I posted about this to the synth-diy list several years ago. I also
believe I made similar comments when Synthesis Technology announced
plans to release a MIDI to CV converter. I don't remember getting
much of a response.
I completely agree that MIDI is outdated and that we need a "fast
event information and high resolution continuous control". However,
I do not think we need a standard so much as we need a product, and I
am surprised that with all of these companies offering analogue synth
modules that no one is offering such a product. Nor do I have any
idea why, as I am sure it would be extremely popular.
I have read that people have used ADAT's as CV recorders by removing
filter caps on their inputs, but I have never tried that myself, and
I do not know how well it would work. Regardless, the issue is
getting from a computer sequencer to CV.
There are two issues here, hardware and software.
I do not know about any cheap audio interfaces that can pass DC.
Perhaps they can. Nor can I respond to what Paul S. might have said
regarding their resolution in the sub-sonic range. As far as I can
remember, the Roland SH-101 used a 6-bit DA driving a CEM3340, and I
never had any trouble keeping mine in tune. Also, afaik, my Kenton
Pro-4 is only 12-bit and seems to work fine. So 16 or 24 converters
-- which are much less expensive now than they were several years ago
-- would have more than enough voltage resolution for accurate pitch
control. Both USB and Firewire have much greater bandwidth than MIDI.
So let's say the hardware is a device that has a USB input and
several CV/gate/trigger outputs. The software issue is that DAW's
are not designed to to generate DC voltages, so recognizing the
hardware as an audio interface would not work. However, all DAW's
support virtual instrument plugins in AU, VST, or some other format.
So by using a plugin the DAW would simply see the hardware as just
another MIDI instrument.
Not only would the resulting CV be as accurate as a MIDI to CV
converter, and allow for much better implementation of pitch bend and
portmanteau, but the triggers and gates generated would have much
better timing than MIDI. Imho, that alone would be worth the price
of admission. Unlike MIDI, it could be able play several several
events exactly at the same time.
I would recommend that the specs for the hardware be publicly
available, and that the manufacturer should consider offering some
sort of SDK to encourage third-party development. Stand-alone
applications could be developed with computer sequencing abilities
that go far beyond the limitations of MIDI. Not only could the
hardware to generate LFO's, envelopes, clocks, and continuously
sweeping voltages, but given the ability to produce frequencies in
the audio range, it can be used as a synthesizer itself. By adding
audio inputs to the software, it could work as envelope follower or
pitch to CV converter.
Further, the hardware would have a multitude of other uses. Software
could be written to use it as a function generator. People could
write applications to use it to control all sorts of things --
lighting, robotics, special effects, home theatre systems, internal
combustion engines, whatever.
On 3/3/06, coyoteous put forth:
>MIDI has been obsolete for more than a decade, but is so implanted
>that it will probably be here long after we are gone. Remember MIDI2
>or ZIPI? MLAN is/was also supposed to be a MIDI killer. What we need
>is a fast standardized duplex protocol for event information and
>high resolution continuous control. Home brew "ghost electronics" +
>MIDI are about as close as it gets or may ever get. I'd still like
>to see a USB 2.0 or (IEEE1394/Firewire) to (and
>from) CV module. But why limit to sub-sonics? Some cheap audio
>interfaces pass DC on one end and/or the other, but as Paul S. has
>pointed out in the past, the resolution in the sub-sonic range is
>probably inadequate for precise 1V/oct pitch control. Planetarium
>laser show controllers have used modified ADATs and sound cards to
>accomplish a flavor of this for a long time. I would think there
>would be some off-the-shelf industrial controller technology that
>would do the job, but in my limited searching I haven't come up with
>anything.