RealNC wrote:For reference: I friend of mine is using a 125Hz mouse (a basic Microsoft Intellimouse) on an Asus VG248QE with LB enabled. It looks just fine, so I would imagine that 500Hz would have no problems whatsoever.
Yes, it looks fine under specific cases, e.g. running Quake Live, where 125fps works well, or using fps_max very close to the mouse poll rate.
However, the big issue with 125Hz mice with LightBoost is that it causes more visible microstutters in other use cases, when you vary the variables, fixed-framecap, variable-framerate, VSYNC OFF and VSYNC ON situations, GSYNC variable refresh rate, and situations where you're trying to optimize framerate == refreshrate == stroberate. This is where you begin to see microstutter differences.
It looks fine until tried against a good Razer DeathAdder or good Logitech, especially during fast left/right movements. When turning left/right, my games look this smooth:
http://www.testufo.com/photo (when viewed as a 120fps animation in a stutter free browser such as Chrome running Aero mode), and my mouse movements become as smooth as keyboard strafe left/right.
125Hz mouse cannot do that in all use cases.
Note: Make sure your friend has "mouse smoothing" turned off, because software based mouse smoothing makes it look as stutterfree as a 1000Hz mouse, at the expense of adding input lag.
During 1000 pixels/second motion, a 125Hz mouse can alias positions to a previous or next frame, which can mean microstutter amplitude of (1000/125) = 8 pixel amplitude of microstutter. This, I see VERY easily when I reduce my Razer DeathAdder or Logitech G9x down to 125Hz, and try to turn left/right with LightBoost enabled. Turning is no longer as smooth as TestUFO motion.
At 500Hz, you have a (1000/500) = 2 pixel amplitude of microstutter at 1000 pixels/second. At 1000Hz, you have only (1000/1000) = 1 pixel amplitude of microstutter at 1000 pixels/second. Of course, fast-flick 180-degrees can go, say, 5000 pixels/second horizontal panning speed, so at 500Hz, that's a 10 pixel amplitude of microstutter and at 1000Hz, a 5 pixel amplitude of microstutter. If you've eliminated all other motion clarity weak links, and you are fast at tracking your eyes, this actually begins to become a human noticeable factor.
If you've got a stuttery game, you won't tell the difference. But if you're playing perfect 120fps@120Hz VSYNC ON, with motion nirvana, the mouse becomes the fully weak link, and the 500Hz-vs-1000Hz difference becomes amplified during the point of zero blur/zero tearing/zero stutter/framerate==Hz. Most competitive players will do VSYNC OFF though, to reduce input lag even at the expense of tearing and stutters (Though you can get very close to the motion clarity of VSYNC ON using various tricks such as fps_max 360 in an attempt to keep frames more synchronized to the refresh rate, while keeping tearing reasonably in check.)
This is a big problem when frame rates vary, so you've got distance between the frame presentation times versus mouse poll times, which creates the microstutters when mouse smoothing is turned off. 1000Hz is also a way of giving hardware based mouse smoothing without the input lag of software-based mouse smoothing.
That said, human factors apply. Likewise, not everyone sees 60Hz vs 120Hz well, not everyone sees tearing, not everyone sees microstutter. I do, though.