[Date Prev][Date Next]   [Thread Prev][Thread Next]   [Date Index][Thread Index][Author Index]

Re: MIDI problems ???

>If you want to learn some more about some midi alternatives and possible
>future musical networking technologies try these:
>the ZIPI home page (actually the "what happened to zipi home page"):
>  http://cnmat.cnmat.berkeley.edu/ZIPI/

I still disagree with you, Kim.  The good parts
of ZIPI are its support for per-note effects and
its performance; as I said before, transmit MIDI
over some faster physical link, and multiplex
it for more channels, and you've got all that.
(Just send every note on its own channel.)
[Ok, ZIPI's approach to pitches is a lot better.]

Most of the rest of ZIPI seems poorly thought
out to me.  Odd/Even harmonic content?  Isn't
that a little overly specific?  In general,
ZIPI seemed obsessed with what I could characterize
as: measure some quality at the controller, transmit
that quality over the network, allow the receiver
to determine how to reproduce that quality.

That sounds totally _wrong_ to me.  In reality,
different sound sources are going to have different
abilities.  You _have_ to remap from one quality
on the controller to a different quality on a sound
source.  In fact, this can be an enormous source
of creativity.  Given that you have to remap, and
given that every controller provides special things,
why even waste your time on "odd/even harmonic
content"?  Do like MIDI continuous controllers--
a large set of entirely unnamed "qualities", and
then manufacturers evolve over time a common set
of _controllable_ properties.  The drive to "name"
the qualities comes from the sound _sources_, not
the controllers.  Then the controller measures
the qualities in the performance and determines
how to remap that onto the known set of things.

As far as I know, this works in MIDI.  Of course
MIDI is broken in terms of per-note things, and
continuous versus just-at-note-start things; I
just mean the notion of which-side-of-the-system
is responsible for "defining" what qualities of
sounds are "interesting".  But none of the ZIPI docs
seemed to provide a rationale for shifting
this remapping onto the receiver, away from the

I didn't reply to your other mail on this subject 'cause
I think we're off-topic, but while I'm here, I strongly
disagree with your definition of "modern synthesis
technology".  The two mail-order catalogs I have
handy seem to have one or two "analog modelling"
synths, and all the rest are sample-playback.
Keyboard's NAMM report lists quite a few faux-analog
synths, but I don't remember seeing any of the "hot
synthesis techniques" you've listed (analog is hardly
new, it's just the digital emulation of it that is).  Those
things may be the bleeding edge in sound technology, and if
that's what you really meant, that's fine.  I thought you
were saying the average everyday "modern synth" had all
this capability nobody was touching because of MIDI.  If
you're saying instead that MIDI is crippling the ability
to do newer next-generation synths... well, you're just
bemoaning the fact that technological innovation is
difficult in this industry, and I still tend to think
that's not 'cause of MIDI, but because of the economic
incentive to stick to the things everybody already knows
(which we've already talked about to death as well).