live midi looping

[Fsksync@aol.com]

I thought I saw (back in the late 80's probably) a product that was a really inexpensive midi/audio codec. I'd love to explore the limits of what could be done with something crude like that- I'm sure it would suggest some new tangents. If I could afford a Kyma, I'd be too distracted to imagine midi loopers, I think  8-)  at least for a year or so. (send donations to my paypal account)

Even if one got something like this going, without some kind of intelligent algorythm to reduce velocity values and then delete notes when they fall below a certain velocity or volume value the system would overload and do "something" eventually (that would be interesting).

If one were to record and delay dc control voltages (I know of at least two examples of functioning units based on this principle for driving analog synths back in pre-midi days) wouldn't that necessarily exclude polyphonic midi, or am I missing something. If one had to multiplex voltages to get polyphony, I don't know how that would compare to converting midi to audio in complexity.


Best,

Tim F 



In a message dated 7/27/03 2:13:56 PM Pacific Daylight Time, dennis@mail.worldserver.com writes:

> I'm posting these together since they're related.
>
> Cheap (not really) MIDI Looper:
> In theory, you could convert particular MIDI messages (say a Note-On/-Off)
> into particular tones (like FSK, frequency shift keying), then loop the
> output tones in a "conventional" looper like the EDP, Repeater, Boomerang,
> or whatever.  You would need a decoder at the output, too, or course.
>
> Decoding will take some time ("latency") since frequency recognition cannot
> be instaneous.  Even better would be to loop a DC voltage level (which
> encodes the MIDI message), but then you would need a DC-coupled looper.
>
> One of my examples in the Looper Construction Kit for Kyma is such a MIDI
> looper.  Kyma loopers are "DC-coupled" (since they are in software) so I
> don't need to do the FSK thing.  I should be able to build a FSK MIDI
> encoder/decoder using Kyma, so you could loop MIDI with an Kyma/EDP combo.
> I assume you can build something similar with MAX/MSP.  But that seems
> rather silly (except for prototyping) since it's easier to loop MIDI
> directly with Kyma or MAX.
>
> A primary difference between MIDI looping and audio looping is that a MIDI
> controlled tone (a "note") is typically described by two events seperated in
> time (Note-On and Note-Off).  You probably want to link the two events
> (manipulating them together) so that you won't have "stuck" notes and such
> as you overdub, insert, decay/fade, chop, dice, and puree the loop.
>
> The technique I've described above does not link the Note-On and Note-Off
> events, so I wouldn't count it as a "real" MIDI looper, though it *does*
> work remarkably well.
>
> And very intriguing thoughts, Mark, on features for a MIDI looper!  Please
> post more thoughts!
>
> Dennis Leas
> -----------
> dennis@mail.worldserver.com
>
>
>
> -----Original Message-----
> From: Fsksync@aol.com [mailto:Fsksync@aol.com]
>
> Does anyone know of a device that encoded midi into an audio tone so it
> could be recorded to an audio cassette (or other recorder) and then decoded
> back to midi, making the cassette machine a crude sequencer? It seems I
> dimly remember such a (cheap!) device flashing by for a millisecond. If
> something like that was available, it might be possible to do "something"
> with a bunch of them and a bunch of digital delays (you couldn't use any
> audio feedback on the delays, though) and midi mergers. Such a thing would
> be basic and crude, no frills, but one might get the most basic live midi
> looping (tm) function going.
>
>
> -----Original Message-----
> From: Jesse Ray Lucas [mailto:jlucas@neoprimitive.net]
>
>
> That's a really cool idea though...
>
>
> >At 03:29 PM 7/26/2003, Dennis W. Leas wrote:
> >>Is the audio path AC or DC coupled?  In particular, will the EDP
> >>record/playback a suitably scaled CV for instance.  (Twisted thought, I
> >>know.)
> >
> >it is ac coupled. you can't loop a dc voltage.
> >
> >kim