Midi event handling and voice handling when using tail-off in Synth

Hi forum.

I’ve come across the following issue:

If I implement a sort of tail-off (fade out), for a synth, which to my understanding means that a note-off will not immediately clear the corresponding voice, I would have to provide twice the number of voices for a number of sounds that can play “simultaneously”.

That is because, through testing, I observed that if the synth has as many voices as the maximum number of sounds in it, and I loop a midi clip with more than half that number of notes from start to end, the midi event block will handle the note off and on per midi note (as they are all in the same subblock). That means that eventually some of the notes will not be played because the engaged voices have not yet been freed (because of the tail off delay).

The obvious solution, as I mentioned is to double the number of voices. I would like some input on this, to see how this is generally handled.

Thank you!