talk algorithmic chord theory with me

[xoxos]

hello,

i'm working on a set of sem modules for synthedit (which will be released as dev material and as finished freeware vst) for algorithmic sequencing.

i am not facile with chord theory.. i've been reviewing material for a decade but am not an instrumental performer or theory based composer, so atm i'm still indifferent to the multivarious chord.. or key.. prog schema others describe.

because of this, i'm going to program several different engines.

what do you say advice, experience, requests? atm modules will probably return in the format of "priority-weighted" pitch values to be effected by a second class of sequence generators.

[JumpingJackFlash]

hello,

i'm working on a set of sem modules for synthedit (which will be released as dev material and as finished freeware vst) for algorithmic sequencing.

i am not facile with chord theory.. i've been reviewing material for a decade but am not an instrumental performer or theory based composer, so atm i'm still indifferent to the multivarious chord.. or key.. prog schema others describe.

because of this, i'm going to program several different engines.

what do you say advice, experience, requests? atm modules will probably return in the format of "priority-weighted" pitch values to be effected by a second class of sequence generators.

I'd be lying if I said I understood all that technobabble, but have you seen:
An Introduction to Music Theory
Scales, Modes and Chords

[dannthr]

hmmmm, is this for sequencing harmonies algorithmically or the whole composition?

If it's just the former, you'll want to pay attention to leading tones to hopefully let the user push the engine one way or another.

[jcivil]

I just have to say that unless you severely restrict the note choices to four, MAYBE five pitches (and no diminished or augmented seconds, I mean black key pentatonic, like) in a given octave your result will be so chaotic we won't call it musical

UNLESS YOU ARE VERY FACILE WITH CHORD THEORY, in theory and in practice

or you are kidding us

[thagoocher]

I just have to say that unless you severely restrict the note choices to four, MAYBE five pitches (and no diminished or augmented seconds, I mean black key pentatonic, like) in a given octave your result will be so chaotic we won't call it musical

I think this is more subjective than you make it seem. Even totally chromatic stochastic/probablistic music can be interesting to some people.

I work in Plogue Bidule which offers pretty extensive tools for algorithmic music. I don't know if you're familiar with it, but it may be good for you to look at to get some ideas. One thing that might be cool is a stochastic chord generator, based around some huge group of chords that jazz guys might use in a given key. You could narrow down what chords you wanted to use out of the pool. Where to get this huge list of whacky chords, I don't know. Ask one of those jazz guys.

[disturb]

MASSIVE -1 to anything stochastic when it comes to chords and algo stuff...

there are already plenty of options out there for that kinda thing, and what we really need, is a better way to navigate through music theory in a meaningful way.

[Laguna Rising]

wrong thread

[xoxos]

i am reading the replies..

[pd]

I haven't really had time to read it in detail but this looks interesting.

"Composer reveals musical chords' hidden geometry "
http://www.physorg.com/news71417403.html

Can't wait to see what you come up with.

[xoxos]

it's ~amazing to actually be doing it :p

i've found several such schema to digest.. (swirling eyeballs smilie) thanks. seems like the wavelength.

[jcivil]

I just have to say that unless you severely restrict the note choices to four, MAYBE five pitches (and no diminished or augmented seconds, I mean black key pentatonic, like) in a given octave your result will be so chaotic we won't call it musical

I think this is more subjective than you make it seem. Even totally chromatic stochastic/probablistic music can be interesting to some people.

If you're just dealing with lines, I'd tend to agree more, but (& maybe this is just me), music is of the ear, by the ear, and for the ear, and chords can be real sticklers for creating expectations for an ear, and very very very quickly tend to get in the way of one another outside of the ear's actual guidance.

My take may, on the other hand, seem anti-intellectual to you.

[disturb]

I just have to say that unless you severely restrict the note choices to four, MAYBE five pitches (and no diminished or augmented seconds, I mean black key pentatonic, like) in a given octave your result will be so chaotic we won't call it musical

I think this is more subjective than you make it seem. Even totally chromatic stochastic/probablistic music can be interesting to some people.

If you're just dealing with lines, I'd tend to agree more, but (& maybe this is just me), music is of the ear, by the ear, and for the ear, and chords can be real sticklers for creating expectations for an ear, and very very very quickly tend to get in the way of one another outside of the ear's actual guidance.

My take may, on the other hand, seem anti-intellectual to you.

I think you're making a totally valid point, which arises a very difficult question :
HOW can one design a system that will produce a non cahotic result, that will please the ear without being completely predictable...?

Every time i've fiddled with such things, i had to SEVERELY limit the system so that it didn't get outta hand...
So unless you come up with a genius idea, it's better to have a set of very limited 'modules' (whatever those are) that perfrom some predictable actions, and actually control the signal flow and/or vary a few params of said modules over time.
I think this view is pretty much what xoxos meant with his first post, and i'm eager to see/hear whatever he comes up with.


And to anyone interested in this, you don't need crazy dope coding skills or anything, ableton live is actually pretty fun in this regard, and basically anything modular that handles midi well opens up a world of possibilities.

[jcivil]

I think such a system has to at least be able to emulate intelligence (learn/adjust), and of course I haven't the first clue what that means in execution

[xoxos]

here's my analysis -

we know that mathematics is.. i hesitate to say synthetic ..every day mathematicians are discovering relations, patterns between numbers.. by extrapolating and abstracting.. make an abstraction.. something interesting happens in it.. often something fairly simple and straightforward.. suddenly it's a new principle that has "always been true.." but wasn't liminal/known previously.

one thing music/ears/mind is about is oc patterns.. the experiential focus in sound can be highly subjective.. most people can report their interests in music changing, often that something once completely unpalatable was 'understood' et c.

7et might not sound harmonic.. maqamat might not sound harmonic.. mongolian folk singing et c.

dissonance, masking are energy dynamics.. many orchestral pieces use both..

algorithmic composition fascinates me in discerning what dynamism to give to various aspects.

i learned one of the fundamental applications of music from catching a few minutes of some film.. kingboy d is playing the harp to sooth someone and then he plays a dissonant, abrupt chord. in the mind of the listener, they are brought to a place of ease and then 'snapped out of it..' bringing focus on the cause of unease..

..so i've listened to a fair amount of algorithmic progression, and acclimatised..

similarly, i use algorithmic lyric engines that generate sentences by rules (atm.. content is easy to implement as well...) which generally emits sentences that don't "make sense in english.." however.. it's my baby, and with continued listening i find that the 'random' semantic conjunctions are often amusing through subtle or abstract interpretations of the words..

you may have heard my "when everyone should receive an instrument." the lyrics were generated entirely algorithmically, but have a fairly coherent meaning if you're semantically limber and forgiving.

saying something "illogical" can circumstantially bear discrete meaning as well. how intelligent is the universe?

unless the output is extremely fortuitous, i generally can't listen to anything purely 'stochastic' (eg. even weight chromatic) either.

'breathcube' has a very simple tritone major/nat chord engine. i tried a progression engine but found that it was more interesting with no progression weighting. this is imo a function of the entire density.. because most elements are simple (4/4 only et c.) this weighting increases dynamism without too much 'chaos' to my sensibility.

this kind of discernment is an artistic process, and is perhaps the crucial infusion? :p

an algorithm to give weight to dynamics in an algorithmic engine.. a dozen years from now

coffeetype,

[thagoocher]

Well, if you read this thread or are familiar with Harmony Navigator, I think that's similar to what I was suggesting, except with some probability weighting.

If you set it to a given key, it would give you a set of chords for each scale degree that somebody has decided "work" (of course, any set of notes is a chord and has some relation to the key, even if it's so distant as to seem wrong to our ears). These chords could be arranged in a way (as suggested by the theorist in the link down there) that the "safest" chords would be clustered together (whether actually, in the GUI, or just conceptually), and as you got farther away from that "safe" center, the chords would become more contorted and possibly more dissonant. How far to stray from safety could be controlled, as well as how quickly to move between areas, and so forth. If you've ever seen Leafcutter John's program called Forester, it could even be a visual environment like that, with a virtual person wandering around a little forest of harmony, and its movement away from the safety of its cabin and into the terrifying wilderness (tritone substitution ) could be controlled by some probability factors.

I guess my point was that using an algorithm to come up with chords is probably being used to break away from the chord habits we all get into. Limiting the amount of chaos possible seems counter-productive, because "chaotic" and "bad" are in the mind of the musician and listener. Anything like this should be able to go beyond what the designer may decide is listenable, while being scalable to whatever realm of harmony works for the user.