korman wrote: ↑
Thu Sep 19, 2019 11:33 am
OK, so I tried to make a practice track for Dundungbe in Percussion studio. In that program you can pick either a binary grid: each 8th note dot divided into 4 (smallest subdivision is a 32-note), or if you press a Shift key ternary grid is available which divides 8th note by 6 (smallest subdivision is a 16th note triplet). So, the desired SML swing could be made quite hard by shifting second and third eight note by a 32-note (25/33/42 relationship), or a little bit softer by shifting by a 16th note triplet (28/33/39 relationship). However, neither version sounded quite right, so I'm thinking it must be the case that different instruments (and perhaps even different beats) are swung differently.
Has anyone been able to produce a realistic sounding PercussionStudio file of dundunba rhythms? Can you share the example file or at least give me a clue?
Similar to what ternarizator did some years ago, I analyzed two instructional recordings Niels Fleurke's "Ritmes uit Baro" and Marten Schepers "Rhythmes de Guinee - Dundunbas". There you have tracks of just the kensedeni against the click or on-the-beat accompaniment. Sure, a village recording would be better, but it's much easier to do the analysis when the sound is so clear.
Average for Fleurke is 26%/36%/38% and Schepers 26%/39%/36%, so, yes, the swing is a rather pronounced S-L-L one
I also recently bought the Zoom R8 recording device to record myself playing and to create backing tracks to practice solo. It has a built in sequencer with the smallest resolution 48 ticks per beat. So, with a little tinkering I was able to make kensedeni loop according to these specs, and - voila - it sounds good to me (I prefer the first version with last subdivision longer than the second one).