In a message dated 2/2/99 11:41:42 AM, Thomas Hudson <
thudson@...>
wrote:
>
>Okay, I've given this some thought. I would a sequencer that
>supported downloadable "personalities." For example, I may
>want to read from the ADC and select a stage according to
>voltage, i.e. voltage control stage select. I may want to
>play with random numbers or markov chains, e.g. set up
>probabilities for each stage based on previous stage, etc.
>
>One interesting idea is to be able to download code via midi.
>I nibblize my program and pack it in a sysex message and when
>sent to the device it gets flashed to ram. Imagine people
>posting new "personalities" to the web. Could really drive
>sales.
I really liked Thomas's notion of "personalities," although I'm not sure what
he means by it. Below is a slightly edited list of things I sent to Paul after
his initial request for input on the MOAS.
I believe that a sequencer's features (maybe more than any other part of a
modular) may lend itself to one or another type of music, and my preference in
electronic music, has been towards probabalistic types of music within well
defined parameters. I also have no basic idea of what you're thinking about
for implementation, so I'll just make a wish list of features.
1) Quantization of pitch, at least chromatic; better chromatic, diatonic, or
pentatonic, best programmable tempered system. This could be defeatable since
continuous control is nice, but it does get tedious to have to tune the beast
after a while.
2) I, long ago, became convinced that while it was good to have one large
sequencer, it was better to have several small sequencers (many just pulse
sequencers) interacting with each other. It seems like your MOAS could act
like several small sequencers running in parallel.
3) I rarely like to deal with long sequences, 16 stages is enough to deal with
in tuning an analog. It would be better to allow for several separate phrases
which could be indexed into directly (a variation on 2 above). Preferably the
lengths of these could be independent of one another which would be good for
polyrhythms, for example. I'm also thinking of independent lists for sequences
and durations for similar reasons.
4) At a minimum, I would want these inputs and outputs in an analog sequencer:
a basic pulse out for each stage, and the following inputs: up/down, hold,
reset, random select. It would also be nice to have a stage select (or,
referring to 2 & 3 above, a phrase or sequence select), or multiple
programmable reset points.
5) I've also been thinking of two different but related ideas involving
voltage capturing.
a) A digital sequencer which will take a sequence of voltages from the
(analog) input and store it digitally, and output the sequence one note per
clock pulse.
b) A voltage capturing router. Same idea as above but having multiple inputs,
which can be routed to different outputs. The idea is to have several voltage
inputs from say a voltage pedal, a keyboard, theremin, and other controllers
which could be digitally stored and then routed to say a filter or panner or
other vc input. A set and forget kind of device.
>Paul Schreiber wrote:
>> Here is a brief outline of the direction I am going:
>>
>> 1) It's modular
>> 2) Has full MIDI control, but all outputs are analog
>> 3) Patch memory
>> 4) I can be BIG. I mean ∗REALLY BIG∗ - 256 stages x 64 rows
>> 5) The modules are tied to a local LAN than reconfigures the modules, ie
>> Plug-&-Play.
>> 6) This being MOTM and all, no cheesy pots. We're talking optical encoders,
>> baby!
>> 7) uP will be in C programming, with "open source code" posted on the
>> website.
I'm really interested in him explaining the details about these ideas!
John B.
Apologies for hogging the MOTM onelist bandwidth!