Originally Posted by prava
My point being: its impossible to get a 99th percentile of 40ms AND and average of 60ms unless you go with negative numbers, that is
. Its simply not possible to say that the 99th of the distribution sits at 40ms or below yet the average distribution sits at 16.6ms
. Impossible. IM-PO-SSI-BLE. Mathematically it makes sense because you can work with negative data... but you can't have negative frametimes and, thus, none of this makes any sense.
I think you might be wrong in your interpretation of 99th percentile in this context. Here it is the time that 99% of frames are less than
. So it should be completely possible to have an average of 16ms and a 99th time of 40ms. It would look like this: 12, 15, 12, 40, 15, 40, 13, 12, 65, 12, 13, 12, 9, 11, 13, 16, 12, 9, 10, 9
Your average is 16ms, and you throw out the 65ms outlier, so your 99th time is 40ms because 99% of the frames are less than 40ms. Obviously you'd need a bigger data set, but you get the point.
And I'm assuming the bolded part is a typo and you meant to say 60 FPS and not 60ms. Or did you do your math based on 60ms instead of converting the 60 FPS to a 16ms average? You wrote it both ways there.Edited by Forceman - 12/13/12 at 11:16am