For what it is worth, some synths employ a lowest volume algorithm instead of least recently used. Sometimes a voice may be sounding even if it is the oldest and strange voice stealing effects can be heard (we have all heard it). This is of course based pon and tied to an output envelope generator of some type. Stealing the softest signal is much less noticable but much harder to implement.
-----Original Message----- From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] On Behalf Of Fons Adriaensen Sent: Sunday, January 18, 2004 6:20 PM To: The Linux Audio Developers' Mailing List Subject: Re: [linux-audio-dev] Project: modular synth editor On Sun, Jan 18, 2004 at 07:00:19PM -0500, Dave Robillard wrote: > Ack. Speaking as a musician and not a hacker for a moment, I don't > think a new note should ever, ever, ever be discarded! It's the simplest way to tell the user that (s)he needs more voices :-) > But you did say optionally... I just had to say that. :) Reusing the > oldest voice should definitely be the default behaviour. Agreed 100% -- FA