[gmpi] Re: Drilling down into MIDI->GMPI conversion

>> The real issue is that there is no *standard* way of doing 
>> these things.  So mainstream sequencers don't support them.
>> 
>> We need a standard way to do per-voice controllers.
>
>Just FYI, the MMA provides a forum for companies to propose
>such things and develop them in cooperation. No one has ever
>made such a proposal to my knowledge... whether that is
>because there aren't (and never have been) any MMA members
>who agree with your statement, or whether it is just a matter
>of everyone having more important things to do, I don't know.

when you build a piece of hardware, its very difficult to provide
per-voice control elements as part of the control surface. how are you
ever going to pack them all in? "bank switching" isn't really an
viable option. moroever, even if you could build them, somehow, the user
doesn't have enough body parts to manipulate them :)

when you build software that models hardware, as has been the norm in
the non-academic electronic music world for the last 5 years or so
(the period in which real-time synthesis has become truly viable), you
tend to follow the hardware model. and even if you don't, you still
tend to follow the idea that a person is going to control all the
parameters in real time using some combination of a mouse, the
keyboard and/or some (MIDI :) hardware control surface. my K5000S
(again!) has 16 parameter knobs - more than i can control, and more
than most hardware synthesizers. yet they are still applied to every
audible voice, for obvious reasons.

however, in the world derived from music-N languages (notable Csound
and MusicKit, but several others too), where real-time synthesis
wasn't part of their original design space, it made perfect sense to
allow per-voice control of multiple voice parameters, evolving over
time. the user never manipulated them in real time - they used a text
editor to create a description of the parameter changes. some of the
major works of early computer music were all created in this way, and
there are still plenty of sound designers who operate in analagous
ways today.

however, contemporary sequencers represent the fusion of both these
worlds. they can be used to drive virtual instruments (doing realtime
synthesis in a variety of ways), but they are free of the limitations
implied by realtime human control. today's MIDI-driven virtual
instruments use the MIDI model in which a NoteOn creates a new voice,
but there is no deterministic way to identify or access "the voice" so
created. this means that the sequencer can't control the voice in the
way, that say, a Csound orchestra file could. all it can do is to use
some ad-hoc mapping of various MIDI messages to the voice, hoping that
the virtual instrument will map them to the correct voice (cf. mike
berry's explanation of LIFO/FIFO issue for voice killing).

i believe that this is the core of the limitations that michael gogins
and others have alluded to.

--p


----------------------------------------------------------------------
Generalized Music Plugin Interface (GMPI) public discussion list
Participation in this list is contingent upon your abiding by the
following rules:  Please stay on topic.  You are responsible for your own
words.  Please respect your fellow subscribers.  Please do not
redistribute anyone else's words without their permission.

Archive: http://www.freelists.org/archives/gmpi
Email gmpi-request@xxxxxxxxxxxxx w/ subject "unsubscribe" to unsubscribe

Other related posts: