your results posting a basic description of your video card chipset and CPU speed would help.

Thanks!

Da`Nacho · Jan 27, 2004, 06:34 PM

Don't everyone volunteer all at once or anything, lol.

toddsmack2k · Jan 27, 2004, 06:39 PM

Give it some time man for some responses. Do you get artifacts in any other stress tests or games?

Da`Nacho · Jan 27, 2004, 07:06 PM

Quote:

Originally posted by toddsmack2k
Give it some time man for some responses. Do you get artifacts in any other stress tests or games?

Nope, which is why I tend to believe that this 'artifact tester' isn't really all that accurate. I was hoping to get a few other people to run the software test on their systems and guage wether other people got weird numbers even at stock video card speeds.

Omegadrive · Jan 27, 2004, 07:37 PM

The perfect way to make sure your card is not too overclocked is to run 3dm2k3, then in the troll's lair game test, look for white dots on the first library scene and many artifacts when she waves the sword in the library. If you see them, the card is too overclocked, even if you don't see the artifacts in other games/benchmarks.

I inlcuded Artifact Tester as a base line for OC, if you start to get artifacts in the 2D test, then your is heavily OC.

Da`Nacho · Jan 27, 2004, 07:55 PM

OK, let me try and rephrase this in the hope that at least one person will understand.

I recently installed the OmegaCorner drivers based around the Cat 4.1's and started playing around with the copy of 'Artifact Tester 5' that is included with the package. The software has 2 artifact testing modes, one of them is labeled 'Software testing' and the other is labeled 'DirectX7 Testing'. At stock Radeon 9800 Pro speeds when running the 'Software' test with any of the three bitmap sets I get a result of usually around 50 - 100 ArtifactMarks. If I overclock the card to a known safe and artifact free speed and run the software test that number increases slightly to like 75 - 200 ArtifactMarks. Meanwhile if I run the DirectX7 test features in the same program at either speed I get 0 ArtifactMarks.

At stock speed or at the same overclock used in the above scenerio I get no artifacting in any games or 3DMark2001SE / 03.

So the question I'm pondering is wether the 'Software' artifact test is a load of boloney or not, and wether I should even pay attention. I am hoping to get a few other people here to run the 'software' test in artifact tester 5 and see if they get a number higher than 0 even at stock speeds.

Clear as mud, guys?

Omegadrive · Jan 27, 2004, 08:07 PM

Well, in my experience, it does work, at least when running the software tests, if I OC too much, I start getting artifacts.
As for my score, rght now I have my 9700 Pro card OC to 364/330 and the software or D3D test shows 0 artifacts, but if I go to 3dm2k3, in the test I mentioned, I get artifacts, but no more than 10 pixel errors, which idicates to me the card is running at the maximum speed posible, if I OC one notch more I get 100's of artifacts.
Bare in mind thou, that is the ONLY palce I get a few artifacts, no artifacts have ever shown on any other D3D or OGL game at this speed.
So to answer the question, yes AT works, if you are getting artifacts at stock speeds, soemthing is wrong with your hardware OR there is some sort of communication problem between the card and the AGP bus.

ViRuS2k · Jan 27, 2004, 08:47 PM

Quote:

Originally posted by Da`Nacho
OK, let me try and rephrase this in the hope that at least one person will understand.

I recently installed the OmegaCorner drivers based around the Cat 4.1's and started playing around with the copy of 'Artifact Tester 5' that is included with the package. The software has 2 artifact testing modes, one of them is labeled 'Software testing' and the other is labeled 'DirectX7 Testing'. At stock Radeon 9800 Pro speeds when running the 'Software' test with any of the three bitmap sets I get a result of usually around 50 - 100 ArtifactMarks. If I overclock the card to a known safe and artifact free speed and run the software test that number increases slightly to like 75 - 200 ArtifactMarks. Meanwhile if I run the DirectX7 test features in the same program at either speed I get 0 ArtifactMarks.

At stock speed or at the same overclock used in the above scenerio I get no artifacting in any games or 3DMark2001SE / 03.

So the question I'm pondering is wether the 'Software' artifact test is a load of boloney or not, and wether I should even pay attention. I am hoping to get a few other people here to run the 'software' test in artifact tester 5 and see if they get a number higher than 0 even at stock speeds.

Clear as mud, guys?

===========================

hmmmm just run some tests and there seems to be something wrong with your setup or card

i run the software test under overclocked settings with my 9800 XT

0 Artifactmarks

then i run the stress testing dx test under Weak and avrage then intense and finaly hardcore..

0 artifact marks again.

now i have to do omega`s 3dmark 2003 test .... ill brb

by the way omega there are 2 trolls tests ? what one do i use ????? the ragdoll or trolls layer ?
im not shure exactly what to look for. ?

also that artifact tester 3 only tests DX 7 ? what exactly does that mean ? in terms of Stability ? i may be paranoid but splinter cell for me in teh court yard i see little white sprites ? like there floating or something hehe >.... i realy want a shure fire test so i know for definatly my card is ok and not COOKED lol 70c under default and overclocked settings and 75c under gameing.! i think thats to high other peoples 9800 XT`s are under clock 60+c never over 70c ?

ohh one other thing the tests above where done with omega 4.1`s
and under a clock of 445 core 380 memory
Hercules 9800 XT 256.

Omegadrive · Jan 27, 2004, 09:14 PM

The first one, I said the Troll's lair, not the Rag Troll

You will see what I mean as soon as the test starts to run, a lot of white dots against the black background.

As for the Splinter Cell problem, If you undeclock the card, The white sprites go away? If they do then you are OC too much, if they don't go away it could be just a bug in the Cat 4.1 drivers.

ViRuS2k · Jan 27, 2004, 09:49 PM

ok i did a test with 3dmarks and i cant see any faults at all on test 1 and 2 and 3 but the 4th test Nature is haveing a problem the water on where the camera is going up the stream where the butterfly is flying up i can see white sprites like litte white dots flickerying off and on and off and on !!! only on the water effects part .... SHITTTTTTTTT i think this card is BORKED and if i send it back to overclockers.co.uk for a RMA what happens if they dont see a fault ????? i will get fined 20 pounds testing charge.

hmm im getting all PISSED OFF LOL becuse this card cost a bomb

Omegadrive · Jan 28, 2004, 12:55 AM

Did you OC the bios directly?
No card is dammaged if you are seeing the artifacts when OC, all cards have different limits, you just have to find the max you can go without artifacts.
If the problem shows on default clocks, then the card IS dammaged.

~~BMG_Cya=-~~ · Jan 28, 2004, 02:54 AM

Into the fray...

I don't O'clock my card much. When I first get a new one, or if I add some
cooling or something.

Usually keep my card at stock or very slightly O'clocked.
I'd never done this "find the card limit" thing though so...wtf...

==========================================

Test Card:
Gigabyte Maya Radeon 9800 Pro (128)...
Tweakmonster Ramsinks applied with AS5 Thermal Epoxy
F5 Tornado 80mm Fan (100 CFM) VPU cooler
Vantec Tornado 80mm Fan (84 CFM) Case/VPU cooler
9 80mm case fans (Antec...Enermax...Evercool, etc.)
Yes...it sounds like a small airport in here

Ambient room temp. 56F

All tests run with 16x Aniso and 4x AA...forced on.

Clock speeds were:

Core: 402.90

Memory: 743.54 (371.77 x 2 = 743.54)

=============================================

Artifact Tester 5
Ran each test 10 times consecutively...all the bitmap sets. 0 artifacts

Ran the DX7 test twice on Intense...once on Hardcore (Long test that one)
0 articfacts

Immediately ran first 4 benches in 3DMark03 twice
No articfacts or any other glitches

===============================================

Repeated the above tests sequence 3 times consecutively...same results

Noticed rise in case/aux exhaust temps of 2.8 degrees C at end of tests.

Didn't find the card's limits...didn't want to bother or maybe f*ck up
a $400+ card. You guys can do that.

The day the R420 hits the shelves...it's mine.
This card will be a nice backup...I guess/think.
Running now at usual 385/715...roughly.

Oh yeah...Omega drivers 2.4.78XP (Cat 3.7 based)

Da`Nacho · Jan 28, 2004, 12:56 PM

Well, I went to a friend's place yesterday who has nearly the exact same setup as mine:

AthlonXP Barton core 2500+
EPoX 8RDA+ 1.1
2x256 PC3200
BBA Radeon 9800 Pro (stock speeds)
etc

And what do you know, the same thing happened when running Artifact Tester 5. Running the software artifact test at a desktop resolution of @ 1280x1024 results in ~50 artifactmarks using the default set of bitmaps. Interesting enough, if I lower the resolution to 1024x768 I get 0 artifactmarks running the same test.

For reference, my system:

AthlonXP Barton core 2600+
EPoX 8RDA+ 1.1 (A1 chipset)
2x256 PC3200 (Buffalo tech)
BBA Radeon 9800 Pro
etc

After doing some further investigating around the web it seems that for the most part Artifact Tester 5 shouldn't really be trusted, especially not the software test. I saw several cases just like mine where the tester would return a numvber of 'artifactmarks' greater than zero on a setup with completely stock speeds.

I can get my card up to 430/365 without seeing ANY artifacts in everything I throw at it (3DMark03, NFSU, that weird ATI demo with the spheres that I can't remember the name of at the moment, etc). I guess I shouldn't let the scores returned by Artifact Tester 5 bother me in my case.

ViRuS2k · Jan 28, 2004, 02:05 PM

Quote:

Originally posted by Omegadrive
Did you OC the bios directly?
No card is dammaged if you are seeing the artifacts when OC, all cards have different limits, you just have to find the max you can go without artifacts.
If the problem shows on default clocks, then the card IS dammaged.

no all these problems are with the card stock defaults.
i tested with my overclocked settings aswell

both show artifacts in splintercell also with 410 and 310 drivers.

this card is borked im getting it RMA`d As soon as i possibly can....
might go for a saphire 9800 XT instead.

BBThumper · Jan 28, 2004, 11:01 PM

......like Omega, I use 3dMark2k3 with the trolls scene and the Nature scene. If your overclocked too high then the water in the Nature scene will show white dots.........

Quote:

Originally posted by ViRuS2k
===========================

ohh one other thing the tests above where done with omega 4.1`s
and under a clock of 445 core 380 memory
Hercules 9800 XT 256.

.......you have your card overclocked to 445 core and 380 memory? Are you on a water set-up?Where there any artifacts when you were overclocking at say, 420-425-430-435 ? . I just had my 9700pro go that way. Ran great for awhile with a couple of artifacts and then died

. Taught me to watch o/cing more closely.........

pRAyNz · Jan 29, 2004, 12:16 AM

IMO rthdribl is the best test to check if the GPU is still working without heat related errors (little white dots which get highlighted by the program and are not related to the flying round things)
And to check the grafic RAM it`s OK to use something like Aquamark3 (and look for strangely stretched or flickering textures - but it would be optimal if nearly the whole grafic RAM is in use at any time during the tests because if only a single chip isn`t able to run at the current speed you can only detect it if you use it to store data and check it later).

ViRuS2k · Jan 29, 2004, 04:06 AM

yes i tested with my overclocked settings and default core and memory clocks.

both showed artifacts in 3dmarks pixel shader 2.0 tests
also in splintercell and in 3dmarks nature test but the nature test one was rare and the errors happened on the water effects.

i tested in 3dmarks 2003 in 5 mhz incresements the core of 455 gave artifacts so i reduced 5 mhz then i got slight artifacts then i reduced 445 and the errors/artifacts vanished.

but when i first did all this i didnt even run the pixel shader 2.0 test. just the games tests ...
thats why i didnt notice untill now of the problem with the pixel shader 2.0 test...

:/ never the less im getting RMAd ASAP

Quote:

Originally posted by BBThumper
......like Omega, I use 3dMark2k3 with the trolls scene and the Nature scene. If your overclocked too high then the water in the Nature scene will show white dots.........

.......you have your card overclocked to 445 core and 380 memory? Are you on a water set-up?Where there any artifacts when you were overclocking at say, 420-425-430-435 ? . I just had my 9700pro go that way. Ran great for awhile with a couple of artifacts and then died . Taught me to watch o/cing more closely.........

BBThumper · Jan 29, 2004, 04:54 AM

.........good luck in your RMA and please report back when you get the new card.........

ViRuS2k · Jan 30, 2004, 04:50 PM

I sent my card in yesterday so i will let you all know the status when i get it..... you watch i will have all the unluck in the world and get a card back that does the same thing lol you watch :/

Jan 27, 2004, 02:27 PM	#1
Da`Nacho DriverHeaven Junior Member Join Date: Nov 2002 Posts: 20 Rep Power: 0	Artifaqct Tester included with Omegas, question? With my Radeon 9800 Pro I get a varying number of artifactmarks when running the software test, while the DirectX test returns zero artifactmarks. This happens wether the card is overclocked or not though the artifactmarks number is slightly higher overclocked, so I'm wondering if it's normal behavior and/or if 'Artifact Tester 5' is just crap. If you have the omega drivers could you run artifact tester and do the software test, then share your results here? Along with your results posting a basic description of your video card chipset and CPU speed would help. Thanks!

Jan 27, 2004, 06:39 PM	#3
toddsmack2k It Never..... Join Date: Nov 2002 Location: Kentucky Posts: 3,173 Rep Power: 0	Give it some time man for some responses. Do you get artifacts in any other stress tests or games? __________________ [color=blue]*Bleeding From the Inside Pouring Out, The* *DriverHeaven* Spiral, We Shall Never Let Go Of.[/color]

Jan 27, 2004, 07:37 PM	#5
Omegadrive Drivers? What Drivers??? Join Date: May 2002 Location: Puerto Rico Posts: 1,526 Rep Power: 73	The perfect way to make sure your card is not too overclocked is to run 3dm2k3, then in the troll's lair game test, look for white dots on the first library scene and many artifacts when she waves the sword in the library. If you see them, the card is too overclocked, even if you don't see the artifacts in other games/benchmarks. I inlcuded Artifact Tester as a base line for OC, if you start to get artifacts in the 2D test, then your is heavily OC. __________________ DRIVERHEAVEN ADMINISTRATOR <<E-MAIL>> <<PC SPECS>> GOD BLESS YOU ALL!

Jan 27, 2004, 08:07 PM	#7
Omegadrive Drivers? What Drivers??? Join Date: May 2002 Location: Puerto Rico Posts: 1,526 Rep Power: 73	Well, in my experience, it does work, at least when running the software tests, if I OC too much, I start getting artifacts. As for my score, rght now I have my 9700 Pro card OC to 364/330 and the software or D3D test shows 0 artifacts, but if I go to 3dm2k3, in the test I mentioned, I get artifacts, but no more than 10 pixel errors, which idicates to me the card is running at the maximum speed posible, if I OC one notch more I get 100's of artifacts. Bare in mind thou, that is the ONLY palce I get a few artifacts, no artifacts have ever shown on any other D3D or OGL game at this speed. So to answer the question, yes AT works, if you are getting artifacts at stock speeds, soemthing is wrong with your hardware OR there is some sort of communication problem between the card and the AGP bus. __________________ DRIVERHEAVEN ADMINISTRATOR <<E-MAIL>> <<PC SPECS>> GOD BLESS YOU ALL!

Jan 27, 2004, 09:14 PM	#9
Omegadrive Drivers? What Drivers??? Join Date: May 2002 Location: Puerto Rico Posts: 1,526 Rep Power: 73	The first one, I said the Troll's lair, not the Rag Troll You will see what I mean as soon as the test starts to run, a lot of white dots against the black background. As for the Splinter Cell problem, If you undeclock the card, The white sprites go away? If they do then you are OC too much, if they don't go away it could be just a bug in the Cat 4.1 drivers. __________________ DRIVERHEAVEN ADMINISTRATOR <<E-MAIL>> <<PC SPECS>> GOD BLESS YOU ALL!

Jan 27, 2004, 06:34 PM	Thread Starter #2
Da`Nacho DriverHeaven Junior Member Join Date: Nov 2002 Posts: 20 Rep Power: 0	Don't everyone volunteer all at once or anything, lol.

Jan 27, 2004, 07:55 PM	Thread Starter #6
Da`Nacho DriverHeaven Junior Member Join Date: Nov 2002 Posts: 20 Rep Power: 0	OK, let me try and rephrase this in the hope that at least one person will understand. I recently installed the OmegaCorner drivers based around the Cat 4.1's and started playing around with the copy of 'Artifact Tester 5' that is included with the package. The software has 2 artifact testing modes, one of them is labeled 'Software testing' and the other is labeled 'DirectX7 Testing'. At stock Radeon 9800 Pro speeds when running the 'Software' test with any of the three bitmap sets I get a result of usually around 50 - 100 ArtifactMarks. If I overclock the card to a known safe and artifact free speed and run the software test that number increases slightly to like 75 - 200 ArtifactMarks. Meanwhile if I run the DirectX7 test features in the same program at either speed I get 0 ArtifactMarks. At stock speed or at the same overclock used in the above scenerio I get no artifacting in any games or 3DMark2001SE / 03. So the question I'm pondering is wether the 'Software' artifact test is a load of boloney or not, and wether I should even pay attention. I am hoping to get a few other people here to run the 'software' test in artifact tester 5 and see if they get a number higher than 0 even at stock speeds. Clear as mud, guys?

Jan 27, 2004, 09:49 PM	#10
ViRuS2k HardwareHeaven Extreme Member Join Date: Mar 2003 Location: Uk Posts: 8,334 Rep Power: 72 System Specs	ok i did a test with 3dmarks and i cant see any faults at all on test 1 and 2 and 3 but the 4th test Nature is haveing a problem the water on where the camera is going up the stream where the butterfly is flying up i can see white sprites like litte white dots flickerying off and on and off and on !!! only on the water effects part .... SHITTTTTTTTT i think this card is BORKED and if i send it back to overclockers.co.uk for a RMA what happens if they dont see a fault ????? i will get fined 20 pounds testing charge. hmm im getting all PISSED OFF LOL becuse this card cost a bomb

Jan 28, 2004, 12:55 AM	#11
Omegadrive Drivers? What Drivers??? Join Date: May 2002 Location: Puerto Rico Posts: 1,526 Rep Power: 73	Did you OC the bios directly? No card is dammaged if you are seeing the artifacts when OC, all cards have different limits, you just have to find the max you can go without artifacts. If the problem shows on default clocks, then the card IS dammaged. __________________ DRIVERHEAVEN ADMINISTRATOR <<E-MAIL>> <<PC SPECS>> GOD BLESS YOU ALL!

Jan 28, 2004, 02:54 AM	#12
~~BMG_Cya=-~~ Banned Join Date: Mar 2003 Location: U.S.A. Posts: 718 Rep Power: 0	Into the fray... I don't O'clock my card much. When I first get a new one, or if I add some cooling or something. Usually keep my card at stock or very slightly O'clocked. I'd never done this "find the card limit" thing though so...wtf... ========================================== Test Card: Gigabyte Maya Radeon 9800 Pro (128)... Tweakmonster Ramsinks applied with AS5 Thermal Epoxy F5 Tornado 80mm Fan (100 CFM) VPU cooler Vantec Tornado 80mm Fan (84 CFM) Case/VPU cooler 9 80mm case fans (Antec...Enermax...Evercool, etc.) Yes...it sounds like a small airport in here Ambient room temp. 56F All tests run with 16x Aniso and 4x AA...forced on. Clock speeds were: Core: 402.90 Memory: 743.54 (371.77 x 2 = 743.54) ============================================= Artifact Tester 5 Ran each test 10 times consecutively...all the bitmap sets. 0 artifacts Ran the DX7 test twice on Intense...once on Hardcore (Long test that one) 0 articfacts Immediately ran first 4 benches in 3DMark03 twice No articfacts or any other glitches =============================================== Repeated the above tests sequence 3 times consecutively...same results Noticed rise in case/aux exhaust temps of 2.8 degrees C at end of tests. Didn't find the card's limits...didn't want to bother or maybe f*ck up a $400+ card. You guys can do that. The day the R420 hits the shelves...it's mine. This card will be a nice backup...I guess/think. Running now at usual 385/715...roughly. Oh yeah...Omega drivers 2.4.78XP (Cat 3.7 based)

Jan 28, 2004, 12:56 PM	Thread Starter #13
Da`Nacho DriverHeaven Junior Member Join Date: Nov 2002 Posts: 20 Rep Power: 0	Well, I went to a friend's place yesterday who has nearly the exact same setup as mine: AthlonXP Barton core 2500+ EPoX 8RDA+ 1.1 2x256 PC3200 BBA Radeon 9800 Pro (stock speeds) etc And what do you know, the same thing happened when running Artifact Tester 5. Running the software artifact test at a desktop resolution of @ 1280x1024 results in ~50 artifactmarks using the default set of bitmaps. Interesting enough, if I lower the resolution to 1024x768 I get 0 artifactmarks running the same test. For reference, my system: AthlonXP Barton core 2600+ EPoX 8RDA+ 1.1 (A1 chipset) 2x256 PC3200 (Buffalo tech) BBA Radeon 9800 Pro etc After doing some further investigating around the web it seems that for the most part Artifact Tester 5 shouldn't really be trusted, especially not the software test. I saw several cases just like mine where the tester would return a numvber of 'artifactmarks' greater than zero on a setup with completely stock speeds. I can get my card up to 430/365 without seeing ANY artifacts in everything I throw at it (3DMark03, NFSU, that weird ATI demo with the spheres that I can't remember the name of at the moment, etc). I guess I shouldn't let the scores returned by Artifact Tester 5 bother me in my case.

Jan 29, 2004, 12:16 AM	#16
pRAyNz DriverHeaven Junior Member Join Date: Dec 2002 Posts: 29 Rep Power: 0	IMO rthdribl is the best test to check if the GPU is still working without heat related errors (little white dots which get highlighted by the program and are not related to the flying round things) And to check the grafic RAM it`s OK to use something like Aquamark3 (and look for strangely stretched or flickering textures - but it would be optimal if nearly the whole grafic RAM is in use at any time during the tests because if only a single chip isn`t able to run at the current speed you can only detect it if you use it to store data and check it later).

Jan 29, 2004, 04:54 AM	#18
BBThumper DriverHeaven Junior Member Join Date: May 2003 Location: Philly Posts: 81 Rep Power: 0	.........good luck in your RMA and please report back when you get the new card......... __________________ P4C800-E Deluxe 2.4C @ 3.5ghz 2x256 OCZ EL Gold pc3700 ATI 9800 XT X-pider Turbo case w/alot of pretty lights 3DMark2k1 Aqua Mark 3dMark2k3

Jan 30, 2004, 04:50 PM	#19
ViRuS2k HardwareHeaven Extreme Member Join Date: Mar 2003 Location: Uk Posts: 8,334 Rep Power: 72 System Specs	I sent my card in yesterday so i will let you all know the status when i get it..... you watch i will have all the unluck in the world and get a card back that does the same thing lol you watch :/