RE: dual loop technique? errata

[Greg Meredith <gregmer@MICROSOFT.com>]

Bob,

Just for the record, in the last sentence, i meant to say: 'i.e., it's not
even algebraic.' Obviously, there are irrational numbers which are not
transcendental. 

--greg

        -----Original Message-----
        From:   Greg Meredith 
        Sent:   Thursday, January 21, 1999 1:42 PM
        To:     'Loopers-Delight@annihilist.com'
        Subject:        RE: dual loop technique?

        Bob,

        i work with a single JamMan. However, when i do multiple loops, i
often do it manually. That is, i do it using polyphonic instruments, e.g.
stick or piano, and/or i employ multiple players. So, this may or may not 
be
on target as the kind of technique in which you might be interested. 

        However, what i do is to think very much about the 'phase' 
structure
of the piece. i relate each loop to a common tick (finest audible
subdivision of the beat). But, i try to emphasize a different pulse in each
loop. This gives the soloist or improvisational voice the opportunity to 
mix
pulses. i look to create loops where the different pulses get maximally out
of phase and then come together for a kind of tension-release effect.

        For example, i may have one loop with a strong 5/4 pulse and 
another
with a strong 4/4 pulse. If the shortest note (in both loops) is an 1/8th
note, then they are synchronized by an 1/8th note tick. In this case, if
both loops are 1 measure long, then they would come together every 40 ticks
(assuming the measures don't repeat internally). i then try to write my
loops to maximize the tension, via harmonic structure, dissonance, 
dynamics,
etc. just before the 41st tick, and then resolve on the 41st tick.

        This technique gets more interesting when the loops are more than
one measure long. In the example above, 40 ticks is 5 measures of 4/4, and 
4
measures of 5/4. That is, if we repeat a single measure of 4/4 5 times it
will line up exactly with a loop that repeats a single measure of 5/4 4
times. If we then make the 5/4 loop be two measures long, each measure
observably distinct, then we don't have a real line up, i.e. pulse and note
values line up, until 80 ticks go by. But, we have the pulses lining up at
the 40 tick mark. Similarly, every 20 ticks there is a mini-"node point",
where a subdivision of the beats of the two loops come together. By playing
with mounting the tension and resolving, slightly on these internal node
points, you can create really interesting effects. 

        Note: i have intentionally left out what i think increases tension
because i think that's different for different ears. Tritones, for example,
make me happy.

        Another interesting technique is to create an 'outer' loop which
uses the inner loop as 'events'. So, we could use the same 5 against 4
structure in the example above as a guide for triggering the 5/4 loop or 
the
4/4 loop. The question you have to decide is the duration of the event. If
you set it equal to the duration of the longest loop, you only get the 
loops
playing at the same time every 20 'events'. If you set it shorter, you get
more overlap. 

        (This approach, imho, seems much a much more promising application
of fractals to music. The dimension along which the piece is self-similar 
is
time. The application above says what happens to the structure of the piece
as you increase the time dimension. Clearly, you see a similar structure. 
As
you diminish time, the same thing out to happen. More specifically, as you
diminish the time, you get closer to the pulse of the frequencies that give
rise to our experience of the note. My (untested!) belief is that if this
pulse is related by a number of iterations of an IFS to the musicians
experience of the basic pulse(s) of the piece, then you get appealing
noise.) 

        Finally, it may seem to many that one must relate each loop to a
common tick. But, i would suggest to a person that feels this way to listen
to the work of Conlon Nancarrow, who uses player pianos to do looping and
other interesting and related things. He sets the rhythmic relationships
between the voices (take this to mean a generalized notion of a loop) to be
non-integral numbers. In fact, some of his pieces have voices progressing
with respect to each other at a rate measured by a transcendental number
(i.e., it's not even rational). 

        i hope this helps.

        --greg

                -----Original Message-----
                From:   Bob Campbell [SMTP:astropulse@hotmail.com]
                Sent:   Wednesday, January 20, 1999 11:08 PM
                To:     Loopers-Delight@annihilist.com
                Subject:        dual loop technique?


                I'm interested in hearing ideas about how 2 loop devices 
                can be used creatively together. I use JamMan's so am
specifically
                interested in exploiting that device, but any generic
technique
                ideas are of interest.

                Bob




                ______________________________________________________
                Get Your Private, Free Email at http://www.hotmail.com