Hi Folks, I was reading through this thread, and agreeing with Mark and Jack with regards to the comments on natural variation in playing style. It got me thinking about how to visualize this. I’m not going to hijack this thread or anything, but I wanted to share visuals of what the variation of natural e-drumming might look like. I queued up a rock track and e-drummed along to it in SD3 while recording. Then I exported the midi and wrote some Python to analyze the velocity variations on a per-instrument basis. Given the genre of music, the snare, hi-hat and kick were dominant, and I was trying to be somewhat light-handed with the snare even with the louder hits. Attached you can see a histogram of snare velocity values (the blue histogram) across the whole track (again these are my midi velocity values recorded while I was playing along – x-axis is snare velocity and y-axis is the number of hits at that velocity). The histogram appears to have two peaks, one around 30 and one around 50, but you can see the spread is roughly 5-85.
However, what is more interesting, and relevant to this thread, is the red histogram. This shows the velocity *change* from one hit to the next across the entire track (for the snare, on the x-axis, with the number of hits at that delta value on the y-axis). Basically my variation in trigger velocity, whether intentional (ie transitioning from backbeat to ghost) or just my own inconsistency in maintaining consistent velocity (which is the point I’m trying to make here). Notice the note-to-note velocity variation between hits, while most in the +/- 25 range, can vary up to +/- 50 (these larger ones are undoubtedly intentional transitions). Interestingly I see a peak around +50, and not -50, which seems to imply to me I was consistent with my return to the backbeat from ghost/light notes, but not the other way around.
When I listen back to this drum track, it is not an award-winner, but it is obviously a live capture and not programmed. I would argue that (and there are research publications that back this up) listening to music with natural percussion variation is more pleasing to the ear than programmed percussion lacking variation, or even programmed “random” variations that slot into bins. That is not to say one is good and the other is bad (often discomfort in the listener is an effect to pursue), but that they are indeed discernible.
Anyway, hope this is of interest to someone, and happy holidays to all!
3
Thanked by: LeanderL, Bear-Faced Cow and drumjack52
