Originally Posted by fateswarm
Can you direct me to specific URLs proving the supposed superiority of G400? I only found a fan-based benchmark test from 2006 using an LP player with an LP cloth mat. One of the mice apparently had the sensor G400 has.
Well the G400 and other ADNS 3090 mice definitely aren't perfect but they aren't that far from it when it comes to 1:1 tracking although they might have some jitter at higher DPI settings (like at 3600 or 4000 DPI or whatever their highest setting is). When it comes to mice with the Avago 9500 or 9800 sensor (like the Logitech G700, G9x or the SteelSeries Xai as shown in the picture) the tracking looks like this:
On the x-axis you can see the speed at which the mouse is moved and on the y-axis you can see how many counts it tracked. That means that at for example 2000 dpi it should register 2000 counts when you move the mouse for 1 inch (obviously ^^). Ideally you want the mouse to always register the same amount of movement independent at which speed you move the mouse. In the graph that means you ideally want a perfectly straight line at 100% without deviations.
Looking at the graph above you can first off recognize the inconsistent tracking as it deviates from the 100% a lot (sometimes it registers more and sometimes it registers less counts as you can see). For example reading from the graph above (lets still keep the 2000 dpi setting as an example) if you move the mouse for just 1 inch at a speed of 0.65 metres/s it will register roughly 5% less counts than it should, which would result in about 1900 counts.
On the other hand at just a little faster hand movement of about 0.85 metres/s it will register roughly 5% more counts than it should, resulting in about 2100 counts. So all in all you could say that at 2000dpi there's a deviation of 200 counts when you move the mouse for 1 inch at two actually very similar speeds of 0.65 and 0.85m/s.
Also if you check the trend of the graph it is slightly rising and after 1 m/s it's always over the ideal 100% line at a peak of around 108% which is precisely the amount of positive acceleration a lot of people experience with this sensor. Also this is not a worst case scenario (the opposite) because on a lot of cloth mousepads things can get even worse where you don't experience positive acceleration at higher speeds but negative acceleration (the graph falling noticeably below the 100% line).
G400 and similar mice have much less inconsistency and therefore no noticeable acceleration.Edited by MONVMENTVM - 5/18/13 at 4:03pm