Oh dear, I wasn't aware that negative p3 was monophonic. I hereby amend my
proposal for a 1 year duration of p3 until note off.

I still vote for an extension of MIDI. Another field, such as "note ID",
could take care of the problems and opportunities raised by Dobson.

Alternatively, however, one could turn to Java and develop new network
protocols for music control; these could become regular Csound "line events"
with all of the power that implies. MIDI controllers could be adapted to
feed this new network protocol. Actually, now that I think about it, that
sounds very attractive, and it is even easier to do than MIDI itself. The
content of the message would simply be a snatch of Csound score - one or
more lines - to be evaluated with their p3 relative to the current time, as
in the existing line in event. It would be quite simple, using Java, to
receive MIDI input and parse it out into Csound statements. The pushdown
stack thing I mentioned could be used together with the fractional p1 "tie"
feature to provide realtime note IDs.

-----Original Message-----
From: Richard Dobson 
To: csound@maths.ex.ac.uk 
Date: Tuesday, July 27, 1999 7:03 PM
Subject: Re: sequencer


>A negative p3 does not only make the note hold indefinitely, it makes it
>monophonic - it is the primary mechanism for creating legato
>instruments. A zero value also already has a special meaning (see the
>manual). Pfields are floating-point, so cannot be 'unsigned' in the way
>an integer can.
>
>It is also possible to run simultaneous and independent notes on the
>same instrument, something that is only possible under MIDI by using the
>sustain pedal, with no control thereafter. More to the point, thanks to
>the duration parameter the user has independent control of the life of
>each of these notes, something, again, which MIDI does not support. The
>classic example of this is additive synthesis using multiple instances
>of a single sine-wave instrument.
>
>The negative p3 instantiates a monophonic voice able to support
>portamento and amplitude ramping, essential for most legato styles. This
>is probably the area in which Csound and MIDI diverge the most. A
>'Cquencer' which facilitated the programming of simultaneous independent
>legato streams would be formidable indeed.
>
>Portamento, in particular, requires knowledge of the target note. There
>is a mechanism in Csound (the 'next-pfield' instruction) which supports
>this; it could conceivably have been added to the MIDI spec as another
>Channel message had it been deemed important enough, but as MIDI was not
>developed to support ~composition~, but merely to enable keyboards to
>control each other, it was left out (assuming anyone thought about it at
>all!).
>
>I see MIDI as essentially a sort of simple telemetry for keyboard
>controls, using wires. It has a grammar, perhaps even syntax, but not
>semantics. Choir and string sounds(etc) invariably sound very false
>under MIDI, however good the sounds may be in themselves, as there is no
>way to program the independent polyphonic voice-leading which is
>idiomatic for them. Pitchbend needs to be associated with notes, not
>just with Channels.
>
>This is one of the many problems with MIDI  - it's grammar is not voice
>based, but controller-based. I want to be able to play four voices, each
>starting at middle C, and each performing a portamento to a different
>pitch. MIDI thinks they are still somehow middle C, but I think they are
>four voices which have changed pitch smoothly. The grammar for Csound is
>the opposite, even if most people end up writing fixed-pitch notes, most
>of the time.
>
>It is all very well adding little tweaks to MIDI here and there,  but
>this fundamental limitation of MIDI is set to remain, and nothing short
>of reinvention (ZIPI?) will obviate it. You are given MONO mode, and
>different Channels, and a few instruments have been made which can be
>configured as sixteen monophonic voices, but without portamento
>messages, and sub-idents for multiple voices on one 'note', the scope
>remains very limited.
>
>My though for a Csound sequencer then is something which (somehow) not
>only supports the arbitrary pfield system, but also full polyphonic
>legato per instrument. With careful design, even this could be handled
>in real-time, though almost certainly not via MIDI.
>
>Richard Dobson
>
>Paul Barton-Davis wrote:
>>
>> In message <001d01bed823$365752a0$79d496c0@Realizer.ngt.sungard.com>you
write:
>> >The p3 issue is a real one but it is not insuperable. The solution is
for
>> >realtime score events to have p3 set negative (infinite duration) until
a
>>
>> zero is more desirable, since it allows duration to be represented by
>> an unsigned value, and thus have a larger range.
>>
>
>--
>Test your DAW with my Soundcard Attrition Page!
>http://wkweb5.cableinet.co.uk/rwd (LU: 6th July 1999)
>CDP: http://www.bath.ac.uk/~masjpf/CDP/CDP.htm (LU: 14th June 1999)