[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: [microsound] AI & rhythm perception - 'groove' heuristics?
David Powers wrote:
As far as velocities being an essential part of creating a good
groove, well that's true. But neither is it mathematically
quantifiable. I like FLStudio (yes some think it is a toy) because I
can draw in velocity curves really quickly, and many times very
contradictory sets of velocities will work.
How do you know they're contradictory? Do you find you need to do some
tweaking to make them the 'right kind' of contradictory? I know exactly
what you mean, and it is my hunch that they appear contradictory but in
fact they express the same thing, just in different ways; like maybe one is
the derivative of the other, or the log of the other, or something.
Datamining algorithms run over mathematical representations of beats would
help uncover this, I think. Or demonstrate that it doesn't exist.
(It is also very easy to
delay a particular beat, and I often have swing vs. straight 16ths at
the same time, or 16ths with different swing, or swing that is drawn
in by hand at points where it feels correct.)
I used to work with Ableton Live, in a very freeform way with a live band
that I'd sample on the fly, and sometimes I'd end up writing a piece of
music that was in 3/4 on a 4/4 grid at 1.5x the tempo Live thought it was
running. Trying to manually marry a bunch of different swing feels into one
timeline on top of this is a nightmare.. When it worked it made it feel
more 'natural' but not necessarily groovier.
And, following from this, you must especially consider things like the
actual mix and the compression on the mix, which have a HUGE effect on
the accents and psychoacoustic effect, but are quite hard to quantify
and aren't usually considered part of the composition of the groove
per se. All this stuff goes into making a good groove.
Yeah, I agree with you that compression and mixing plays a big part in
adding the final touches to a groove; but a good groove sounds like a good
groove regardless of how it's compressed or produced. Mixing and
compression can only bring out its goodness, they can't create goodness
from nowhere. Unless you're using excessive compression to make particular
things duck using side chains (cf Rhythm & Sound/Basic Channel), but that
means using compression at the composition end to mechanically specify your
velocities for you, rather than using it as the kind of compression
normally associated with mastering.
This idea of groove still allows Fela Kuti to work, too. You can get any
disparate collection of musicians to play a Fela Kuti groove and, so long
as the musicians know how to play their instruments, it will sound and feel
like a Fela Kuti groove. Here you've got different mixing, compression, and
mastering parameters being roughly equivalent to different sets of musicians.
--
Damian Stewart
+64 27 305 4107
f r e y
live music with machines
http://www.frey.co.nz
http://www.myspace.com/freyed
---------------------------------------------------------------------
To unsubscribe, e-mail: microsound-unsubscribe@xxxxxxxxxxxxx
For additional commands, e-mail: microsound-help@xxxxxxxxxxxxx
website: http://www.microsound.org