Lowering graphics is a good idea, as it increases your FPS, but what about resolution? I use 1152x864 myself, but I’ve heard that really good CS players use a somewhat low resolution - a different game, but still. Anyone got some input on this?
Another question: Why would I want to cap my FPS? I thought the higher the better. Please explain.
I run at 800x600 cause that’s all my video card and moniter will handle. That works fine. I can see everything, it runs smooth and it looks pretty good. People who “need” 1900x1200 or whatever it is are just prissy little girls.
set your res to the highest you can while still maintaining the framerate and quality you want.
Capping framerate is good because you can jump slightly farther at certain frame rates. It can also give your game a smoother feel, since various things are frame rate dependant. Having a steady 76 FPS may be easier to play consistantly than having it go from 76 to 300 depending where you stand. There may also be some other minor gains to be had in conjunction with other settings. For why the jumping thing is the way it is, see
http://ucguides.savagehelp.com/Quake3/FAQFPSJumps.html
To squash a few old legends before they get repeated:
setting your com_maxfps to a value higher than your video card normally produces will NOT give you better FPS in general. It just a limit. If your card cannot produce frames fast enough to reach the limit, the setting has no effect.
FPS does NOT control your rate of fire. There may be some second order effects like the jumping thing, but if there is, they are minor.
Are you sure on that? I very much get the impression that the engine checks the keyboard once per frame and hence if your FPS is lower than the firing rate of your weapon (which has to be pretty bad FPS) then you will fire slower than someone with higher FPS.
Are you sure on that? I very much get the impression that the engine checks the keyboard once per frame and hence if your FPS is lower than the firing rate of your weapon (which has to be pretty bad FPS) then you will fire slower than someone with higher FPS.[/quote]
Go somewhere where you get good FPS.
Set com_maxfps to 10
Empty your SMG, by holding the trigger down.
Now set com_maxfps to 100
Again, empty your SMG, by holding the trigger down.
The second time, does the SMG empty 10 times as fast ? Obviously not.
Therefore, FPS does not control your rate of fire.
BTW, I doubt very much the game works as you describe. More likely, it knows when you pulled the trigger, and fires at the correct rate until you release the trigger (for full auto weapons, of course). Even if it did work as you describe, the SMGs fire at about 360 rds/min, or 6 rds/sec. That means that to see the effect, you would have to get less than 6 FPS. 6 FPS = unplayable by any sensible definition. The ROF of the semi auto weapons is similarly limited, even if you pound the keys really fast.
The higher your resolution, the better will you be able to spot enemies in far distance.
Weapons are not fps dependent, although flamethrower seems serverside-fps dependent.
Is there btw a way to synchronise clients to sv_fps like with g_synchronousclients in RTCW?
I get about 30-40 FPS on average (higher sometimes), with most of my settings on either high or medium with a resolution of 1280x1024. I tried lowering that and lowering detail level and my FPS shot up…however, the quality of the graphics made it really difficult to play. I have bad eye sight already and trying to tell if that jaggy thing at the other side of the map is an enemy or a rock is much harder with those lower settings.
What’s that have to do with anything? Dunno, but for me, getting 90 FPS vs. 35/40 isn’t worth it.
Aritificially limiting the FPS with com_maxfps doesn’t count - even if you do /com_maxfps 1 that doesn’t make your computer really take 1 second to draw a frame.
I used to regularly play with < 6 FPS until I got my GeForce 4 
Agrado, if you set com_maxfps to 1, it WILL draw only one frame per second.
ROF has nothing to do with your FPS.
Bad ping can sometimes alter (very slightly) the way pistols/rifles fire, but never automatic guns.
no, AFAIK, it will just make it sit around and do nothing for the rest of the second, which is equivalent.
I used to regularly play with < 6 FPS until I got my GeForce 4
Uhhh, good for you ? 
I have heard that if your FPS is less then your cl_maxpackets, you will get less smooth gameplay. If that is true (I have no information one way or the other on that) that would seem to mean that the game only samples input and/or sends once per frame, so at the very low end there might be some influence.
He knows that. But there’s a difference between “one frame per second” and “take 1 second to draw”. Restricted with com_maxfps it draws the frame as fast as possible, and then twiddles thumbs until it’s time for the next frame - the lower FPS are simulated by waiting, not by throttling the CPU and GPU etc.
Now the question is, does it read user input during this delay?
Does any of this even matter?
He knows that. But there’s a difference between “one frame per second” and “take 1 second to draw”. Restricted with com_maxfps it draws the frame as fast as possible, and then twiddles thumbs until it’s time for the next frame - the lower FPS are simulated by waiting, not by throttling the CPU and GPU etc.[/quote]
Absolutely… correct!
It mattered to me when I was playing at 4-5 fps 
It doesn’t matter to me now except as intellectual curiosity :blah:
I understood that was true in Quake2, but not in Quake3 or other games based on it (like RTCW / ET). Just to be on the safe side however, I have cl_maxpackets 76 and com_maxfps 76 in my config. 
I understood that was true in Quake2, but not in Quake3 or other games based on it (like RTCW / ET). Just to be on the safe side however, I have cl_maxpackets 76 and com_maxfps 76 in my config. :)[/quote]
if i increase cl_maxpackets from lets say 30 to 85 it will result in ping 50 instead of ping 30. So lower cl_maxpackets seems to give the smoother gameplay
Man it really puts me to shame when i hear all this talk about 76 fps, 300 fps when i have never gotten more then 30 :\ And that is with all the settings as low as they go, including the resolution.
Here are my system specs
Proccesor: 1.4 ghz intel celeron
ram: 512mb
bus: pci
vid card: nvidia geforce 4 mx 420
vid ram: 64mb
Os: Windows XP
When im in closed spaces i’ll get a good 30 fps, but when im looking out from the command post room (in oasis) to the city ruins and in BATTERY i get like 10 fps 
I would think that my system should have had no probs with this game. Is my system sucky? Or should my system be able to perfrom rather good and i might have some internal error?
I’d have thought you could do better than that, although not by huge amounts. The GF4MX is really a GF2 in disguise and not really that great a card for games, and the 1.4 Celeron isn’t that great a CPU either, so you’re not going to get spectacular results.
Try a little tweaking:-
seta cg_atmosphericeffects 0
seta cg_wolfparticles 0
seta r_picmip 3
seta r_lodbias 2
Those might help (you add / edit these lines in \etmain\profiles[i]yourname[/i]\etconfig.cfg).
Thanks for the replies.
Btw, what’s the difference between using ‘set’ and ‘seta’?
What does that do? I’ve been using the other’s you listed, but could couldn’t see any difference with r_lodbias0, 2, 100, or 2000 for that matter.
set = uses for now
seta = written to cfg
lodbias = google found:
OpenGL computes a texture level-of-detail parameter, called lambda
in the GL specification, that determines which mipmap levels and
their relative mipmap weights for use in mipmapped texture filtering.
This extension provides a means to bias the lambda computation
by a constant (signed) value. This bias can provide a way to blur
or pseudo-sharpen OpenGL's standard texture filtering.
This blurring or pseudo-sharpening may be useful for special effects
(such as depth-of-field effects) or image processing techniques
(where the mipmap levels act as pre-downsampled image versions).
On some implementations, increasing the texture lod bias may improve
texture filtering performance (at the cost of texture bluriness).
The extension mimics functionality found in Direct3D.
which totally clears that up. erm, maybe better try:
http://www.beyond3d.com/articles/gf4ti4600lodbias/index1.php
scoreboard ping is broken in every Q3 engine game, or at least it is if you mess with cl_maxpackets; notice how often in clanwars players sometimes appear to be on a LAN with totally identical pings for entire game, even though theyre using different ISPs, and even different country. Even if scoreboard ping was 100% accurate, does it make any sense that the trip time of one packet is the sole, or even primary explanation of gameplay smoothness? if you set rate to 5000 you’ll probably also see lower pings, but if you think 10 less ping can compensate for transmitting 1/3 the amount of data…
So ‘set’ has nothing to do in cfg - because then it would be permanent? Only when you’re typing them in manually in the console. Am I right?