Synthetic Sensor Testing
This section is an evolving endeavour to improve the objectivity of mouse reviews here at Play3r, it shouldn’t necessarily be used to compare mice apples to apples but it should help us get a better feel for the quality of sensor implementation. The programs being used here are homebrew, so these results should always be taken with a pinch of salt but I think it is essential to supplement the subjective content of my reviews with some (slightly) more scientific testing.
All my testing is done on an OcUK Mega Mat XXL
Enotus Mouse Test v 0.1.4
Enotus mouse test can give us a few bits of sensor information such as DPI and polling rate. The methodology used in this benchmark is to always run it at the maximum DPI the mouse supports and also at 1000Hz.
Right starting off with Enotus mouse test and we can see why I ask you to take a pinch of salt with the results of these homebrew programs. First off the DPI was spot on so no complaints there, the speed was also fine though as ever we won’t know whether the sensor is failing or not so take the result lightly.
Therein is also the first issue that gets thrown up, the software seems to think the polling rate of the mouse capped out below 800Hz. I can’t really explain this but I know it is an error because Mouse Movement Recorder is more reliable and says otherwise but we’ll get onto that in a minute. The precision result is a test of acceleration and amusingly, despite numerous attempts using the standard methodology I couldn’t get this test to complete; not because the mouse performed badly, but because there was so little acceleration the program didn’t register it!
Finally the smoothness test looks pretty dire but I have an explanation for this in the conclusion of this section so read on for that.
Mouse Movement Recorder
This benchmark gives us a direct feed of sensor reading versus pointer movement and it also gives us an approximated polling rate. This benchmark is done on an open and run basis, but we run it at both 500 and 1000Hz to compare the two.
As you can see the account of polling rate here is variable and the single error is representative but not concerning. The variation in the frame rate here looks worse than it is, when mice keep a dead straight bead on the polling rate it tends to be because the firmware is smoothing inputs, not because the sensor is performing flawlessly.
As you can see polling definitely was more stable at 500Hz but the error is again representative. Overall there is nothing to worry about on the polling rate side of things, the results show us windows is getting a good raw reading from the sensor.
Paint jitter testing
This is test is a bit less scientific, but drawing lines in paint gives a good visual representation of any jitter present. The methodology here is to perform the test at 1600 DPI, 3200 DPI, and finally the maximum DPI of the particular mouse which gives us means to keep results more consistent between optical and laser sensors. All tests are run at 1000Hz.
As you can see there is no noticeable jitter, and perceived jitter is actually due to the fact that I struggled with the small size of the mouse and not that there is a fault.
Again no real jitter here that isn’t my fault alone.
Same story here too, it looks worse than it is; curse my large mitts.
So, as you can see the synthetic tests look a bit concerning in places on the face of things, namely in Enotus mouse test where the Hz didn’t seem right and also in Enotus where the smoothness was rated as ‘bad’. However, as we can see from Mouse Movement Recorder the report rate is fine, hence I believe Enotus moust test was just misbehaving there (please refer back to the foreword at the top of the page) so that isn’t anything to worry about. When it comes to the smoothness test I believe this is due to the fact that the mouse is so small in my ape hands that I struggle to grip and point it smoothly hence why the paint jitter test also looks bad at first glance.
Therein is the answer really, using a mouse that is too small for my large hands in a grip I am less accustomed to has resulted in black and white tests throwing some flags up which is not surprising at all when you think about it. Lets move on to some real world testing and see if it confirms my theory that it is me that is to blame for the dodgy synthetic results, not the sensor.