Looper's Delight Archive Top (Search)
Date Index
Thread Index
Author Index
Looper's Delight Home
Mailing List Info

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

Re[2]: MIDI problems ???



     Kim... after all you've said below, why aren't you mentioning ZIPI? 
Is 
     that completely dead as a doornail? Did you have any role in it's r+d?
     
     -Miko


>> >From:       matthew hahn [SMTP:esker@mail.utexas.edu]
>
>> >>And of course the other problem lies between the front and the end. My
>old
>> >>friend midi. If you actively tried, you would not be able to design a
>worse
>> >>networking protocol for musicial instrument control than midi. That
>alone
>> >>has hampered a lot of potential innovation in musical controllers.
>
>
>i dunno - maybe i'm shooting too low, but midi serves me well - i find my
>casio gtr triggers excellent-ly, and my midi mitigator does a plethora of
>cool stuff, obidiently, like prog change, note on/off info, chord sends,
>start/stops, etc. I get to change programs on my ADA MP1, quadraverb, 
>Kawai
>synth and even analog old pedals thru my rockman octopus - all useful,
>timesaving stuff for what i do -

well, here's an example: On your guitar controller, if you play an
artificial harmonic, does the synth respond in some appropriate way? When
you apply different types of mute techniques, does anything change in the
synth patch? What if you pick near the bridge, or use the flesh of your
thumb instead of a pick. Or pick near the bridge. Anything? nope. Most of
the expressive control over a sound that a guitar gives you is lost, and
about all you have left is pitch and 128 volume levels. Now realize: modern
synthesizers are quite capable of providing huge and subtle varieties over
the timbre of a sound. Modern sensor tehcnology is readily able to detect
all of the things you are doing on a guitar string. But when you try to
send all of that information from your musical and expressive controller to
your musical and expressive synth, you totally swamp midi. It can't handle
all the control data. It's incredibly slow, even by standards at the time
it was created in 1981. Also, it's "description language" for describing
musical events is terribly shortsighted and limited. (it was only ever
intended for a piano key trigger to start a simple synth sound). The right
words just aren't there to describe most types of musical events.

In the world of modern instrument design, midi is the bottleneck, and has
been for years. That is why you so rarely see innovative new electronic
instruments, and why the ones you do see are so rarely satisfying. Midi has
been the immovable object in the middle.

for simple uses, midi is fine. It serves some purposes well. But in many
ways, its a thing where you don't know what you are missing, because you've
never been able to try it.

As far as a network design, midi is just stupid. It's unidirectional, the
topology leaves you with at mass of wires everywhere, addressing events on
any particular device is clumsy and difficult, and you always end up with a
device B that can't talk to device A without rewiring everything. The list
goes on, and there are others who have expressed it much better than I. I
think there are places on the web you can find if you care.

As far as loop specific problems, we would ideally want the capability to
transmit audio data along with control information, in real time. We would
also want the ability to easily define a loop oriented control language,
rather than forcing keyboard commands to do it or spending 6 years in
committee meetings at the MMA. These things are beyond midi, and the music
industry is not able to get itself organized enough to do anything about
it. It doesn't really matter now, because Microsoft is happily changing
everything to suit their needs, and I don't expect anybody will be able to
do anything about that. I'm holding out hope that firewire will finally
come into vogue, and at least speed things up a bit. It's the only
technology out there that has both the bandwidth for mass data transfers
and low enough latency for useful control activity, without being
overwhelmingly expensive. any year now.....

kim


______________________________________________________________________
Kim Flint                   | Looper's Delight
kflint@annihilist.com       | http://www.annihilist.com/loop/loop.html
http://www.annihilist.com/  | Loopers-Delight-request@annihilist.com