Internally, a voice is not played if its mix volume is zero. This is in order to not tie up that voice from being used if it's not being heard. When the voice doesn't play, it also doesn't send MIDI notes. In my opinion, it's not important enough to change this, which would require internally changing the software voice architecture to play the voice mostly but not completely, because it could introduce new bugs in a system that I think is already too complex. Fortunately the workaround is easy: simply set the mix volume to 1 or higher.
ok, i understand the problem. thanks for the answer.
may i ask how is it that the voice allocation software works? is it possible to program a sound with the lowest priority possible, but with a level of 1 or higher for the sequence to be sent out, so we could load this sound to the external sequencer pad and not have voice steal for it due to low priority?. That way we would be able to use an external synth effectively as a 7th fixed voice for the tempest, hence my inquiring.
may be it is programmed so if a sound gets voice stealed the sequence wont run? then it's just not possible and an external synth will just replace one of the tempest voices still limited to 6 at a time total.
or may be the priority for the sequencer out pad is maxed out so it always steal a voice?
thanks again for your answer.