What is this oscillator?

[FIX]

For pitch changing, like a doubler/chorus, a triangle wave pretty much gives a constant up and down shift in pitch, where a sine wave gives a sharp change then more gradual one as the wave moves along. Noise is also a good feed for doubling, filtered so it's really slow, gives that random feel...

 

[Bo Deadly]

For pitch changing, like a doubler/chorus, a triangle wave pretty much gives a constant up and down shift in pitch, where a sine wave gives a sharp change then more gradual one as the wave moves along. Noise is also a good feed for doubling, filtered so it's really slow, gives that random feel...

For delay based effects, it is noteworth that the control voltage is controlling the speed of both the encoder and decoder whether it be bbd, tape or digital memory. Meaning if you encode (record) signal using a rate that is sweeping, you're also simultaneously decoding (recovering) the delayed signal at that same sweeping rate. This compounds the effect of the control voltage which, depending on the speed relative to the delay, can tend to cause the effect to flop from one pitch to another and generally not be continuous.

A "solution" to this would be to use a tape machine that buffers the tape with momentary slack like an old cinema projector so that the tape is sweeping over one head but moving at a constant rate over the other. This would be impossible to do with a BBD and very difficult with digital memory but probably quite easy using an entirely software based implementation.

 

[JohnRoberts]

For delay based effects, it is noteworth that the control voltage is controlling the speed of both the encoder and decoder whether it be bbd, tape or digital memory. Meaning if you encode (record) signal using a rate that is sweeping, you're also simultaneously decoding (recovering) the delayed signal at that same sweeping rate. This compounds the effect of the control voltage which, depending on the speed relative to the delay, can tend to cause the effect to flop from one pitch to another and generally not be continuous.

You are making a pretty sweeping generalization, the different technologies will have different characteristics.

For today's TMI, I worked for a company back in the early 70s that made a voice pitch shifter for speeding up playback of talking book recordings (for the blind) at faster speed but restored to normal pitch speech. That company (VSC) used a BBD delay line with a sawtooth ramp controlling the clock frequency such that the samples were clocked in at one rate and out at another which created the pitch shift. At the end of each sweep we had to discard the delay line full of bad data that was clocked in at an unusable rate. For pitch shifting lower this discard was relatively short because it was clocked in at low frequency but out at high frequency, for pitch shifting in the opposite direction the discard was much longer and harder to ignore, but this is getting off into the weeds.

A "solution" to this would be to use a tape machine that buffers the tape with momentary slack like an old cinema projector so that the tape is sweeping over one head but moving at a constant rate over the other. This would be impossible to do with a BBD and very difficult with digital memory but probably quite easy using an entirely software based implementation.

I have probably mentioned this before but bell labs made a rotating head machine (not unlike VCR heads) that worked remarkably well... BBD approaches could be made to work with mild trickery, digital was much easier to manipulate for splicing samples together seamlessly.

JR