In the Expression maps dialog the switches can have a LENGTH % setting. If a switch is set to 25% and applied to an 8-note (or less) the playback lenght of the note is set to 25%, but when applied to longer note values the playback percentage is always MORE than 25% of the notated value (according to the playback grid) … is there a good reason for this?
I believe there’s a clamp on the amount by which these modifications are allowed to modify the total duration of the note. I’ll ask Paul to confirm.
I asked Paul about this, and he told me there’s different behaviour for whether a note is greater or less than a crotchet. The motivational use case is how the “normal” note length works with longer notes. If the normal note length is 90% then that means you want a small amount of detachment. So for notes <= quaver, they’ll play at 90% length. However, if you have a note tied over 4 bars, that would finish 2 beats early. So if you have long notes you don’t want the effect to be proportional to the length. Instead, it works out the shortening as a proportion of a crotchet length.
In the case of staccato-like quantities like 25% this doesn’t make a lot of sense because it’s unusual (in most idioms of music, at any rate) to mark a note longer than a quarter note with a staccato articulation.
Daniel, just a thought.
You already have a clever algorithm to adjust note spacing based on duration. Could the same (similar) not also be applied to played length? So for a staccato minim it might play 98%, whereas a staccato semiquaver might play 50%. Similarly a tied note over 4 bars might play 99.9%. There’s probably a curve, much like your dynamic curve.
(FWIW I encounter staccato minims frequently in standard repertoire. Usually they just intend separation. How much separation depends on the underlying tempo)
Daniel, I hope you can revisit this whole topic thoroughly and consider implementing the means to change only notes of certain durations for the purpose of helping swing playback without affecting all quavers and semiquavers. at all. And another fully independent setting to govern shortening the last note under a slur. Please do let me know when you’ve had a chance to absorb all the things in those several postings I made earlier.