MOTM

I remember reading an essay a long time ago about modulating
frequency-domain devices. It claimed that a sinusoidal LFO was a
less-than-ideal source for modulation because of the log nature of the
frequency/pitch spectrum. A sin (or triangle, if we're being cheap) tends to
"rush" through the lower frequencies and take "too long" to sweep high,
giving a vertically off-balance cycling.

This article claimed that the best way to modulate these kinds of devices
was a negative-cycle full wave rectified sine. In other words, a wave that
looks kind of 'spiky' like this:

(format this diagram with a monospaced font such as Courier)

  |    |    |
_/ \__/ \__/ \__


Just thought I'd throw that into the discussion.

In fact, let me throw in another "beat the dead horse" comment:

Phasers work great with Shepard function generators as modulators...