Is there any fact behind the whole vsync mouse lag issue?
I mean, I would love to just have a 60 refresh rate and vsync on. I see no possible need for anything greater, and honestly I've never noticed any mouse lag at all with vsync. From what I can understand vsync supposedly drops sections of mouse inputs in between screen draws or something because it forces everything to be done sequentially and skips everything outside of the sequence. I've also been told there might be a way to fix it. Any top-notch resources on this?
Now, about pmove_fixed. Is this considered unfair play? I mean, people use it for tricking right? Do they not use it in competitive play, tdm etc.? Are their any good descriptions of what this does, what it compensates for? I mean, it's supposed to give you the optimum jump height every time or something, right? Well, if you don't usually get that optimum with pmove_fixed 0, then what do you get instead? And this is considered a cheat by punkbuster, right?
Just looking for some facts and clarification, I've heard lots of uninformed opinions on both matters.
vsync and pmove_fixed questions
pmove_fixed isn't always used for tricking and never for competitive play. It is used for tricking with overbounces, yes, but it gives an insane speed boost though.
Q3's physics are framerate dependant (evaluated at every new frame i guess), unless you use pmove_fixed, which forces every frame to last as long as pmove_msec (defaults to 8 milliseconds). That equates to 125 fps and if your videocard renders less than that, things get choppy. You move quite a lot faster than without pmove_fixed at 125, which makes doing strafepads in trickjump practice maps a lot easier.
Btw, CPMa has replaced the pmove physics code and it's much more framerate independant than any VQ3 or any other mod. It has no magic com_maxfps numbers that give a speed boost (only some weird things happening at very low framerates).
About vsync affecting mouse input: i find the effect very noticeable. Perhaps the fact that i use my Razer Diamondback at 500 Hz USB polling rate makes it more noticeable. When i enable vsync (r_swapinterval 1), i experience an significant slowdown when i flick the mouse from left to right (wrist movement). The effective sensitivity goes down and the mouse just seems not responsive enough to pick up flick moves correctly (they're translated to smaller crosshair movements).
Q3's physics are framerate dependant (evaluated at every new frame i guess), unless you use pmove_fixed, which forces every frame to last as long as pmove_msec (defaults to 8 milliseconds). That equates to 125 fps and if your videocard renders less than that, things get choppy. You move quite a lot faster than without pmove_fixed at 125, which makes doing strafepads in trickjump practice maps a lot easier.
Btw, CPMa has replaced the pmove physics code and it's much more framerate independant than any VQ3 or any other mod. It has no magic com_maxfps numbers that give a speed boost (only some weird things happening at very low framerates).
About vsync affecting mouse input: i find the effect very noticeable. Perhaps the fact that i use my Razer Diamondback at 500 Hz USB polling rate makes it more noticeable. When i enable vsync (r_swapinterval 1), i experience an significant slowdown when i flick the mouse from left to right (wrist movement). The effective sensitivity goes down and the mouse just seems not responsive enough to pick up flick moves correctly (they're translated to smaller crosshair movements).
Oele did you also try r_finish for testing vsync? I know r_finish and r_swapinterval are independent of each other, perhaps r_finish can give you a different result?
And yeah, overall vsync enabled can give you worse performance in games, for me, it drops frames noticeably in every game I've tested it, so I always leave it off. (Via videocard control panel, 'vsync always off state'.
And yeah, overall vsync enabled can give you worse performance in games, for me, it drops frames noticeably in every game I've tested it, so I always leave it off. (Via videocard control panel, 'vsync always off state'.
pmove_fixed also drops frames in osp. When some people had it enabled on the server I frequent the game started to feel very jerky.
Like oeloe mentioned, it's a server setting anyways, in all the major mods like baseq3/osp/cpma, so you don't have to worry about it. I think most baseq3 servers have it off (but of course there is no quick way of knowing unless you type it to check) and osp/cpma can callvote it with pmove.
Like oeloe mentioned, it's a server setting anyways, in all the major mods like baseq3/osp/cpma, so you don't have to worry about it. I think most baseq3 servers have it off (but of course there is no quick way of knowing unless you type it to check) and osp/cpma can callvote it with pmove.
CPMA doesn't have the pmove code anymore so you can't vote for it either.dzjepp wrote:pmove_fixed also drops frames in osp. When some people had it enabled on the server I frequent the game started to feel very jerky.
Like oeloe mentioned, it's a server setting anyways, in all the major mods like baseq3/osp/cpma, so you don't have to worry about it. I think most baseq3 servers have it off (but of course there is no quick way of knowing unless you type it to check) and osp/cpma can callvote it with pmove.
Yes i did try r_finish but it doesn't seem to do anything (no visual effect anyway). r_swapinterval 1 just synchronizes framerate with monitor refresh rate to eliminate tearing. If you constantly get more FPS than your refresh rate, there shouldn't be any framedrops but most people can't have 120 (or 125) Hz at the screen resolution they want to use, so they can't have the 125 fps speed boost (not a problem in CPMA).
-
^misantropia^
- Posts: 4022
- Joined: Sat Mar 12, 2005 6:24 pm
-
black8ball
- Posts: 113
- Joined: Sat Jan 21, 2006 2:13 pm
