Initialization and deintialization sequences

When MuseScore commences to play a score, it first sends:

a. Basic setup information for each staff (program number, bank select, etc.)
b. Then, for each staff, on the MIDI channel for that staff:
b1. A CC64-0 (Sustain pedal OFF)
b2. Note OFF messages for all MIDI pitches, 0-127 (128 messages)
b3. A CC123=0 (All notes off).
b4. A Pitch Bend = 64 (no bend)

At the end of play, the sequence in b1-4 above is sent. along with some other things.

Considering just the start of play sequence in item (b), for a 10-staff score, that would come to 1310 MIDI messages before the first note messages of the score proper.

I will first discuss the need for such a thing (focusing on the start of play situation). Seemingly, the 128 Note OFF messages are intended to silence any notes that might have been "stuck on" in the synthesizer (what an organist would call "ciphers"). Evidently, there is not faith that the ALL Note OFF commands would take care of that.

The scenario is in any case not very realistic. If my synthesizer had somehow gotten stuck sounding a note or several, I would probably not look to playing a score to solve it. Rather, it would be very good to have on the MuseScore control panel a "MIDI panic button", which would send such a "clearing" sequence, presumably on every one of the 16 possible channels. (Note that the current scheme will only clear stuck notes on channels used by the current score; the stuck notes might in fact not have been on any of those channels, and so would remain stuck.)

So, is this arrangement a bit silly but harmless? Sadly not. Many MIDI devices here do not take well to such a flood of MIDI messages (arriving at the fastest rate MuseScore can send them). The misbehavior this causes is often complex and tricky, and I will not bore the reader here with tedious war stories. Some devices consider such a flood to be implausible, and conclude that there has improvidently come into existence some sort of "MIDI feedback path", which the device disables by blocking the path altogether.

In certain cases, when an external MIDI device is nominated, but we have its output muted, listening instead to the MuseScore indigenous synthesizer, the anomalous behavior afflicts the play as heard through the indigenous synthesizer. I do not know how this can be, but it is.

So, overall, I refresh my urging that the initialization (and initialization) sequence play of MuseScore be revised and reconsidered.

Best regards,

Doug