Overclock.net › Forums › Components › Mice › The unspecific Mouse Thread
New Posts  All Forums:Forum Nav:

The unspecific Mouse Thread - Page 53

post #521 of 2027
Nah, in motion the quantization actually causes worse problems than the jitter. It's literally the same thing as shrinking the grid to hide the noise. Now you have grid ripple.
Edited by wareya - 2/5/16 at 9:21pm
post #522 of 2027
Quote:
Originally Posted by wareya View Post

Nah, in motion the quantization actually causes worse problems than the jitter. It's literally the same thing as shrinking the grid to hide the noise. Now you have grid ripple.

You are thinking only of sidways ripples and not the rectilinear parts of the jitter though.
post #523 of 2027
If quantization hides jitter, then it's introducing stronger warping artefacts than was the jitter.

I don't understand what weird conception you have about something that quantization fixes without adding anything else, but it's certainly wrong.
post #524 of 2027
What about for 2D games?
post #525 of 2027
The right DPI to get your sensitivity. Maybe double with 50% scaling in the OS.

But osu uses subpixel cursor precision, AFAIK, but don't quite me on that.
post #526 of 2027
Quote:
Originally Posted by wareya View Post


I don't understand what weird conception you have about something that quantization fixes without adding anything else

I did not say that.
post #527 of 2027
Hmmm I'm going to ask an osu dev about it tmw. Can you explain to me what subpixel cursor precision is?
post #528 of 2027
Placing the cursor not aligned to pixels on the screen.
Quote:
Originally Posted by daniel0731ex View Post

I did not say that.
I didn't understand what you were trying to say. I thought this was what you were talking about.

Let's say you're trying to track someone and the warping from jitter is moving your crosshair left and right as you track. If you lower the DPI, you still get warping, because the sensor locks to particular speeds in its efforts to stay aligned to the grid.

http://i.imgur.com/6et91sj.png

Math tells us that as long as the jitter is visible, the jitter will overtake the quantization. Once the quantization overtakes the jitter, the quantization's own warping overtakes the strength of the jitter's warping. This quantization is different from quantizing the raw deltas (it's sort of error-diffuse), but it still has the same rules.
Edited by wareya - 2/5/16 at 10:20pm
post #529 of 2027
So if the pixel that represent the cursor's centre on the display is not aligned with the actual centre of the cursor? inb4 newly discovered cause of mouse drift
post #530 of 2027
Think like FPSs where the game logic doesn't really have pixels.

Mouse drift is because jitter in the sensor's internal DPI space.
New Posts  All Forums:Forum Nav:
  Return Home
  Back to Forum: Mice
Overclock.net › Forums › Components › Mice › The unspecific Mouse Thread