HELP: Disappoing FPS even with RADEON 9700 vidcard

[Guest]

I have just upgraded my video card from a Asus GeForce 2 MX400 32 Mb ram toan ATI 9700 (non Pro) with 128 Mb of ram. I did this mainly to get higherframe rates in FS2002.As expected, the new card gave an impressive boost in Mad Onion 3D scores;from less than 3000 to 8500. Based on that I expected a decent boost inframerates (fluctuating between 7 fps - 14 fps with the old card).Disappointly, the framerate has only increased to a range between 12fps -17fps - and there is still some jerkiness in scenery with a lot of eyecandy. To get rid of the jerkiness I had to back off 1 "click" for most ofthe display settings in the FS2002 options!Is this what I should expect with this highly rated video card? (The demoson the Mad Onion 3D were awesome with extremely high 3D framerates and verysmooth video without any jerkiness, and that leads me to believe that theproblem is with my setup:Asus P4B motherboardIntel Pentium 4 CPU running at 1.6 gig (10 percent overclocked)512 Mb of RAMIntel chipset with latest driversWindows 2000 Pro (SP3)ColorPro Radeon video Card with Catalyst 3.0 driversFS2002 running full screen at 1024x768 with 32 bit color refreshingat 100 MhxI tried to experiment with all the different options for the video card -but to no avail. Still disappointing framerates.I also tried a card from Nvivia - the new ASUS ti4200S with 666 ms highspeed memory - and that was even worse....to get a higher framerate I had toset the Animorphic(?) settings to the max..and that resulted in blacksquares around all blinking lights (and the sun) ...so that card went backto the dealer.Can anyone help me out here - is there a way to get better framerates (say25fps at 1024x768 full screen 32bits color) and smooth video in FS2002 with this type of card without backing off all the FS2002 display settings?Thanks for any input.

[Guest]

Increase the resolution to the max your monitor will handle. Not above 1600x1200x32 though. Decrease your refresh rate to 85hz. Or less. I read over at rage3d that folks had preformance problems with very high refresh rates. No need to run at 100hz. Mine runs at 1600x1200x32 at 85hz. I have the Pro card and also a bigger CPU with RDRam so there's no problem running with the sliders maxed. Except AI. How's your pic quality??

[W2DR 1,859]

Actually, I'm surprised the fps increase is as large as it is. But it does speak to the issue of very poor performance from the early MX's (and the newer one's aren't any better). The 1.6Ghz isn't going to push the 9700 anywhere near it's capability. There are really only two options to get further substantual gains - either move the sliders further to the left or increase the CPU speed - or both. I don't think you'll see a smooth 25 fps any other way.Doug

[Guest]

Well, first of all, how many times has it been pointed out around here that the key to FPS in FS2002 is the CPU, not the video card? Big bucks spent on the latest and greatest video card, while leaving the CPU where it is at, is almost always a disappointment. FS2002 is just too processor bound.That said, some things you can do to make the best of the situation have already been said. Accept that you will have to back off the settings at least somewhat.Don't blame the video card for the lack of performance. The bottleneck is your CPU (and chipset).-Basil

[Tabs 2,790]

FS2002 is severely CPU limited - you'll see a big boost if you upgrade your processor...

[vscimone 16]

I just upgraded to an Nvidia gforce4 128mb from a gforce 2 gts 32mb. I had a HUGE increase in performance, and my CPU hasn't changed (1.7 ghz). To the original poster... hope you get everything sorted out...Vin

[rhumbaflappy 113]

Hi CVJ2.You've increased you FPS from 7 to 12 FPS? That's a 71% increase in FPS. Not bad.I had a somewhat similar setup with an original average of 28 FPS. with a Radeon64DDR. Added a Radeon 9500Pro, and now get an average fo 38FPS.. but now also with 4x antialiasing and 8x anistropic filtering! Now I turn my sliders to max, and set the ceiling at 24FPS... smooth. Note: I turn my ATC slider to zero.Something else besides the videocard may be holding back your system... and the stutters may indicate a problem with background tasks, or even a sound card that needs it's 'quality' setting turned down a notch.Also, the 9700 and 9500 cards are Directx9 compliant, and should be using that, and the latest drivers from ATI.Dick

[Guest PaulL01]

>Hi CVJ2. >>You've increased you FPS from 7 to 12 FPS? That's a 71% >increase in FPS. Not bad. >>I had a somewhat similar setup with an original average of >28 FPS. with a Radeon64DDR. Added a Radeon 9500Pro, and now >get an average fo 38FPS.. but now also with 4x antialiasing >and 8x anistropic filtering! Now I turn my sliders to max, >and set the ceiling at 24FPS... smooth. >>Note: I turn my ATC slider to zero. >>Something else besides the videocard may be holding back >your system... and the stutters may indicate a problem with >background tasks, or even a sound card that needs it's >'quality' setting turned down a notch. >>Also, the 9700 and 9500 cards are Directx9 compliant, and >should be using that, and the latest drivers from ATI. >>>Dick Hi Dick,Nice possitive spin on the original post. :)It does sound like there might be some progs running in the back.">and now >get an average fo 38FPS"38 Average? Whole Cow dude! I dont see even 30 FPS average on highly tweaked 3.4Ghz(p4) Pro9700 equiped boxes including no ATC (as you have to have that off for Benching) unless I'm flying in the middle of nowhere.you surely must have other settings turned down...mesh,..shadows...effects..quage quality?DX9 is not a good reccomendation for this card or anyother card yet as applied to FS2k2 as the mipmaping is srewed as well as a few other things, unless image quality doesn't bother you. Best wait until DX9a is out. It seems the more they fix the more they break. :-roll

[Guest SD Sim]

Hey Paul,That's quite the little speed demon of a box you got there. It's good to see you back in the forums - hadn't seen you for quite some time. BTW - how are you cooling that XP? It must get pretty toasty in there ;)

[Guest]

>38 Average? Whole Cow dude! I dont see even 30 FPS average on highly>tweaked 3.4Ghz(p4) Pro9700 equiped boxes including no ATC (as you >have to have that off for Benching) unless I'm flying in the middle>of nowhere.>you surely must have other settings turned >down...mesh,..shadows...effects..quage quality?I've only had a chance to make a couple of short flights in my new system, but "in the middle of nowhere" I'm easily seeing two, three, or four times that. That doesn't seem out of line to me, except in the densest areas, with a lot of AI.FWIW, I compared my 3DMark scores with the scores of all P4, Pro9700 equipped published 3DMark scores in the CPU range of 2500 Mhz to 2900 Mhz, to get an idea of the real "performance rating" of my new box. The average P4 speed of the 10 scores just above and just below mine was about 2850 Mhz. So my 2100+, o'clocked to a nominal 2600+, seems to be performing more like a 2800 Mhz P4. Not bad for a $500 upgrade from a 1.4 Athlon T-Bird and Voodoo5 5500 -- and that includes mobo, processor, memory, and video card! Is there any kind of rigorous benchmarking scenario for FS2002? Used to be for earlier versions of FS (Bruce Wilson's FSBench). I think there are lots of places not out in the middle of nowhere where the new system would average 38 FPS or more. I would expect it to fall down into the 20's with dense scenery and AI, but I seeming some phenomenal numbers, and it would be nice to quantify them rigorously.-Basil

[Guest PaulL01]

>Hey Paul, >>That's quite the little speed demon of a box you got there. >It's good to see you back in the forums - hadn't seen you >for quite some time. Thanks David.:-beerchug BTW - how are you cooling that XP? It >must get pretty toasty in there ;) These new Thoroughbred-B's run fairly cool to begin with, but if you want to O/C and dont like noise or want to use water get one of these:http://www.frontiernet.net/~pleatzaw/images/bottom.jpgIt is an Alpha PAL8045T and is still among the best made.I am using it with a quiet Panaflo 80x80mm fan with this O/C I run all day at 36C and 42C max load in a 26C/78F environment.And these temps will only go down as I slowly lower the Vcore under stability testing.Take care. :)

[Guest]

__________________________________________Is there any kind of rigorous benchmarking __________________________________________I think that any accurate benchmarking would be impossible in FS2k2. Reason being is that as the framerate increases, the amount of scenery and objects being drawn on the ground decreases. At 20, it's almost impossible to sit a cessna down anywhere without hitting a building or a tree. Except maybe in a desert out west somewhere. I lock my frames at 20 and pretty much forget about framerates. Smoothness is all that really matters.

[Guest]

Try renaming the cfg file so that fs2002 will be forced to create a new one.

[Guest]

Hi Basil,"Is there any kind of rigorous benchmarking scenario for FS2002?"Well, sort of. The SFC FS Benchmarks:http://www.simflight.com/news5/modules.php...rticle&sid=1317Not exactly "rigorous", but its most likely what you're looking for.Take care,Elrond

[Guest PaulL01]

Hi Basil,>I've only had a chance to make a couple of short flights in >my new system, but "in the middle of nowhere" I'm easily >seeing two, three, or four times that. That doesn't seem >out of line to me, except in the densest areas, with a lot >of AI. I understand your feeling here Basil...But we are talking averages with AA/AF on with all maxed settings (cept ATC).I am no stranger to high FPS in FS2k2 the shot below is my old T-bird GF3 combo running 16x12 with 2xAA and 4xAF at -LOD.http://www.frontiernet.net/~pleatzaw/images/46_fps.jpgHigh FPS are one thing, high FPS averages are another. Use Fraps to count frames, turn up >ALL< the sliders, and fly around, you wont come close.The systems as I mention above that I have benched will do 3DMark into the 20k's and running clean with all sliders maxed will not average anywhere near 38, not Talking about blue sky flying either, for that one can average into the hundreds.>>FWIW, I compared my 3DMark scores with the scores of all P4, >Pro9700 equipped published 3DMark scores in the CPU range of >2500 Mhz to 2900 Mhz, to get an idea of the real >"performance rating" of my new box. The average P4 speed of >the 10 scores just above and just below mine was about 2850 >Mhz. So my 2100+, o'clocked to a nominal 2600+, seems to be >performing more like a 2800 Mhz P4. Not bad for a $500 >upgrade from a 1.4 Athlon T-Bird and Voodoo5 5500 -- and >that includes mobo, processor, memory, and video >card! Just remember high scores in 3dmark do not necessarily make a fast FS2k2 machine, the machine with the better CPU/memory subsystem has the upper hand over the Video advantaged cards that rank high in the 3Dmark scores, no slight against yer Hot Rod Basil.My next upgrade will be the ATI9700/9800-pro or the Gainward GFFX-ultra as their cooling solution is almost silent. >Is there any kind of rigorous benchmarking scenario for >FS2002? Used to be for earlier versions of FS (Bruce >Wilson's FSBench). I think there are lots of places not out >in the middle of nowhere where the new system would average >38 FPS or more. I would expect it to fall down into the >20's with dense scenery and AI, but I seeming some >phenomenal numbers, and it would be nice to quantify them >rigorously. I had used a canned (recorded) flight of the Chicago area for about the first few months after Fs2k2's release and did about hundred benchmarks using some 30 different machines, then I changed to a canned flight of the Las Vegas area as the Mesh and custom 3D models and textures offered a better total mix of what one can run into when flying in the more taxing areas of Fs2k2, I meant to upload it a while back but never got around to it, it is about 2mbs in size.