Conference napalm::commusic_v1

250.0. "Delayed/Modulated Sync" by SIVA::FEHSKENS () Tue Feb 18 1986 13:47

I recently (finally) got around to doing an experiment that I wanted to
do for a while - multitrack the same instrument using a sync track run
through a modulated delay.  The goal is to simulate the minor timing
differences that make an ensemble sound like an ensemble rather than one
loud composite instrument.  This is the effect that a chorus box attempts
to produce (I should compare my results to just running the signal through
a chorus).

Anyway, I laid down a sync track from my Roland TR707 on track 8 of my
Tascam 38.  Then I ran the track 8 output (dbx disabled on this track for
both record and playback) through a Deltalab Effectron-II ADM-1024 digital
delay.  I set the delay in the 256 msec range, dialed in a delay factor of
.5 (to give me a base delay of about 120 msec), and set the modulation
rate to .1 Hz and the modulation amount to about 20%.  I figured this
would give me a delay modulation of 20 msec, about 20% of the time between
16th notes at 120 BPM (120 BPM = .5 sec/quarter = .125 sec/16th), a
variation that ought to be audible and out of the range of flanging and
phasing effects ("automatic double tracking" uses delays on the order of
50 msec to get the doubling effect).

By gosh and by golly if it didn't work.  The TR707 faithfully tracked the
delayed/modulated sync signal, as long as the modulation frequency didn't
exceed about .3 Hz.  At that level, the tempo is audibly reeling like a
drunken sailor.  A little more and the 707 just gives up, I suspect because
the modulation is now pitch shifting the sync track enough to lose tracking.
I didn't see how far I could go in terms of modulation amount, although I did
get as high as 30% or so.

Then I laid down 6 audio tracks of drums playing similar patterns.  I synced
my TR727 and TR909 to the 707 via the MIDI interface, and programmed patterns
into the tomtoms and congas.  For each track I slightly modified the audio
EQ and the delay parameters (base delay, modulation amount, modulation
frequency).  All six tracks played together sound like a massive African
war rally.  The effect is incredible.  A close listen with headphones disclosed
some phasing/flanging as the various delay modulations interacted with one
another, but more interesting was the way some voices seemed to fade in and
out as their position relative to other copies of the same sound shifted in
time (they'd disappear or fuse when they overlapped perfectly).

This technique is a lot of work (it took about two hours to do 6 tracks plus
sync for a three minute piece, with trivial drum machine programming), but the
effort can be worth it if you want a massed choir effect.

len.