Analog Pitch Shifter Idea

[Ingo Debus]

Sean Costello wrote:
> The seperate clocks are driven by the sawtooth outputs of the
> oscillator, such that each clock is modulated by a sawtooth signal that is
> 90 degrees apart from the sawtooth modulating the other clock

But how is this sawtooth shaped? If the clock frequency rises/falls
constantly with time, there won't be constant pitch shift.
Simple example: At t=t0 there is a clock frequency of 20kHz; and when
the signal witten into the BBD at t=t0 has reached the BBD's output at
t=t1 there is a clock frequency of, say, 40kHz. So we get an upward
pitch shift of one octave. The signal that is written into the BBD at
t=t1 must reach the output at t=t2, when the clock frequency is at
80kHz.
If the clock frequency rises constantly, t2-t1 is twice as long as
t1-t0. In fact, t2-t1 has to be shorter than t1-t0, because the delay of
the BBD is shorter then due to the higher clock frequency.
Right? Or did I miss something?

Another problem: The BBD's output signal is unusable when the sawtooth
"reset" occurred between writing the input signal and reading the
output. So each BBD's output has to be muted quite a long time.

Isn't it better to use only two clock frequencies; one for writing into
the BBDs and one for reading them out? As soon as one BBD is filled
completely, the clock is switched to the "read out" frequency. At the
time one BBD is filled with signal, the other one is read out and vice
versa. For upward pitch shift, the "read" BBD has to be looped to get
the signal more than once.

Any comments?

Ingo