Leslie simulations - again

[Mark Hammer]

I was sure I posted this the other day, and again today, but I can't seem to find a trace of it here, so here goes again.

I'm about to acquire a small Leslie speaker from a local guy (actually RE-acquire; I had one just like it but sold it in 1980 and I really miss it), and it got me to thinking about simulating rotating speakers.

Authentic simulations require one to consider the quirks of the device being simulated and going beyond merely the first thing you can find that sounds sort of like it.  In the case of Leslie speakers, our attention is drawn most to the notches created by the Doppler effect and to the cyclical nature of it.  So, some folks found an effective simulation in LFO powered bandpass filters (the PAiA Synthespin), then phasers (P90), and finally flangers near the upper end of their delay range.

The thing of it is, though, that although a spinning spekaer or speaker baffle produces swirly notches, that's not ALL it does.  As it points away from you, it also produces a slight modulation of both timbre (less treble when not facing you) and amplitude (louder when pointed at you).  So, what I'm thinking of is whether any of the existing rotating speaker sims factor in the "throb" of a real rotating speaker.

In principle, one could easily whip up a "throb simulator" module that could tack on to a basic design the way that RG Keen's LERA (at www.geofex.com) mimics the ramp-up/down time of rotor motors.  A couple of optoisolators could easily be used to gently modulate level and treble cut, and the LED half of the optoisolator can be yoked to the on-board LFO.  The nice thing is that since the notches themselves are generated internal to the flanger, the throb generator can be tacked onto the output without having to monkey around too much on the internals.

The thing I am having a difficult time wrapping my head around is the relationship of LFO polarity to throb. What *should* happen is that at the same time:
- the volume takes a little dip
- the treble is rolled back a bit
- the delay time is at maximum in its sweep (not maximum delay possible from the device)

Bear in mind that both treble trimming and notches will already create the impression of slightly reduced amplitude, so whatever amplitude modulation is imposed on top of that will have to be subtle.  Still, my guess is that you'd need to top off what the treble trim and notches do with a modest volume reduction.

Make sense?

Incidentally, it is the moving notch and treble trim that make the phasefilter mod on phasers so interesting.  I've only ever heard it on 4-stagers, and since I know from Mike Irwin's demonstrations that 12-stage phasers sound a lot more flanger-like, I'm wondering whether a 12-stager with the last 2 stages converted to lowpass (5 notches and a 2-pole lowpass filter) might do a passable Leslie simulation.