Comment #33 on issue 687 by [email protected]: Enhancement: inequal MIDI quantization of equal durations (swing, rubato)
http://code.google.com/p/lilypond/issues/detail?id=687
Well, I suppose so ;-) It looks very promising, although I haven't looked at it yet in detail. I would like to know whether it takes into account all the points I raised in comments #6 and #15, and in particular whether it is capable of coping with the Viennese waltz lilt example. Another question is whether this interface allows the degree of swing to be easily parametrized, and set either manually, or as a function of the current tempo. In jazz, swing tends to become more pronounced (i.e. closer to an exact 2:1 triplet) at lower tempi and straighter at higher tempi (e.g. 3:2, 4:3, 5:4 etc.), but it is typically the reverse for samba - a lower tempo might use something like 6:5:5:6 and a faster one might use something like 3:2:2:3. I don't think it's ideal to force the average lilypond user to have to do these mathematical calculations themselves in order to obtain the right degree of swing. They should just be able to say "make it jazzy" and then specify a number on a sliding scale. That's why I proposed the quadruplets syntax in comment #6: the 4th integer in each quadruplet which represents the degree of displacement can be trivially multiplied by a weighting factor to give this sliding scale.
