Archive of the former Yahoo!Groups mailing list: MOTM

Now I'm going to grit my teeth and contribute my two cents worth.

Paul S. is 100% correct in his evaluation of why this clicking occurs.
I spend countless hours editing sound files in ProTools and I can tell
you for a fact that anytime you cut a waveform it will click unless
the cut occors at the zero crossing. This is because the edit (which
acts exactly as a fast envelope generator drivin a VCA) causes the
waveform to change value instantaneously (or nearly so.) If you
looked at a frequency plot of the result you would see a noise spike
or a click.

There are two ways around this in ProTools. Either you make a fade
instead of a cut (which is the same thing as lengthening the attack
and decay time on an EG) or you make sure the cut is at a zero
crossing. The lower the frequency, the more you may have to shift the
edit point to hit the zero crossing. This tends to cause the edit to
be either early or late.

It may be possible to build an EG that delays it attack until the
incoming waveform crosses zero, especially if the attack is set very
short. It is also possible that if you were using a DCO instead of a
VCO, you could cause the DCO to start its waveform at the leading edge
of a gate signal. Digital synths may in fact do this.

I have owned quite a large number of analog synths over the years and
they have all clicked with too fast of an attack on low notes. I
can't say that my MOTM clicks any worse than any other synth under
those conditions and I do have both 190 and 110 VCAs. I have never
found the clicking that does occor to be objectionable anyway.

I would also agree that the speakers may be part of the problem. A
speaker cone cannot move instantaneously, (I believe for it to do so
would require infinite energy.) When a speaker cone tries to move
faster than it is capable of moving, all sorts of weird transients can
result. This problem may disappear when using a sub, as the subs
frequency response is severely limited and the transients may be
stopped by the crossover. The crossover may also cause enough phase
shift at the transition point to spread out the transient in the
tweeters making it less objectionable.

To expect the VCA/EG combination not to click at low frequencies under
fast transients, without implementing some kind of comsensation
strategy is to expect Paul S. to violate the laws of physics.

Paul H.