SergeModular

I can't say how the Doepfer FS works, but I think most analog frequency
shifters work by implementing the trig identity

cos(x + y) = cos(x) * cos(y) - sin(x) * sin(y)

in analog hardware. Imagine "x" is your input signal and "y" is the
carrier oscillator. If the equation above can somehow be performed
continuously in real time, then your signal "x" gets shifted by whatever
the frequency of the carrier "y" is (y gets added to x).

This is done in hardware by making a sin and cos version of the input
signal with a phase delay network, then ring modding the cos version of
the signal with a cosine carrier wave from a quadrature oscillator (this
generates the cos(x)*cos(y) part) and, in parallel, ring modding the sin
version of the signal with the sine carrier wave for sin(x)*sin(y). The
two ring mod outputs are then subtracted from each other to get the
final output.

-Chris

John Papiewski wrote:

>
> Igor,
>
> Think about it.... One ring modulator yields sum and difference
> frequencies mixed together. The expense in the FS AFAIK is in a
> filtering network that splits the sum out to one output and the
> difference to another. There are several compromises to be addressed in
> this filtering network. Difficult design, a lot of parts and labor to
> get it all tuned right.
>
> That's why it's so expensive.
> The Doepfer FS I believe has a fairly basic version of this network...
> it can do some frequency shifting but it is compromised in terms of
> performance... I am not speaking from direct experience but according
> to the doepfer web site that shifts of < ~ 50 hz are not possible. I
> guess that handling the smaller shifts involves more difficult design
> and construction. Also in the Doepfer, I think I heard that signal
> bleed-thru is a problem (incomplete suppression of signal or carrier).
> So perhaps the application of the Doepfer FS is more limited to doing
> klangy distortion and robot voices than more subtle effects, but hey,
> they offer it at a fraction of the price of anyone else.
>
> Other things to consider:
> The Wilson Analog Delay (WAD) Serge module does pitch shifting, but the
> pitch shift can't be constant... so you can do vibrato and doppler-type
> shifts with it, as long as the "dv/dt" is not zero (sorry, my calculus
> is showing). Blacet makes a similar module (VC analog delay), I think
> Big Briar too. You can do this with some digital delays as well, since
> in each instance you're changing frequency output by varying a master
> clock rate.
>
> ANOTHER module that can do pitch shifting is a notch filter or phase
> shifter. The shift amounts will be very small but if you shift the
> phase you're pushing the wavefront forward or back... again, it has to
> be changing to work (remember dv/dt <> 0). So you can do vibrato and
> some FM types of effects with a phase shifter.
>
> Guys if I don't have this right let me know.
>
> JP
>
> dahlia13@... wrote:
>
> > Thanks! That's as lucid and down-to-earth an explanation as I
> > could've asked for. So basically, a frequency shifter allows you to
> > send the sum and the difference frequencies to seperate outputs (that
> > must sound interesting with one output panned hard left, the other
> > hard right), whereas a "conventional" ring modulator doesn't allow
> > you to seperate the two. That's the main difference?
> >
> > What I was referring to as an octave divider sounds like what's being
> > called a pitch shifter in this thread. For some reason, I always
> > thought a pitch shifter was strictly a digital device that changed
> > the pitch of sampled audio input without altering the length/speed.
> >
> > This may be a naive question, but if a frequency shifter is basically
> > two ring modulators in parallel, why are FS's so expensive? For the
> > price of one Serge FS (external carrier), a person could buy 4 or 5
> > of the ring modulator modules...does anyone have any theories as to
> > why this is? Is seperating the sum/difference frequencies more
> > difficult to achieve than producing both of them combined?
> >
> > Igor.
> >
> >
> > --- In SergeModular@y..., John Papiewski <johnp@w...> wrote:
> >
> >> Hey Igor,
> >>
> >> You can think of a frequency shifter as a ring modulator with TWO
> >> outputs: the upshift output has only the sum of the two input
> >> frequencies, the downshift output has only the difference
> >
> > frequencies.
> >
> >> Serge has two FS's in their catalog, the internal carrier model,
> >
> > which
> >
> >> has an onboard sinewave oscillator, and the external carrier, which
> >
> > lets
> >
> >> you use your own sinewave or some other waveform or complex source.
> >>
> >> Please note that since it is ring modulator based, a frequency
> >
> > shifter
> >
> >> is NOT a pitch shifter! A frequency shifter will add the carrier
> >> frequency to all the harmonics of the sound you want to shift. The
> >> greater the carrier frequency, the more pronounced the harmonic
> >> distortion. This is suitable for, well, distortion effects like
> >> robot/computer/alien voices, etc. At very low carrier frequencies
> >
> > the
> >
> >> FS will simply impart vibrato, mild doppler effects, leslie effects
> >
> > etc
> >
> >> since the harmonic distortion is more subtle and harder to notice.
> >>
> >> A pitch shifter on the other hand will multiply by the harmonic
> >> frequencies, keeping the original sound's harmonic structure
> >
> > intact, so
> >
> >> you can *transpose* the original sound by an octave, fifth, or some
> >> other arbitrary interval. You can't do this with a frequency
> >
> > shifter as
> >
> >> far as I know.
> >>
> >> Make sense??
> >>
> >> John P.
> >
> >
> >
> >
> >
> > To unsubscribe from this group, send an email to:
> > SergeModular-unsubscribe@yahoogroups.com
> >
> > Keep on Patchin'!
> >
> >
> > Your use of Yahoo! Groups is subject to http://docs.yahoo.com/info/terms/
>
>
> To unsubscribe from this group, send an email to:
> SergeModular-unsubscribe@yahoogroups.com
>
> Keep on Patchin'!
>
>
> Your use of Yahoo! Groups is subject to http://docs.yahoo.com/info/terms/