|
Author: flies on June 20 2006 Viewed 3197 times. 4 people liked this blog. You can rate it below if you haven't already.
-->
so my computer is dead right now, but that doesn't stop me from having patch ideas.
this one came to me last night while i was listennig to greg davis's album arbor (a fine album, if it does feel a little dated now. his breakwork is sometimes good, sometimes not so good, but most of the time i'd rather there were no drums at all).
basically the idea was to make a patch (that if I make it will fit into my modular format) based on amplitude modulation where the different 'operators' are all upper harmonics in some harmonic series.
the idea is that harmonics form a set that is closed under the operatino of AM, so if you take H_10 x H_12 (where H_n is the nth harmonic of some fundametnal and x represents the operation 'ring modulate'), then the sidebands will be H_2 and H_22, and then i f you cross taht will H_8, then you get H_6 + H_10 + H_14 + H_30, all of the resultant sidebands would be harmonically related.
-- one technical note is that in AM unlike FM, it doesn't matter what order the operations happen in, meaning (H_8 x (H_10 x H_12)) = ((H_8 x H_10) x H_12)
so the idea was to take something 7 sine waves, going from, say, the 5th thru the 11th, and you made the middle one, 8, the 'carrier' and then you used the other ones to amplitude modulate the main one, then you might come up with something interesting. with 6 modulators, this scheme creates up to 729 harmonics (729 = 3^6. this number is inflated b/c it counts overlapping harmonics double, but the point is that it's a rich timbre)
by controlling the amount of amplitude modulation by the 6 modulators (from no modulation to ring-mod), you control the volume of teh sidebands vs the carrier, allowing for continuous variations in tone. and if the amount of AM is controllable with 'CV', then you could get something very interesting. putting the resultant output through a bp filter could again create something pretty cool.
additional variations would include introducing pitch instability into the modulators. this might sound really cool, since it would destabilize the 'closed-ness/closure' of the whole scheme, and the harmonics present woudl 'spread out' - the 729 harmonics would stop overlapping.
one of the nice things about this idea is that it'd be very hard to do with a real modular, where getting 7 sinewaves is expensive, and then tuning them into harmonic intervals is tricky, and having AM that varies from nothing all the way to ring mod is basically impossible so far as i know (there are ring mod modules and there are VCA's, but afaik a VCA cannot produce a ring mod effect), and if it were possible, you'd still need 6 VCAs, and if you wanted to control the amount of AM with CV you'd need 1 additional VCA for each CV input. If you were to do that, i'm sure it would sound way better than it would on a computer, tho, needless to say.
| |
Comments
06/20/06
+
PM |
QUOTE |
PERMALINK |
REPORT
jarvis
so the pitch instability would have to be very small numbers with all that multiplication going on, right?
06/20/06
+
PM |
QUOTE |
PERMALINK |
REPORT
flies
sounds right to me. a small amount of fm on some of the modulators would create a nice subtle instability in the overall timbre, i imagine. more pronounced fm effects would also be possible, and would probably sound cool, but it might not.
my thought had been to put noise~ through a few onepole~ objects with a cutoff at about 10 hz and then use that to fm the modulators by about 0.05 of the modulator frequency. noise~ -> onepole~ would work, also rand~, or just plain cycle~.
06/20/06
+
PM |
QUOTE |
PERMALINK |
REPORT
jarvis
what about running some of the sine waves through wave folders to add some harmonics to the sources? what would that do?
06/20/06
+
PM |
QUOTE |
PERMALINK |
REPORT
flies
could be cool. wave folding definitely sounds like distortion, and so that could be good or bad depending on the application. does wave folding always add harmonic distortion, or does it also add inharmonic distortion?
so if you added wave folding to every sine wave, you'd end up with some serioulsy noisey shit, which could definitley be lots of fun, depending on how well it worked.
another patch idea i had was to make an emulation of the doepfer a137 module, which is a fancy wave-folder link . the tricky thing there is rounding the corners around the folding points. i was thinking that it'd be possible to do that using lookup~, but having the amount of rounding continously variable would be tricky. you couldn't do it with lookup~ unless you just crossfaded btwn the outputs of different lookup~ objects, which might actually work well.
06/20/06
+
PM |
QUOTE |
PERMALINK |
REPORT
flies
btw pong~ makes wave-folding super easy, i used to use it as a trebly distortion back when i made more noisey music - like 8 months ago i started doing more soft droney stuff, but i am trying to fold some of the noise back in.
anyway, it occured to me that haveing a very distorted modulator that had a small modulation strength would add a certain bite to the whole sound, which would otherwise be rather 'clean'. the idea had been to make a clean sound, but something like wave-folding or clipping (or soft clipping) would definitely give the patch a wider palette.
edited: Jun 20 2006
+
PM |
QUOTE |
PERMALINK |
REPORT
jarvis
well, maybe my idea of a wave folder is wrong. what i'm thinking is a sine wave whose amplitude are limited so that when the peak of the cycle hits a certain threshold, the portion that goes above the threshold is inverted and actually goes downwards, like it's adding a cycle to each positive and negative portion of the wave. then you could also have a lower-boundary threshold where as the amplitude of the folded portion of the waveform would be folded back again when it reached a zero crossing.
06/20/06
+
PM |
QUOTE |
PERMALINK |
REPORT
flies
no that's what i understood you to mean, eg if you were to fold a saw wave about the 0 axis you'd get a triangle wave.
06/20/06
+
PM |
QUOTE |
PERMALINK |
REPORT
flies
it occurs to me that i'll want to have some of the modulators be of low, medium and high order, since their behaviour in AM will be different. i mean
H_1 x H_8 = H_7 + H_9
whereas
H_10 x H_8 = H_2 + H_18
so getting a mix of these effects would be good.
something like the following: modulators = H_1, H_2, H_5, H_6, H_13, H_16
carrier = H_8
06/26/06
+
PM |
QUOTE |
PERMALINK |
REPORT
flies
i am really dorky and also enjoy wasting time at work, so i made this script that calculates the strength of every sideband produced with a six modulator system. I also implemented a fold-over system so that aliased frequencies will be calculated accurately.
06/26/06
+
PM |
QUOTE |
PERMALINK |
REPORT
flies
oops forgot the link: link
06/20/08
+
PM |
QUOTE |
PERMALINK |
REPORT
blockhead
Interesting idea. It inspired me to throw together a program. It spits out a wave file of two seconds with randomly generated modulation using ring mod of several sine waves. I was surprised that everything was still harmonic and surprised at what the audio sounds like: those yamaha digital FM synths from the mid 1980s.
Was gonna post it here but i've just learned that the "code" tag does not work in replies to blogs. :/ Poop. Maybe I can put it in the coding section.
Register / login
|
^
EM411 is Copyright 2001-2008 EM411.com
All rights reserved. | Contact | RSS
|
