Sensor

Cursor movement is tracked by one of the best mouse sensors available, the PixArt PMW3360. This sensor is as raw and responsive as it gets, which also depends on the firmware, of course. The ATOM's firmware isn't exactly perfect, and the input lag section will give a detailed explanation on that. Overall, I had no issues with the sensor; I experienced no motion skipping, unusual jittering or any other malfunctions.

The available resolutions are 400, 800, 1200, 1600, 2400, 3200, and 6400 CPI, and you can set polling rates of 125, 500, and 1000 Hz. There are no custom lift-off distance options, and the default value is set to about 1 DVD in height (~1.2 mm/0.047"), but it barely tracks on that (the measurement was done on a plain black cloth mouse pad).

Paint Test

There is no jitter on the reasonable CPI steps, or unwanted angle snapping and measurable (or audible) sensor lens rattle.

CPI Divergence

CPI divergence on the ATOM is relatively high, and it linearly increases with the resolution. You might need to adjust your in-game sensitivity accordingly if you come from a mouse that has more exact values.

Perfect Control Speed

The nominal tracking speed of the HERO sensor is ridiculously high at over 400 IPS, or 10 meters per second. There is absolutely no chance of hitting that speed while in-game, or anywhere within reasonable bounds.

Speed Related Accuracy Variance

This test shows the sensor's accuracy at different speeds. You can see me doing a fast swipe to the right before I slowly slide the mouse back to its original position. Any displacement is almost entirely caused by human error in this test.

Polling Rate

All set values seem to be perfectly stable as there are no suspicious outliers after multiple tests; the ones seen are mere measurement errors.

Input Lag

I measured +2 ms of input lag on and below the 1600 CPI step before it jumped up to +4 ms above 1600 CPI and stayed there. The PMW3360 has the common issue of smoothing above 2000 CPI, but input lag below that value is quite strange, and I think it is a firmware-related issue that could be fixed. I redid the tests numerous times with different mice, on different USB ports and PCs, and the results were always the same. On the graphs above, the control subject is a Ninox Venator.