Don't know where to start
@dmxdex2020: I'll try to tell you when mine arrives
When I tried it, it felt really good because of its low profile.
The G500 has a hugely off-center sensor position. Be sure to test if that's your style before you get it.
@outerspace: Don't think 3.5m/s is a problem if you wanna
do it. However, anything above 1.5-2m/s is far beyond my natural gaming
First off, may I ask what is your source of the M6900 having the Pixart PAW3305DK-H? So far, the ever only mentioning I have seen about this is Bullveyr's post
in this thread. I even tried asking Gigabyte directly but they wouldn't tell me saying it is some kind of secret
A sensor being stuck at 1000Hz sucks imho. If not in general then because of the higher CPU load. However, that kinda approves of it being the Pixart since I have read of the Pyra
having the same property.
A MS-paint picture of the reported jitter would also be nice.About the graphs:
You should have a look (in the .log file) how far the sample points are apart timewise (e.g. 0.001s, 0.002s, 0.008s). This might tell you if the 1000Hz are real. Would also be nice if the graphs showed the time axis.
What I find most disturbing is the huge negative acceleration (clipping) of the M6900, and it's so ironic since it's supposed to be 1000Hz. Even if it still had an 8bit-sensor, it should do way better @1000Hz (like the old MS-mice). One could argue that 1.5m/s is enough for gaming if you are not too far into low-sense but it just doesn't make any sense.
I'd be really interested to see comparable graphs with a Pyra or Kova[+]. Manufacturers should just include these graphs in their marketing About the MX310:
433Hz is a strange polling rate. I assume you manually overclocked the mouse. What value did you set it to? There are cases where, if you overclocked your mouse too high, it would report with irregular polling rate. So if you looked at a tool only reporting the maximum or average poling rate you might get something like 433Hz whereas actually the polling rate is jumping around. This would also match the regular drops to zero seen in your graphs. You should definitely have a go with 250Hz and see if the drops to zero disappear.
The M6900's drop in the flick-shot is probably sensor related or tracking was lost due to a slight unintentional lift-off.
Generally, your desk is possibly not the optimal surface for testing your sensors for flawless behaviour