how to get some order in this atonal melody mish mash?

Isjtar

hi,

I'm working on some generative stuff and was wondering how people on here get some modal/tonal order (what an ugly word) in their melodies and harmonies.
It's all to easy programming some 12tone rows, retrograding them etc., but how do *you* get less atonal melodies?

for example, I could force a random stream of chars to be ordened by a complicated system of adding and substraction, but before I start doing stuff like that; I'd like to know some alternatives.

i've been having some nice results with Cellular Automata generated melodies, which is pretty cool...

Perhaps a probabolistic table with desired notes, but how do you access them not purely at random, maybe some fancy algorithm, attractor style?

another thing is how to get some nice velocity templates into them, I'm sick of 127 and even sicker of random...

and Peter, if you're going to offer me Litter Power, you better get some examples at hand ; )

thanks,

junior _ isjtar

roger.carruthers

Probably a pretty redundant reply, but just in case
you've missed the Pierian Spring that is Karlheinz
Essl's RTC Lib:
http://www.essl.at/works/rtc.html
cheers
Roger

Isjtar

yes, a bit redundant ; )
no, seriously, some nice stuff in there, I especially like the sneak rythm, but I was hoping for something different...

cheers

Gary Lee Nelson

I do a lot of this sorting of thing. Lately I have been using various
iterative function (chaos, Barry Martin, Cliff Pickover, etc.) to generate
what I call "structure" - musically speaking this contains rhythms and
register. Then I apply "surface" which is exact pitches and orchestration.
You may be familiar with Webern's "klangfarben" treatment of a Bach
Ricecare.

One of the most useful methods for applying order to "atonal" structure is
the sieve. I made an abstraction called NewSieve (I had several older
ones). The left inlet takes any MIDI notes you please. The right inlet
takes a list of pitch classes say 0 2 4 5 7 9 11 if you want the output to
coerced into the notes of a C major scale. The right inlet can be varied in
realtime to create harmonic progressions.

One very cool technique is one I call "cross composition" where the right
inlet receives the notes from a midi file collected with bag. You can hear
this in action in the soundtrack for for my film "Death and
Transfiguration." The structure is a 2D fractal based on Barry Martin. The
surface is a midi file playing Strauss' tone poem of the same name. You can
view the film on my website shown in my signature.

Here are the patches. Note that I use Peter Elsea's Lobjects so you'll need
to get those from maxobjects.com.

NewSieve

Max Patch

Copy patch and select New From Clipboard in Max.

And NewSieve.help

Max Patch

Copy patch and select New From Clipboard in Max.

Cheers,
Gary Lee Nelson
TIMARA Department
Oberlin College
www.timara.oberlin.edu/GaryLeeNelson

Peter Castine

On 20-Mar-2006, at 11:36, junior wrote:
> It's all to easy programming some 12tone rows, retrograding them
> etc., but how do *you* get less atonal melodies?

I won't offer Litter Power because this is stuff you can readily do
with standard Max objects. For instance: table.

table 0 2 4 5 7 9 11 12 14 16 17 21 23 24 ...

(continuing the table, setting it up for Ab minor, etc. left as an
exercise for the reader).

Send any numbers you want to the table and out comes something
guaranteed diatonic.

Is that hard?

If what you want is something that sounds like a real tonal melody
with functional tonic/dominant tension/release... now, that's a bit
more tricky. You may find some of Peter Elsea's list objects helpful
for that, and someone has written some set class objects that might
be helpful.

-- P.

PS:
> you better get some examples at hand

Est-ce que tu a dejas vu des examples chez http://www.bek.no/
~pcastine/Litter/Examples/ ?

jbm

I second the Elsea vote... he's also got a nice fuzzy logic tutorial which might give you some ideas.

Peter Castine

On 20-Mar-2006, at 13:09, Vincent Goudard wrote:
> Made me think of indian raga hierarchy of note, where
> any note of the scale has a name and a "probability"

Which reminds me, in turn, of this:

Any college student who was subjected to Piston's Harmony may recall
his description of the likelihood of individual harmonic steps of the
scale, even giving a sketch for a first-order Markov chain of
harmonic progression.

Page 18 in the 3rd (1962) edition.

You could probably get some wandering wallpaper diatonic non-
structure out of this approach.

-- Peter

uyttendaele.jeroen@gmail.com

Hi j,
i don't know what the hell you'r talking about but here's some of my advise:
the only thing i know is, drink less rocheforts and patch some more.
cheers, Dr Jerre

robotic-audio

also look at the midi scale example... the COLLective called 'Harmonic Mysteries'

Isjtar

thank you all for you replies...

Quote: Peter Castine wrote on Mon, 20 March 2006 08:16
----------------------------------------------------
> I won't offer Litter Power because this is stuff you can readily do
> with standard Max objects.

i know, max was pretty much made for this kind of stuff, was just making a bad joke...

> table 0 2 4 5 7 9 11 12 14 16 17 21 23 24 ...
> exercise for the reader).
>
> Send any numbers you want to the table and out comes something
> guaranteed diatonic.
>
> Is that hard?

not at all, I've been doing it, but it always sticks at that same mediocre level, maybe I need to elaborate on that technique a bit...
>
> If what you want is something that sounds like a real tonal melody
> with functional tonic/dominant tension/release... now, that's a bit
> more tricky.

looking for everything really, ideally, i'd like an xy control to fade between tonal/atonal and funk/monk, would be oretty nice, but I guess that will be tricky too : )
> Est-ce que tu a dejas vu des examples chez http://www.bek.no/

no I haven't, will do and you're talking french to me again ; )

Isjtar

Quote: Gary Lee Nelson wrote on Mon, 20 March 2006 05:23
----------------------------------------------------
> One of the most useful methods for applying order to "atonal" structure is
> the sieve. I made an abstraction called NewSieve (I had several older

thanks for that, will come in handy...
next up are dynamics and rythm, preferably in conjunction, so if you have some more, I'd be delighted to deconstruct them.
keeps surprising me how people always have their own custom methods in max. on the other hand, it keeps surprising me how lazy people can get with ma -myself included....

>vg said:
>Made me think of indian raga hierarchy of note, where
any note of the scale has a name and a "probability"
--
>The vadi, samvadi, vivadi and anuvadi namely, sonant, consonant,
dissonant and anuvadi
are compared to the king, the minister, the enemy and the servant,
respectively.
--
>My patch is under construction and just ugly and mixed in a messy
wire net, I can send it to you if you really want..

as you please, sounds nice, don't hesitate if you feel like putting something up...

thanks

ps @j900 nice intro, you spammer : )

Isjtar

Quote: jbmaxwell@btinternet.com wrote on Mon, 20 March 2006 08:26
----------------------------------------------------
> I second the Elsea vote... he's also got a nice fuzzy logic tutorial which might give you some ideas.
----------------------------------------------------

thanks, this is the link to the paper if anyone might be interested http://arts.ucsc.edu/ems/music/research/FuzzyLogicTutor/Fuzz yTut.html