[MIDI Microphone?????]

[Harry Bissell]

Harry Bissell writes: Well Dave, I can steer you AWAY from the wrong path...
128 bandpass filters would be a lot of hardware. Done analog you would never
get the frequency stability. 1/2 step (music) is too close of an interval. And
there would be time domain problems (the filters would not detect instantly.
Switched capacitor filters would be a better choice, but there are still
problems.
    The "bandpass" idea can only work for a sinewive, any harmonics would
excite higher filters (2f, 3f, 4f etc.) You would then have to detect the
lowest filter that is activated. And you need to control the input amplitude.
    For a theremin, you have a major advantage. The oscillator will run at all
times if you tap into the circuit before the VCA. Then there are two good
techniques. One is to digitally divide the input wave into 4 sections
(just use flip-flops). During two of the steps, you make a linear ramp that
increases in voltage. 3rd step. you sample and hold (until the next step
three). 4th step you reset the ramp and wait for step 1 again. This will give
an output voltage that is updated every other cycle. It will be proportional
to the period of the wave, if you want volts per octave you need an
exponential converter (check the archives for one of these). Then you can
convert voltage to midi. (not trivial)
    Two of these ramps could be used to increase the update time if it is a
problem.
    Other solution is to use a high-speed counter and start count at one zero
cross and stop at the next, read the resulting number with a micro and look-up
the appropriate midi note. The faster the clock, the more accurate the
conversion (but the bigger the counter) this gets even closer to the midi
note.
    Or the invention above (patent attempt at humor) embodied as a blazing
fast micro that counts its internal clock cycles from zero cross to zero
cross, and outputs midi note.
    Search the archives for "trauntonium" (sp?) there was discussion a couple
of months back regarding another problem, how to handle a continuous frequency
(a'la theremin) bu using a combo of note on-off and pitch-bend messages. This
part ain't trivial either. If you just want to quantize to the nearest note,
forget this part.
    If you have access to a fax machine, e-mail me your number and I can send
you some schems. I don't have a scanner.
    My wife send her greetings (she's from Canada too)
:-) Harry


dsolursh at georgian.net (dave solursh) wrote:
Hello from canada,  I just finished my first real DIY project, a theramin
from Paia.  I painted it green because it reflects my experience with the
workings inside the black boxes I have been using for years.  I have had
great luck and it is working just fine, but it is time to take this
theramin to the next level.  I would like to use it as a MIDI controler and
have not been able to find any designs for a CV to MIDI converter.  I know
there are a lot of aspects to consider going from CV to MIDI but I am
looking for something pretty crude for the moment.  I have heard through
the grape vine that something exists called a MIDI MICROPHONE that converts
CV to MIDI and that may do just fine.   I can't seem to find any literature
on this MIDI Mic and was wondering if any of you could lead me in the right
direction, or help me design something that could convert a Theramin CV to
a MIDI note on and off message on seperate notes over at least on octave.

        I have one idea, and maby someone could help me start my quest.  is
there an electrical circuit that would be turned on (conduct electricity)
when a certain frequency was introduced into it.  I figure if you have 128
of these (tuned to approx. each note) each triggering an eight bit MIDI
note # and velocity # It could be a CV to MIDI.   Please remember I am
GREEN to this,   Little Dave.



____________________________________________________________________
Get your own FREE, personal Netscape WebMail account today at http://webmail.netscape.com.