MOTM

At 5:28 PM -0500 09/02/01, J. Larry Hendry wrote:
>
>CV outs:
>It seems my basic sequence would need to have at least 2 but better yet 3
>>or 4 CVs so that multiple oscillators could change tuning relationship
>>through the sequence.

Four CV outputs is good -- that way it can be used with double the steps
and half the channels (16*2).

Also, CV outputs are useful for more than the pitch of the VCO's.  They can
be used with PWM, filters, VCA's, waveshapers, LFO's, etc.  You know, about
half the 1/4" holes in your modular :)  One could also be used to control
the rate of the clock driving the sequencer.

># of steps:
>I'm not sure 8 is right,  Can a sequencer be flexible enough to recycle or
>quit at any point along the chain?  How about 12 or 10 or 7 ??

I like the idea of combining a step sequencer with a gate sequencer into
one unit -- ie. having a gate output for each step.  I believe the Wiard
sequencer does this.  By adding a reset jack, patching the gate output from
that step brings the sequencer back to the beginning.

Also, having gate outputs allows one to compose a bassline and a basic
drumbeat on the same sequencer.  Of course, one could also drive a step
sequencer and a gate sequencer from the same clock

>Quantizing:
>I assume this would be essential.  Is this a feature one would expect
>>inside the sequencer or as another external module?  I assume if
>internal, >you would want to be able to switch it on and off as some
>voltages (maybe
>controlling other things from oscillators) you would not want to be
>quantized.

Yes, and for composing in different scales.

>Timing:
>I would think you would want to be able to use an internal or external
>clock.

Yes, the internal clock should have an output to drive other gear, and it
should have an input for an external clock so it can be slaved to MIDI,
SMPTE, etc.

>But, if using an internal clock, would the "time" of each step be a
>quantity fixed as one division of the total sequence?  Or would the time of
>each sequence be adjustable (like on Tony K's UEG?  It seems to me that a
>>12 step pattern where each step got 1/12 of the time might be OK for some
>things and not for others.

The UEG ignores the value of the time knobs and TCV when in step mode
(which is why I built that differentiator to trigger converter to use it in
loop mode).  Ideally, one would want a time knob that can be used with an
external clock.

Most hardware sequencers consider each step a "quarter note" and advance
each step after a fixed number of pulses per quarter note, eg. if I
remember correctly Linn used 96 ppq, Korg used 48 ppq, Roland used 24 pqn,
and so on.  It's my understanding that almost all modular sequencers use 1
pulse per step so that the same clock can also trigger an envelope.

My idea is to add a knob to each step that ignores a number of pulses based
on that knob's setting, such that:

knob setting:     result:
0                 skips that step
1                 normal, holds for one step
2                 holds for two steps
3                 holds for three steps
4                 holds for four steps
etc.

In addition to holding the current CV value, it would hold back a gate or
trigger output, thereby allowing one to compose notes of almost any length.

While a rotary switch would work well for this, one could use a pot, or
even add a CV input that effects the current stage.

You'll also note that setting all the knobs to zero after a certain step
changes the overall length of the sequencer.