There is still a big slowdown even with mirrors off , Its just it seems even worse with mirrors on ,its sometimes hard to get an even fps because fps changes with the movement around the track and the more cars drawn ect ,
are the car reflections drawn on top of the cars texture? and are reflections drawn in mirror?
It seems the V mirror slowdown I get is not so drastic from last place car (with no cars in his mirror).

SD.

You could try LFS S1 and switch on/off simple car to see if the reflections are causing the slowdown.

Right, just tried a quick online race, that was fun at 15fps, not, and if I turn the clocks off I get a big increase, from 15fps to about 45fps, not back to my "normal" speeds but a big increase non the less.
Virtual clocks or clocks "on" make no difference, but clocks OFF does.
Help at all?.

Dan,

I have gone back to 5.11 drivers because of the system taken up by the new CCC in the ATI drivers from 5.12 onwards, the control panel option has been dropped forcing users into the CCC config. I had evil fps drop across the board and used some tuning tips I garnered from a post in "another" forum, but even using the tips, which are a PITA to setup, the drop was still there just not quite as bad. I have recently bought an AX850XT AGP card to replace my 9600xt, still using the 9600xt atm tho cos the 850 was wrecked during delivery, along with an Audigy 4 and some other stuff, by the courier, so I will check again once it gets returned to see if the newer card suffers as badly.


Sys Spec
AMD 3000+ barton
1.5 gig PC2700
256 Sapphire 9600xt
Creative 5.1 digital

Sam

You can use the 5.12/13's sans CCC just by first installing the ATT (ATI Tray tools) and then downloading just the DRIVER from ATI, not that it makes a blind bit of difference to the performance in LFS.....

Dan,

P4 3,0 GHZ
512 MB
ATI X800 pro chip

no single frame less with 5.13, no matter if mirrors are on or off

Is your p4 an HT one?

Quote from iHomer :

Is your p4 an HT one?

yes (im pretty sure :shy

After reading the problems with the 5.12/5.13 drivers i was "afraid" of installing them, but yesterday i tried them and got the following results :

Catalyst 5.11
---------------------------------------------------------------
2006-01-07 21:17:09 - LFS
Frames: 4297 - Time: 103135ms - Avg: 41.664 - Min: 27 - Max: 57

Catalyst 5.13
---------------------------------------------------------------
2006-01-07 22:55:56 - LFS
Frames: 4062 - Time: 103590ms - Avg: 39.212 - Min: 24 - Max: 55


So i experienced a minimal fps-loss of 2-3 frames. (I think its within measuring tolerance)

Its done withe the unofficial LFS benchmark.spr, but with my current config file.

I am NOT using the CCC (biggest POS ever) and my CPU has no dual-core / HT (Athlon XP 3200+).

Mirrors are are always turned on (real mirrors + virtual mirrors).

The Gfx-Card is quite old : ATI Radeon 9500 pro, all test done in 1360x1024x32 with 4xAA and 4xAF.

I've wanted to run some test on my system, to check this issue with mirrors on/off, but unfortunatly (for the solving of this problem) my problems with the performance are gone. Don't know what I've done what could have solved it for me, but now it just runs as fast as ever.

Strange thing that I was affected by it before and now it is gone. It's a pity for Scawen that I couldn't help, but nice for me.

I've ran a benchmark at Fe Green, 1 hotlap with xrg (spr).
AMD Athlon 2000+, 512 MB, Radoen9600XT 128MB, 800*600*32 no AA,AF

VGA driver, that came with the card (V7.991 i think):
2006-01-08 18:39:59 - LFS
Frames: 12281 - Time: 138360ms - Avg: 88.761 - Min: 66 - Max: 93

Cat 5.13:
2006-01-08 18:51:45 - LFS
Frames: 12084 - Time: 138594ms - Avg: 87.189 - Min: 62 - Max: 93

I haven't experienced any performance loss anywhere in LFS.

Thanks everyone for the results.

My programmer friend Alex was able to reproduce this on his computer, in the debug and release versions of DirectX, and there were no error messages output from the debug version. So our current thinking is it still looks like a fault in the ATI drivers. Just strange that other games don't seem to be affected.

Anyway, i've written again to my ATI contact, and if i'm unable to get in contact with him, i will try another person at ATI.

I appreciate the continued support for this issue Scawen, I am glad that your friend was able to reproduce it, shame it cant be fixed at your end, but hopefully ATI will fix it their end.

Dan,

Yes, keep it up. :up:

i, just as others would also be glad to help out further if possible. i'll drop my specs again just for the record,

Windows XP x64
ATI Radeon x800 xt pe (asus)
Athlon a64 3200+

let me know if there's something in particular i can try which might help.

Got a reply over at the Beyond3D forum.Looks like ATI folks are aware of the problem and trying to find a solution.

ATi Radeon 9550 256MB, 1024MB Ram, Intel P4 Prescott @ 2.8 GHz. I'm getting 90+ alone on the track, works like a charm running Cat 5.13s.

Quote from SladiVadi :

Got a reply over at the Beyond3D forum.Looks like ATI folks are aware of the problem and trying to find a solution.

Well, thats a start for sure, lets hope that they can solve it for the next release, don't really fancy dropping back to the 5.11's for the 2nd season of the ESL .

Dan,

P4 EE 3.4
X800XL 256MB AGP
ASUS P4P800-E
Win XP Pro SP1

I used to get 130+ fps with 5.11 drivers. With 5.13 I get 15-20fps!!!!

Turning mirrors off may increase the fps a little, but it is still not playable.

i have tested with mirrors on mirrors virtual and off

mirros on = 20-30fps
mirrors on and virtual = 15 - 25 fps
mirrors off = 60 - 100 fps

machine is xp1700 / radeon 8500 64 meg ddr / 256ddr ram

I was playing with the graphic options and something strange happened with the Simple track (no car shadow) option, when I clicked on ALL my usual 10 fps went down to 2 using the cockpit view

Quote from BLooDMaN :

I was playing with the graphic options and something strange happened with the Simple track (no car shadow) option, when I clicked on ALL my usual 10 fps went down to 2 using the cockpit view

PC Specs?

Just a note for ATI users: we have found the driver issue and have corrected it internally and are currently running it through QA. It will not make it into the Cat 6.1 release but we are looking into the best way to get it out to users.

Our Devrel and driver people should be in touch with the developers already, or shortly.

Thanks for reporting the bug and sorry for the inconveniences. For further problems, you can file a ticket under the www.ati.com website in the "my support" section of the drivers.

Quote from sireric :

Just a note for ATI users: we have found the driver issue and have corrected it internally and are currently running it through QA. It will not make it into the Cat 6.1 release but we are looking into the best way to get it out to users.

Our Devrel and driver people should be in touch with the developers already, or shortly.

Thanks for reporting the bug and sorry for the inconveniences. For further problems, you can file a ticket under the www.ati.com website in the "my support" section of the drivers.

Great news maybe a hotfix? so ... looking forward to the 6.2 release.

Thanks for the customer service.

SD.

Thanks for the info! :up:

Wahoo. Tbh, I'm a bit surprised it's really ATi's fault . But then again I shouldn't, who am I to question Scawens code?