Any kind of script - any kind at all - when run on my PC, will always result in some horrible tearing. I'd hate to use V-Sync, as it gives a noticeable delay between controller input and screen output.

Monitor Refresh is 75Hz. I can play Team Fortress 2/ Burnout/ Starcraft and other games with disabled V-Sync and 75Hz fine... No tearing there.

Is there some code I'm missing that would help solve this tearing issue, or is it just my machine?

fps_max?
but if it gets out of sync you'll have tearing aswell.

...and I'm not 100% sure if this solves the input lag.

Using triple buffering should help reduce the amount of tearing:

http://www.conitec.net/beta/ad3d_triplebuffer.htm

@Kartoffel: The input lag is only noticeable with V-Sync (vertical sync) on. I leave it off to keep input latency at bay, and I don't have problems with tearing with V-Sync off in most games.

Sorry guys; fps_max didn't help, and triple buffering made it worse. oO

Whats your framerate when this happens?

What did triple buffering made worse? The lag or the tearing?

^... yes; + does you fps drop before this happens or do you always have max-fps?

also, you said your monitor has a refresh rate of 75. Did you set the right windows-settings for this?

@Uhrwerk: Triple buffering made the screen tearing worse...

Framerate is always at max, i.e. above 75 fps. F11 display tells me my framerate can go as fast as 250 fps. Windows setting is correct, 75 fps.

Triple Buffering will only help if your framerate is below your monitors refreshrate, not above. So this is one of the cases where you should use v-sync. The controller delay should be no longer than 13 milliseconds (That is already the worst possible case). I doubt this is really noticeable.

Does enabling v-sync give you a constant 75 fps?

Enabling v-sync from the graphics card control panel does give me a constant 75 fps.

I should probably mention now that I just noticed using fps_max does eliminate a lot of tearing, but somehow it alters my physics by raising everything's max speeds higher...

Exploring with V-sync on, no tearing was observed (as expected), but it reveals that my ship's movement is in actuality very 'jerky'. I'm using pXent functions with the ackphysx.h library.

Using time_step atm. Is it a suspect? Time_step vs time_frame much?

time_step is the correct variable to use in many scenarios. Use time_frame only when you want to ignore time_factor. If your movement is jerky at about 75 fps chances are pretty high there is timing fault somewhere. But thats just a guessing game with this few information...