comb filters

Hey everyone,

For those of you who are tuning up their
synths and effects boxes to test this
white noise-to-tone using a delay line
debate... when you set up your patch
with the 100% original in your left ear and
100% delayed in your right ear, after you've
convinced yourself that you don't hear the tone,
play with your set up using very small delay
times. Hear how the noise moves left and right?
That's called ITD - Interaural Time Difference,
one of the components of 3D positional audio.
It sounds smoothest if you have an effects
box that interpolates position when modulating
the delay time.

Just so everyone is on the same page, you don't need
feedback/regeneration to hear this effect - though adding
it increases the Q (resonance) of this digital
filter.

The previous posts are correct in that perception of
the tone comes from not just the 50Hz signal but
also its overtones (so the fundamental may not even be
needed - as in the mentioned case of band-limited noise).
And I do believe, upon reflection
that those partials are simply the peaks in mounds
that are centered about each expected integer harmonic -
the response curve of a comb filter.

When playing with the noise and its delayed signal
mixed together, try sweeping the delay time - you'll
hear the infamous WHOOSH of a comb filter being
swept over noise - it sounds just like an airplane
going by (which I was alluding to in the last post.

delay line length base pitch
20ms 50Hz
1ms 1000Hz
.2ms 5000Hz

These experiments are experiments in physical modelling -
noise is like air across an edge and a delay line is
like a whipple tube - that's why it sounds like a whipple
tube (tube spun around your head to make an airy whistling
tone).

- Jeff