dude dont cap framerate, dont use freesync/g-sync

com_maxfps 250

I use to split my cfg into two parts and create my own qz/rep file and put it on read only, exe my cfg b4 each game and vid_restart. Also after every game would in_mouse 0/in_mouse -1 so my mouse movement always was the same

Damn you got a nice setup. Avoid the laggy mess known as win 10

yeah, i7-9700k is really nice!

/s

"What is the most stable NVIDIA driver for windows 7 to play Quake Live on. the populair 441.41? "

It wouldnt make any difference what you use. The game is so old i doubt anything is changing that effects Quakelive. If you try 3 or 4 different version and still have a problem its not the driver.

Drivers that existed before the 2080 released are probably best for your card, beyond that i dont see it making a diffence.

Stripped driver (whatever that is) wont do anything either. Just custom install (driver and hdmi only) is all you need.

Lots of people get weird stutters. Its generally the games fault nothing more so dont go nutters over your system as a problem.

Game got a bunch of updates (especially the update that gave 250fps and caused fps to lock at multiples - certain numbers) that caused new stutters.

The game has to run at certain multiples. Capping at 138 actually gives 142. Using rivatuner probably messes with it harder cos it cant jump to 142 so then its gonna behave like 125 and do 63 maxpackets, even if you still get 138fps on screen

The only thing ive noticed at 144hz is that fps 126 is smoother than fps 125 or 250.

The game has a slight stutter at 125 and 250, maybe its just the way 125 goes into 144 without vsync. Some people cant see it but I just think they are somewhat blind.

You can do things like 138 and 126 fps, but in the end the games going to run internally at 125, 142, 166, 200 or 250 and there is a slight stutter. If its not full blown locking up its probably normal.

tl:dr - this game isnt actually smooth so forget it.

Maybe you just somehow notice inherent stutter of running at different fps thant monitor hz? have you tried 125 fps with gsync/freesync on 144hz? I don't see anything at 250fps/240hz, but I know there are 10 "stuttering" frames per second.

" inherent stutter of running at different fps thant monitor hz? "

Yes. That or more or less what I attribute to causing it.

However, its not quite so simple because I used this monitor @ 144hz playing at 125fps for a long time and there was no stutter. It was at the 250fps update that I suddenly had a small stutter at any of the desirable framerates (125, 142, 166, 200, 250).

Its not even a very bad stutter. Cant really notice it unless I stop and look for it but it is there. Only way I really notice it is to strafe left right infront of an archway (dm6 quad archways etc) and I can just see the edge of the wall visually break apart slightly. Its a kind of vertical tearing (down the screen as opposed to horizontal vsync across screen type tearing) and also see pixels jump 4 or 5 pixels to the side by the next frame when it should really only be 2 or 3.

I can make it butter smooth by just typing com_maxfps 126 thats how im easily making this comparison. (edit: should point out that setting 126 actually gives me 142... thats the kind of weirdness that makes it hard to fix the problem, not to mention setting 142 stutters, but setting 126 and still getting 142 does not stutter)

Anway its fine. For long time I played at 126 (downside being that causes 63 maxpackets) but now I just play at 250 to get the 125 maxpackets. I dont consider the difference to affect my playing, its just a slight visual annoyance if I look hard at it.

It could be that 126fps is double 63maxpackets, while 125 is not.

As for gsync/freesync. I have one of the early benq 144hz so it does not suport those so best I have is just vsync off.

Anyway just trying to put down what I knew about it because I thought OP was barking up the wrong tree trying to fix it with the nvidia drivers.

com_maxfps in Q3 (and probably QL as well) unfortunately does not give you control down to an exact number of frames.

It basically just controls the minimum number of milliseconds between frames, which is an integer equal to

1000 / com_maxfps

with the fractional part removed.

so
1000 / 125 = 8, giving you 125 FPS.
1000 / 126 = 7, giving you 142.9 FPS.
1000 / 142 = 7, giving you 142.9 FPS.

So one would expect 126 and 142 to give the same result. Perhaps, com_maxfps ties into some new stuff in QL. Maybe it is com_idlesleep related, which is in QL but not Q3, and also well known to cause stuttering issues.

If OP is trying to keep FPS under 144 for g-sync, then 142.9 is probably too close. Normally, 140 or 141 is recommended. Probably best to just stick with 125. It's also better for max packets.

Yea i was not sure how those fps were being caluculated so thanks for explaining.

Dont have the gaming pc setup right now but maybe sometime ill see if I can do anything better with it. I think I tried 143 already, maybe i could cap at 142.9 but with the way maxpackets works I dont think there is any real solutions.

Maybe when I get newer screen it will go away lol. Have had the problem over several systems using same monitor however. Interesting that my gen1 i5 didnt need the idlesleep command but my newer system did because it was lagging hard, but after that the slight jitter persists on both systems.

ohwell.

Do you happen to know if com_maxfps 125 gives 125 maxpackets? for some reason I thought it was half or maybe the halving that only applies to the higher framerates like 166 and 250?

com_maxfps sets the maximum. It doesn't guarantee that you get that all the time. To get constant 125 packets per second, you need constant 125 or 250 FPS.

You get suboptimal performance between 125 and 250 FPS.

Frames that are 7ms apart is the worst case for anything over 125 FPS. That results in one packet sent every 14ms (2 frames)---so at 142.9 FPS, you get 71.45 packets per second.

Anything less than 7ms will still result in a packet sent every second frame, but they get sent more frequently because two frames take less time.

It's 20 year old game. just set com_maxfps 250, busywait 1 and it's fine. Do you think rapha uses "stripped down win 7" to get this last edge over his opponents? Or tox, or clawz? I've got similar hw + win 10 and 240hz monitor and ql runs great, no problems whatsoever. I also played on some old laptop with intel graphics and fps stutter before that and my accuracies/elo didn't even budge.

I think it has to do with the NVIDA drivers cause my win 7 is stripped down naked with no windows update etc.

Welcome to the botnet. Unless you bought this setup just to play quakelive and don't do anything else on it... Don't fall for this "voodoo gaming".