Originally Posted by Carnaged
Thanks for the information, maybe I am getting this wrong but does that mean that if I put the windows step to 5/11 that it would be closer to 1800 cpi which is what I would normally use but without the lag due to the 2300cpi default of the mouse?
That would be 0.75*2300 = 1725 CPI. However, the 5/11 step is inconsistent in tracking and will not feel right. I would use 4/11. 5/11 doesn't uniformly discard counts from the mouse. For example, if 1s are counts and 6/11 gives: 11111111 for 8 counts, 5/11 gives: 11101110 discarding every 4th count which wont' feel right. 4/11 will give you: 10101010 skipping every other count, effectively halving your CPI. I would stick with 2300 CPI and 4/11 with 1000 Hz, then adjust your sensitivity higher numerically by about 56% in your game from what you were using at 1800 CPI. (1800/1150 = 0.56).
If your game supports raw input, USE it. There's no reason not to use the full 2300 CPI of the mouse and then reduce your in game sensitivity and bypass windows entirely. I know you're used to a certain CPI but in reality a higher CPI and a lower in game sensitivity only helps you here, as it increases the granularity of movements in FPS games. I actually play on 5670 CPI on my Sensei right now at a very low in game sensitivity to give me 60cm/360 just because the mouse performs exactly the same at all CPI steps, so I just use the highest native one with raw input.