[sdiy] FM

[Sebastian Kuehnl]

Hello all,

--- Grant Richter <grichter at asapnet.net> wrote:

[snip]

> I haven't actually heard anything musical done with a frequency
> shifter. I
> have serviced the Bode units and had a chance to play with them,
> but they
> essentially render any sound "clangorous" by changing the harmonic
> ratios.
> (IMHO) If there is some great musical quality about frequency
> shifters, I
> fail to understand it.

And I fail to understand what you mean :-) You would have to double
that impression with ringmodulators then?? The sound of a single
sideband is free of the interference of the two sidebands in a
ringmod sound. I'm trying to say that frequency shifting even a
simple sawtooth wave is as perfectly controllable and ~musical~ (are
woggle bug noises musical ? ;-) as filtering or nonlinearly shaping
the wave. It doesn't necessarily have to be a crass distortion
effect. It's only about compressing the harmonics. It can alter the
sound from barberpole phase shifting over subtile shift to radical
"clangering". Frequency shifting an entire melody line with a static
carrier pitch sounds boring indeed of course (but you can still use
that for static textures like in "space music"). However if the
frequency shifter can use an external carrier, or if the internal
carrier VCO has a CV input so it can track the modulator frequency,
the compression ratio of the harmonics will be the same at all
pitches; thus the "clangerous" quality becomes systematic; thus it's
a sound processing method equally or even more useful than others.
(Can someone summarize that longish rambling into a clearer phrase
please? ;-)

>From my understanding the biggest technical challenge with musical
frequency shifters is the precision of the phase delay networks or
dome filters. There seems to be no satisfying solution for delaying
subsonic carriers precisely enough to enable the through zero phasing
effect. To my knowledge only Juergen Haible and Chris MacDonald have
built frequency shifters with this broadband ability.

My own idea has been to replace the PDNs with wavetable readouts. If
the modulator and carrier each read a wavetable (like the Wiard
Waveform City or Blacet Miniwave) containing a 1/4 phase delayed
sawtooth wave, the output will be the required 90 degrees delayed
versions of the signals.

I'm just theoretisizing since I don't have the aforementioned
modules, so if someone tried this I would be very curious to hear the
results! To my understanding the samplerate & bitrate limitations of
the wavetables may introduce stepping to the barberpole effect as
well as noise to the final sound and thus may be no real replacements
for PDNs in terms of overall precision (only in terms of broadband
operation).

> It was probably once useful for "robotic" voices in the same way as
> vocoders
> were once popular for radio station IDs.

Funny you would mention voices. I found that frequency shifting is
interesting for speech synthesis. I have no analysis to
scientifically back this up but it seems that compressing
respectively expanding the harmonics of a vowel sound is enough to
morph it into any other vowel - quite similarly to altering the
bandpass filter frequencies in the comb filter producing the vowel
sound. You can get mp3s of a few results at the URL
http://briefcase.yahoo.de/skuehnl

> Can anyone point out a recorded example of a frequency shifter used
> musically?

Chris MacDonald?


Best,

Sebastian Kuehnl

__________________________________________________________________

Gesendet von Yahoo! Mail - http://mail.yahoo.de
Sie brauchen mehr Speicher für Ihre E-Mails? - http://premiummail.yahoo.de