I've been trying to avoid posting this because i'm probably the only one with this problem for a long time and i tried all possible solutions i could think of, maybe there is none...

No matter what ping i have (28,...,100) i always have same feeling of huge latency so i doubt it's network lag.
I notice it more during fights, when i get hit screen seems to be very slow and delayed and mouse/keyboard tend to be alot less responsive.

For example, in DM2 i can't move well across the map because i keep hitting walls when i try to do a tight curve, just don't know if its the mouse/keyboard or video being displayed delayed or incorrect timings.
Maybe what is being rendered isn't exactly what is happening so i think i'm seeing one thing and something else is occurring (leading me to think its input lag).

If anyone has any clue or explanation i would be eternally grateful. Posting all possible important information:


XP SP3
Nvidia ForceWare 175.16
Mouse software not installed
Ezquake 1.9 Stable (build 2253)


//Player Settings
rate "25000"


//FPS and EyeCandy Settings
cl_independentPhysics "0"
cl_maxfps "77"
cl_nolerp "0"


//Texture Settings
gl_anisotropy "16"
gl_ext_texture_compression "0"
gl_max_size "2048"
gl_texturemode "gl_linear_mipmap_linear"


//OpenGL Rendering
gl_clear "1"
gl_finish "0"
gl_ztrick "0"


//System Settings
sys_disableWinKeys "0"
sys_highpriority "0"
sys_inactiveSleep "1"
sys_yieldcpu "1"


//Video Settings
vid_colorbits "0"
vid_conheight "480"
vid_conwidth "640"
vid_customheight "800"
vid_customwidth "1280"
vid_depthbits "0"
vid_displayfrequency "75"
vid_glDriver "opengl32"
vid_mode "-1"
vid_vsync "0.0"


//Sound Settings
s_khz "44"
s_mixahead "0.1"
s_noextraupdate "1"
s_precache "1"
sys_inactiveSound "0"


//Input - Mouse
in_mouse "3"
m_filter "0"


//Network Settings
cl_predict_half "1"
cl_predict_players "1"


What does the "-nohwtimer" cmdline parameters actually does ??

Your vid_conwidth/vid_conheight and resolution ratios are drastically mismatched, which results in distorted geometry, your conres is 640*480 which is 4:3, and your resolution is 1280*800 which is 16:10. Change vid_conheight to 400

In this particular case, the fix applies (the player has widescreen monitor). However this does NOT apply to all situations, and since people can understand it wrong, more information has to be added.


CRT monitors:
Users normally strecth the resolution to fill all the screen area, hence the picture is always 4:3 on CRT monitors, regardless of the situation. With CRT monitors players do not need to make any additional changes, unless they intentionally do not stretch the resolution to fill the whole screen area. Screenshots, however, are affected unless your picture viewer doesn't scale the picture to fill the screen. This applies to QW, your desktop image will have broken geometry and other games might or might not be affected.


With TFT monitors it depends on the scaling mode: (examples concern 16:10 monitors and default 4:3 conwidth/height)

1:1 pixel mapping (do not scale): depending on the resolution, each pixel always fills one pixel of the screen, resulting in wrong geometry if not corrected by conwidth/height settings. 4:3 resolutions require no changes but 16:10 resolutions require geometry correction.

Aspect ratio scaling (fill the screen while keeping aspect):
4:3 resolutions always keep 4:3 scale, and this will not break geometry. 16:10 resolutions on the other hand will always break geometry and require correction.

Full scaling (stretch to fill the screen):
Will always break the geometry, regardless of the resolution, since it fills the whole screen which is 16:10.

Now that the resolution/geometry related stuff was made more complete, I have to ask what's the point of replying this half a year old (dead) thread, and not even replying to the question that were originally asked?

It's always nice to help people when they ask, but ignoring the question asked and telling something other instead isn't all that fun to the one asking. Unfortunately I can't give any insights for the original problem, and it seems noone else have had the same problem or solution for it.