[MMM] My first composition.

I wrote an algorithm based on my harmonic permutation JI theory.
Still very simple, I will add many more things to the algorithm over time.
But I allready really enjoy it's output.

5-limit permutation just intonation sung by computer
choir.<http://sites.google.com/site/develdenet/mp3/9-12-2009_5limit_harmonic-permutation-ji_choir.mp3?attredirects=0>

Should the link not work (which sometimes seems to be the case with google
pages direct links) you can find the mp3 at www.develde.net

Hope you enjoy!

Marcel

[Non-text portions of this message have been removed]

Wow Marcel,

this sounds really nice to my ears! Not at all what I expected as I find most algorithmic music I've heard disappointing. This doesn't sound particularly random at all, not even as a whole which is really surprising. I haven't been that impressed with your retunings of existing pieces but it seems that your theory has a lot of potential for creating new music. I'd be really interested to see the details of your algorithm and how they relate to your theories.

Kalle

--- In MakeMicroMusic@yahoogroups.com, Marcel de Velde <m.develde@...> wrote:
>
> I wrote an algorithm based on my harmonic permutation JI theory.
> Still very simple, I will add many more things to the algorithm over time.
> But I allready really enjoy it's output.
>
> 5-limit permutation just intonation sung by computer
> choir.<http://sites.google.com/site/develdenet/mp3/9-12-2009_5limit_harmonic-permutation-ji_choir.mp3?attredirects=0>
>
> Should the link not work (which sometimes seems to be the case with google
> pages direct links) you can find the mp3 at www.develde.net
>
> Hope you enjoy!
>
> Marcel
>
>
> [Non-text portions of this message have been removed]
>

This is absolutely amazing. How did you do this? Is it constantly modulating
between related keys at random, just keeping one tone common each time?

Right off the bat you go Gmaj Cm/Eb Dbmaj7 Cmaj Ebmaj Abmaj/Eb Gmaj/Eb Abmaj
Dbmaj Gbmaj

This is a bit "out" harmonically considering you were telling me a while ago
that the note F# does exist in the key of G major, and then you go from Gmaj
to Dbmaj7 like that.

I think what I really want to see is for you to run this program again with
the set of all 7-limit pitches and hear what amazing chord progressions
we're missing.

This is getting spammed on Facebook.
-Mike

On Wed, Dec 9, 2009 at 12:47 AM, Marcel de Velde <m.develde@...>wrote:

>
>
> I wrote an algorithm based on my harmonic permutation JI theory.
> Still very simple, I will add many more things to the algorithm over time.
> But I allready really enjoy it's output.
>
> 5-limit permutation just intonation sung by computer
> choir.<
> http://sites.google.com/site/develdenet/mp3/9-12-2009_5limit_harmonic-permutation-ji_choir.mp3?attredirects=0
> >
>
> Should the link not work (which sometimes seems to be the case with google
> pages direct links) you can find the mp3 at www.develde.net
>
> Hope you enjoy!
>
> Marcel
>
> [Non-text portions of this message have been removed]
>
>
>

[Non-text portions of this message have been removed]

Hi Kalle,

Wow Marcel,
>
> this sounds really nice to my ears! Not at all what I expected as I find
> most algorithmic music I've heard disappointing. This doesn't sound
> particularly random at all, not even as a whole which is really surprising.
> I haven't been that impressed with your retunings of existing pieces but it
> seems that your theory has a lot of potential for creating new music. I'd be
> really interested to see the details of your algorithm and how they relate
> to your theories.
>
> Kalle
>

Thank you! :)
I'll reveal all the details of my theory and how I aplied it so far to the
algorithm soon.
I'd paste it here now but it will be hard to make sense of for people so I
figured it'd be better to set up a proper website with all the info of my
theory, and a running version of the algorithm where one can control many
parameters and get a scala sequence file as the output (and pitch tuned midi
as well later in the future).
But first nex days I'm going to implement a few more things in the algorithm
and make a few more songs.
Can't do everything at thesame time but I'll try to hurry with the website.

Marcel

[Non-text portions of this message have been removed]

Hi Mike,

This is absolutely amazing. How did you do this? Is it constantly modulating
> between related keys at random, just keeping one tone common each time?
>

Thank you! :)
Yes it's modulating at random.
It's indeed keeping atleast one tone common between consecutive harmonies
right now, but this is only because it was easyer to write to start with.
I'm diving a bit deeper into the permuation math right now and making a
version that doesn't nessecarily keep a tone common between consecutive
harmonies.

>
> Right off the bat you go Gmaj Cm/Eb Dbmaj7 Cmaj Ebmaj Abmaj/Eb Gmaj/Eb
> Abmaj
> Dbmaj Gbmaj
>

Oh you have good ears :)

>
> This is a bit "out" harmonically considering you were telling me a while
> ago
> that the note F# does exist in the key of G major, and then you go from
> Gmaj
> to Dbmaj7 like that.
>

Ah yes no that was by an old definition of tonic.
Also with that old definition thesame chord progression was possible, only I
would have said it was not in one tonic.
The old tonic definition was 6-limit permutations from the lowest voice
(lowest voice stays fixed)
This gives the 1/1 9/8 6/5 5/4 4/3 3/2 8/5 5/3 9/5 15/8 2/1 scale when
reduced to one octave.

My new definition of tonic is all permutations from a limited harmonic
series.
So 5-limit = 1/1 2/1 3/1 4/1 5/1
And all permutations of these do not only permutate from the 1/1 but also
from 2/1, 3/1 4/1 and 5/1.
This gives 1/1 16/15 9/8 6/5 5/4 4/3 45/32 3/2 8/5 5/3 9/5 15/8 2/1 as a
scale when reduced to one octave.
Full 12 tone scale from the tonic 1/1! It's amazing :)

>
> I think what I really want to see is for you to run this program again with
> the set of all 7-limit pitches and hear what amazing chord progressions
> we're missing.
>

I wish to hear this myself too, but i have a lot more work programming
permutation rules before I can do this.
6-limit running randomly allready gives too many chromatic steps making it
less tonal. I know rules that prevent this but have to write them.
7-limit is much harder, i don't know yet how to write permutation rules for
this that would make sense.
I've rendered a 7-limit output allready but it sounds like atonal random
nonsense right now, not very interesting.
Though i have good hopes for 7-limit in the future! :) But first get 5-limit
written properly, then 6-limit, and then (with a lot of thinking) 7-limit.

>
> This is getting spammed on Facebook.
> -Mike
>

Oh cool :)
A facebook group you mean? Can you give me the link?

Marcel

[Non-text portions of this message have been removed]

Marcel,

By permutation do you mean...of inversions or what?

>"This gives 1/1 16/15 9/8 6/5 5/4 4/3 45/32 3/2 8/5 5/3 9/5 15/8 2/1 as a
scale when reduced to one octave."

Beside having one note always in common between chord changes...what constitutes a valid chord IE
A) Does the each following chord need to have a root with a common denominator as the one before it?
B) Does any part of the following chord (IE one, two, three...notes) have to be an inversion of the chord before it?
C) It seems all the fractions in your scale have either a numerator and/or denominator that is a power of 3 or 2...do you take advantage of that in any way?
D) etc.

Anyhow...it sounds fantastic...seems to change the distance between notes in the chords very little between chords for easy/smooth listening and yet cover many tonal colors over very little time.
If there's any way I could have a copy of the code as well I would love to look it over.

________________________________
From: Marcel de Velde <m.develde@...>
To: MakeMicroMusic@yahoogroups.com
Sent: Wed, December 9, 2009 2:03:54 PM
Subject: Re: [MMM] My first composition.

Hi Mike,

This is absolutely amazing. How did you do this? Is it constantly modulating
> between related keys at random, just keeping one tone common each time?
>

Thank you! :)
Yes it's modulating at random.
It's indeed keeping atleast one tone common between consecutive harmonies
right now, but this is only because it was easyer to write to start with.
I'm diving a bit deeper into the permuation math right now and making a
version that doesn't nessecarily keep a tone common between consecutive
harmonies.

>
> Right off the bat you go Gmaj Cm/Eb Dbmaj7 Cmaj Ebmaj Abmaj/Eb Gmaj/Eb
> Abmaj
> Dbmaj Gbmaj
>

Oh you have good ears :)

>
> This is a bit "out" harmonically considering you were telling me a while
> ago
> that the note F# does exist in the key of G major, and then you go from
> Gmaj
> to Dbmaj7 like that.
>

Ah yes no that was by an old definition of tonic.
Also with that old definition thesame chord progression was possible, only I
would have said it was not in one tonic.
The old tonic definition was 6-limit permutations from the lowest voice
(lowest voice stays fixed)
This gives the 1/1 9/8 6/5 5/4 4/3 3/2 8/5 5/3 9/5 15/8 2/1 scale when
reduced to one octave.

My new definition of tonic is all permutations from a limited harmonic
series.
So 5-limit = 1/1 2/1 3/1 4/1 5/1
And all permutations of these do not only permutate from the 1/1 but also
from 2/1, 3/1 4/1 and 5/1.
This gives 1/1 16/15 9/8 6/5 5/4 4/3 45/32 3/2 8/5 5/3 9/5 15/8 2/1 as a
scale when reduced to one octave.
Full 12 tone scale from the tonic 1/1! It's amazing :)

>
> I think what I really want to see is for you to run this program again with
> the set of all 7-limit pitches and hear what amazing chord progressions
> we're missing.
>

I wish to hear this myself too, but i have a lot more work programming
permutation rules before I can do this.
6-limit running randomly allready gives too many chromatic steps making it
less tonal. I know rules that prevent this but have to write them.
7-limit is much harder, i don't know yet how to write permutation rules for
this that would make sense.
I've rendered a 7-limit output allready but it sounds like atonal random
nonsense right now, not very interesting.
Though i have good hopes for 7-limit in the future! :) But first get 5-limit
written properly, then 6-limit, and then (with a lot of thinking) 7-limit.

>
> This is getting spammed on Facebook.
> -Mike
>

Oh cool :)
A facebook group you mean? Can you give me the link?

Marcel

[Non-text portions of this message have been removed]

[Non-text portions of this message have been removed]

Hi Micheal,

> By permutation do you mean...of inversions or what?
>

Permutation is all possible configurations of the intervals making up the
harmonic series.
So 3-limit is 1/1 2/1 3/1
It has the following permutations:
1/1 2/1 3/1
1/1 3/2 3/1 (change of interval order centered on 1/1)

1/1 2/1 3/1
4/3 2/1 4/1 (change of interval order centered on 2/1)

1/1 2/1 3/1
1/1 3/2 3/1 (change of interval order centered on 3/1)

It's not an inversion, for instance 4 limit is 1/1 2/1 3/1 4/1, a
permutation is for instance 1/1 3/2 3/1 4/1.
Inversions make use of octave equivalence, this is not yet in my theory. I
mean offcourse one can use octave equivalence but it has a few consequences
which I have not yet worked out in permutation rules.

>
>
> >"This gives 1/1 16/15 9/8 6/5 5/4 4/3 45/32 3/2 8/5 5/3 9/5 15/8 2/1 as a
> scale when reduced to one octave."
>
> Beside having one note always in common between chord changes...what
> constitutes a valid chord IE
>

No no it doesn't need to have a note allways in common between chord
changes.
That's just an easy trick I used to write the algorithm fast.
I'm writing now the better permutation rules that also give certain chord
changes that have no note in common, yet still all voices move diatonically.

> A) Does the each following chord need to have a root with a common
> denominator as the one before it?
> B) Does any part of the following chord (IE one, two, three...notes) have
> to be an inversion of the chord before it?
> C) It seems all the fractions in your scale have either a numerator and/or
> denominator that is a power of 3 or 2...do you take advantage of that in any
> way?
> D) etc.
>

I will explain everything indepth soon in a new website I'm designing.
I can't explain it all right now in an email sorry.

>
> Anyhow...it sounds fantastic...seems to change the distance between notes
> in the chords very little between chords for easy/smooth listening and yet
> cover many tonal colors over very little time.
> If there's any way I could have a copy of the code as well I would love to
> look it over.

Thanks! :)
I just posted the code on the tuning list.
But it's messy and not a good representation of or good way to learn what
it's really about, better to wait a few weeks for the website where it'll
all be explained in an easy to understand way.

Marcel

[Non-text portions of this message have been removed]

Marcel, for computer generated creativity this is simply outstanding.

- Is the point that every "true JI" chord can follow another (to modern ears) and still have beauty?

- Could be cool to work your algorithm from a traditional cantus firmus

What do you use for the choir? Sounds nice too.

Thanks,

Chris

--- In MakeMicroMusic@yahoogroups.com, Marcel de Velde <m.develde@...> wrote:
>
> I wrote an algorithm based on my harmonic permutation JI theory.
> Still very simple, I will add many more things to the algorithm over time.
> But I allready really enjoy it's output.
>
> 5-limit permutation just intonation sung by computer
> choir.<http://sites.google.com/site/develdenet/mp3/9-12-2009_5limit_harmonic-permutation-ji_choir.mp3?attredirects=0>
>
> Should the link not work (which sometimes seems to be the case with google
> pages direct links) you can find the mp3 at www.develde.net
>
> Hope you enjoy!
>
> Marcel
>
>
> [Non-text portions of this message have been removed]
>

Hi Chris,

Marcel, for computer generated creativity this is simply outstanding.
>
Thank you too :)
I'm amazed that there's universal praise up till now :)

>
> - Is the point that every "true JI" chord can follow another (to modern
> ears) and still have beauty?
>

No.
It works because these are mathematical rules that explore the possibilities
of true JI chords following eachother.
There are rules which chord can follow which chord.
For instance the C E G, C E A, D F A, D G B chord sequence my algorithm
doesn't do yet and it contains permutation combinations (D F A beeing
combination of a permutation containing F A and D comming from another
permutation.) I have not worked the rules out for such things.
In these rules which chord can follow which chord there are also rules that
make each voice move "diatonically" (for a lack of a better word). And in my
algorithm there's no crossing of the voices.
This is all the algorithm does right now.
Though I doesn't really do this correctly yet because there's allways a held
note at the moment. I'm making a version today which doesn't need to hold a
note anymore.
My new version will also allow for instance the following chord sequence:
v1 v2 v3 v4 v5
1/1 2/1 3/1 4/1 5/1
4/3 2/1 8/3 10/3 20/3
3/2 9/4 3/1 15/4 15/2
1/1 2/1 3/1 4/1 5/1

The mp3 I posted doesn't do this yet.

> - Could be cool to work your algorithm from a traditional cantus firmus
>

Yes I thought thesame too :)
It will work. But it's easyer to first put the cantus firmus in JI and make
a choice incase there are several JI interpretations which one to use.

>
> What do you use for the choir? Sounds nice too.
>

I've downloaded a "free try before you buy" version of Native Instruments
Komplete.
The choir is in the library that comes with it for Kontakt 4. I simply ran 5
instances of it so it does all the voices correctly with pitchbend tuning.
I don't have any money now to buy anything, but saving up for Synful. Looks
like it can be tuned by pitchbends too.

Marcel

[Non-text portions of this message have been removed]

maybe because your music sounds better than your theories?
Sorry Marcel, I could not resist!!!!
:-)

--- In MakeMicroMusic@yahoogroups.com, Marcel de Velde <m.develde@...> wrote:
>
> I'm amazed that there's universal praise up till now :)

that's nice indeed! I would add more rhythmic variety, though!

--- In MakeMicroMusic@yahoogroups.com, Marcel de Velde <m.develde@...> wrote:
>
> I wrote an algorithm based on my harmonic permutation JI theory.
> Still very simple, I will add many more things to the algorithm over time.
> But I allready really enjoy it's output.

> maybe because your music sounds better than your theories?
> Sorry Marcel, I could not resist!!!!
> :-)
>

Hehe :)
No but it's really the other way around.
It's a lucky accident that it sounds so nice while doing things so
completely random.
My theory actually goes much deeper and gives non random algorithmic control
over all the voices and harmonies etc.
In the future there could still be for instance a random seed at the start
of a song for several parameters. But this will give a very diffirent song
every time, while right now the algorithm is so simple that the output will
sound fairly similar every time it is run.
I guess the reason why it kinda works right now while doing things
completely random is because the possibility set is rather small.
Later when the possiblities are much larger it will require more control I
think, but this is good because writing control is exactly the thing this
algorithm / my theory is good at :)

Marcel

[Non-text portions of this message have been removed]

>"maybe because your music sounds better than your theories?
Sorry Marcel, I could not resist!!!!
:-)"

Funny thing, I've noticed within the past month Marcel's theories have gotten notably better...so I doubt it's just the composition method here that makes things sound so smooth.

The other thing...I actually prefer Marcel's original work...to his re-tunings of classical music. But I doubt that is coincidence: IMVHO, each truly new tuning has its own mood and, as such, does best with songs made for it rather than simple re-tunings of already made music.

________________________________
From: Carlo <carlo@...>
To: MakeMicroMusic@yahoogroups.com
Sent: Thu, December 10, 2009 7:54:18 AM
Subject: Re: [MMM] My first composition.

maybe because your music sounds better than your theories?
Sorry Marcel, I could not resist!!!!
:-)

--- In MakeMicroMusic@ yahoogroups. com, Marcel de Velde <m.develde@. ..> wrote:
>
> I'm amazed that there's universal praise up till now :)

[Non-text portions of this message have been removed]

Just to make it more clear:

There are rules which chord can follow which chord.
> For instance the C E G, C E A, D F A, D G B chord sequence my algorithm
> doesn't do yet and it contains permutation combinations (D F A beeing
> combination of a permutation containing F A and D comming from another
> permutation.) I have not worked the rules out for such things.
> In these rules which chord can follow which chord there are also rules that
> make each voice move "diatonically" (for a lack of a better word). And in my
> algorithm there's no crossing of the voices.
> This is all the algorithm does right now.
> Though I doesn't really do this correctly yet because there's allways a
> held note at the moment. I'm making a version today which doesn't need to
> hold a note anymore.
> My new version will also allow for instance the following chord sequence:
> v1 v2 v3 v4 v5
> 1/1 2/1 3/1 4/1 5/1
> 4/3 2/1 8/3 10/3 20/3
> 3/2 9/4 3/1 15/4 15/2
> 1/1 2/1 3/1 4/1 5/1
>

The C E G, C E A (or C F A), D F A, D G B sequence which is a simple examle
of the comma shift problem.
My theory is a mathematic framework of what's possible and what isn't.
My harmonic permutation JI sais there is no possible solution for the C E G,
C E A (or C F A), D F A, D G B chord sequence without combining
permutations. In other words, it's impossible to tune in pure low ratio JI.
To give an example:
C (1/1) C (2/1) G (3/1) C (4/1) E (5/1) - the C E G chord
F (4/3) C (2/1) F (8/3) A (10/3) A (20/3) - the C F A chord (can also
make it E (5/2) instead of F (8/3) makes no difference)
Bb (8/9) Bb (16/9) F (8/3) A (10/3) D (40/9) - a possibility of the D F
A chord, "consonant low-ji" version. and a modulation.
Now to go from here to any chord containing D G B is not possible! There is
no permutation solution which will give a chord containing a D, G and B!
One can try all permutations pssible to see if there's a solution in there,
but there isn't. This chord sequence is simply not a low ratio single
permutation structure chord sequence (sorry for all the silly names)
Only with combined permutations does this sequence become possible, and for
instance the D F A chord become 9/8 4/3 5/3.
There is no "comma drifting" solution either.
(btw all the above examples do not use octave equivalence, offcourse one can
use octave equivalence, but this has consequences for certain things and I
have not worked out everything yet (which may or may not become a very
difficult thing i don't know yet have to think about it))

My theory works because it contains in it the inherent logic of what's
possible mathematically, and it has a way to work with this.

Marcel

[Non-text portions of this message have been removed]

Marcel,

>"Only with combined permutations does this sequence become possible, and for
instance the D F A chord become 9/8 4/3 5/3."

So let me get this right
A) combined permutations result in chords with differing greatest common multiple denominators (IE your above example has 9/8 where 8 = 2^(a power) and 4/3 where 3 = 3^(a power) but not 2^(a power)?

B) F (8/3) A (10/3) D (40/9) (the lowest denominator version possible in your scale that has a common GCD across all fractions) must be re-arranged in a combined permutation to make the above (and again, I'm having a tough time figuring out what's different between a "permutation combination" and a simple inversion)?

> "And in my algorithm there's no crossing of the voices."
So in other words each note in the second chord can't be higher than the note over that note in the first chord OR lower than the note under that note in the first chord (thus making the chord have to either "slide" either up or down or consendense/expand to an extent, but never do both at the same time to a huge extent)?

If so it would make sense, I remember Dmitri Tymoczko (professor at Princeton University) figured out that good musicians naturally pick chords where the average movements between notes in a chord and the nearest notes in the last chords are minimized...and that minimization makes music easier to listen to. You guys might be on the same track....

________________________________
From: Marcel de Velde <m.develde@...>
To: MakeMicroMusic@yahoogroups.com
Sent: Thu, December 10, 2009 8:59:51 AM
Subject: Re: [MMM] My first composition.

Just to make it more clear:

There are rules which chord can follow which chord.
> For instance the C E G, C E A, D F A, D G B chord sequence my algorithm
> doesn't do yet and it contains permutation combinations (D F A beeing
> combination of a permutation containing F A and D comming from another
> permutation. ) I have not worked the rules out for such things.
> In these rules which chord can follow which chord there are also rules that
> make each voice move "diatonically" (for a lack of a better word). And in my
> algorithm there's no crossing of the voices.
> This is all the algorithm does right now.
> Though I doesn't really do this correctly yet because there's allways a
> held note at the moment. I'm making a version today which doesn't need to
> hold a note anymore.
> My new version will also allow for instance the following chord sequence:
> v1 v2 v3 v4 v5
> 1/1 2/1 3/1 4/1 5/1
> 4/3 2/1 8/3 10/3 20/3
> 3/2 9/4 3/1 15/4 15/2
> 1/1 2/1 3/1 4/1 5/1
>

The C E G, C E A (or C F A), D F A, D G B sequence which is a simple examle
of the comma shift problem.
My theory is a mathematic framework of what's possible and what isn't.
My harmonic permutation JI sais there is no possible solution for the C E G,
C E A (or C F A), D F A, D G B chord sequence without combining
permutations. In other words, it's impossible to tune in pure low ratio JI.
To give an example:
C (1/1) C (2/1) G (3/1) C (4/1) E (5/1) - the C E G chord
F (4/3) C (2/1) F (8/3) A (10/3) A (20/3) - the C F A chord (can also
make it E (5/2) instead of F (8/3) makes no difference)
Bb (8/9) Bb (16/9) F (8/3) A (10/3) D (40/9) - a possibility of the D F
A chord, "consonant low-ji" version. and a modulation.
Now to go from here to any chord containing D G B is not possible! There is
no permutation solution which will give a chord containing a D, G and B!
One can try all permutations pssible to see if there's a solution in there,
but there isn't. This chord sequence is simply not a low ratio single
permutation structure chord sequence (sorry for all the silly names)
Only with combined permutations does this sequence become possible, and for
instance the D F A chord become 9/8 4/3 5/3.
There is no "comma drifting" solution either.
(btw all the above examples do not use octave equivalence, offcourse one can
use octave equivalence, but this has consequences for certain things and I
have not worked out everything yet (which may or may not become a very
difficult thing i don't know yet have to think about it))

My theory works because it contains in it the inherent logic of what's
possible mathematically, and it has a way to work with this.

Marcel

[Non-text portions of this message have been removed]

[Non-text portions of this message have been removed]

Hi Michael,

>"Only with combined permutations does this sequence become possible, and
> for
> instance the D F A chord become 9/8 4/3 5/3."
>
> So let me get this right
> A) combined permutations result in chords with differing greatest common
> multiple denominators (IE your above example has 9/8 where 8 = 2^(a power)
> and 4/3 where 3 = 3^(a power) but not 2^(a power)?
>

I haven't yet figured out exactly how to use combined permutations.
But one can say a permutation is a 1/1 2/1 3/1 4/1 5/1 chord only with the
order of the intervals shuffled in some way.
All possible permutations of 1/1 2/1 3/1 4/1 5/1 give the scale 1/1 16/15
9/8 6/5 5/4 4/3 45/32 3/2 8/5 5/3 9/5 15/8 2/1 when reduced to one octave.
By combined permutation i mean a chord that has notes in it that belong to
different permutations.
So by defenition this means that this chord can only have notes from the 1/1
16/15 9/8 6/5 5/4 4/3 45/32 3/2 8/5 5/3 9/5 15/8 2/1 scale.
So for instance 9/8 4/3 5/3 becomes possible, where 4/3 and 5/3 belong to
one permuation and 9/8 belongs to another permutation (for instance the V
chord that comes after the 4/3 5/3 chord)

> B) F (8/3) A (10/3) D (40/9) (the lowest denominator version possible in
> your scale that has a common GCD across all fractions) must be re-arranged
> in a combined permutation to make the above (and again, I'm having a tough
> time figuring out what's different between a "permutation combination" and a
> simple inversion)?
>

Forget permutation combinations for now :)
Normal permutations are better to focus on first, I'm not bother myself with
permutation combinations yet.
An inversion is something completely different. It works by simply shifting
notes of a chord up or down by an octave.
One can do this aswell (octave equivalent shifting) with the notes that come
from the permutations (as long as one remembers and works with the original
permutation structure and doesn't reinterpret octave shifted notes as a
permutation chord which would often result in not allowed permutations after
that) although this will offcourse have some effect on voices, I haven't
worked out additional rules for this.
Btw I'm not minding gcd or anything like that at all, it doesn't have any
use really in working with my theory.

> > "And in my algorithm there's no crossing of the voices."
> So in other words each note in the second chord can't be higher than the
> note over that note in the first chord OR lower than the note under that
> note in the first chord (thus making the chord have to either "slide" either
> up or down or consendense/expand to an extent, but never do both at the same
> time to a huge extent)?
>

The voices can become higher or lower than it's neighbour voice previously
was, in a next chord, but then the neighbouring voice will drop aswell.
What I mean is that for instance the highest voice will allways stay the
highest voice and can't yet walk all the way down to become the lowest
voice.

One thing to easily work out a permutation possiblity is like this:
Take the 1/1 2/1 3/1 4/1 5/1 chord.
it has the intervals 2/1,then 3/2 then 4/3 then 5/4
A permutation is to simply shuffle the order of these intervals, so first
3/2, then 5/4, then 2/1, then 4/3 for instance.
Now this new permutated chord can be relevant to the 1/1 (1/1 stays thesame)
or 2/1 (2/1 stays thesame) etc.
When relevant to 2/1 you get the following chord sequence:
1/1 2/1 3/1 4/1 5/1
4/3 2/1 5/2 5/1 20/3

Now the way my algorithm did the mp3 was simply by doing thesame permutation
thing again and certering it on any of the 5 tones (so keeping atleast one
voice in common)
Please do not that voice one can center on voice one, voice 2 can center on
voice 2, but voice 1 can't center on voice 2!

So for instance:
1/1 (2/1) 3/1 4/1 5/1
4/3 (2/1) (5/2) 5/1 20/3
1/1 5/4 (5/2) 15/4 5/1

(to avoid confusion, note that even though 5/1 is on each of the chords, it
is first voice 5 that is 5/1, then voice 4, then voice 5 again. so there is
no voice that plays 5/1 constantly)

The way my mp3 worked wasn't complete.
There are additional permutation possibilities which I'll explain now.
What I should have written (and will write later today) was this:

Instead of going directly from 4/3 2/1 5/2 5/1 20/3 to 1/1 5/4 5/2 15/4 5/1
by a permutation centered on voice 3: 5/2
One should really go from 4/3 2/1 5/2 5/1 20/3 to a 1/1 2/1 3/1 4/1 5/1
chord first (but not play it!) which is centered on any of the voices of the
previous chord. And THEN do a permutation of this 1/1 2/1 3/1 4/1 5/1 chord
(which IS played (and can offcourse beceme 1/1 2/1 3/1 4/1 5/1 again as
itself as one of the possiblities).
This then is a correct purmutation, it includes all of the results of the
previous method, and adds several other possibilities where there is not a
note connected.
So for instance
1/1 (2/1) 3/1 4/1 5/1
4/3 (2/1) (5/2) 5/1 20/3
-> [15/16 15/8 5/2 15/4 75/16] (this is a 1/1 2/1 3/1 4/1 5/1 chord
centered randomly on 5/2, it is NOT played it is simply for working out the
permutation)
1/1 5/4 (5/2) (15/4) 5/1 (one of the possible permutations of [15/16 15/8
5/2 15/4 75/16] centered on for instance voice 3 or 4)
or alternatively (15/16) (15/8) 45/16 (15/4) (75/16) (see this one has no
note in common with 4/3 2/1 5/2 5/1 20/3)

Marcel

[Non-text portions of this message have been removed]

Oef sorry a little mistake here!

> 1/1 (2/1) 3/1 4/1 5/1
> 4/3 (2/1) (5/2) 5/1 20/3
> -> [15/16 15/8 5/2 15/4 75/16] (this is a 1/1 2/1 3/1 4/1 5/1 chord
> centered randomly on 5/2, it is NOT played it is simply for working out the
> permutation)
> 1/1 5/4 (5/2) (15/4) 5/1 (one of the possible permutations of [15/16 15/8
> 5/2 15/4 75/16] centered on for instance voice 3 or 4)
> or alternatively (15/16) (15/8) 45/16 (15/4) (75/16) (see this one has no
> note in common with 4/3 2/1 5/2 5/1 20/3)

Offcourse not [15/16 15/8 5/2 15/4 75/16] and (15/16) (15/8) 45/16 (15/4)
(75/16) ugh.
That should have been for instance [5/6 5/3 5/2 10/3 25/6] and (5/6) (5/3)
20/9 (10/3) (25/6)

So:

1/1 (2/1) 3/1 4/1 5/1
4/3 (2/1) (5/2) 5/1 20/3
-> [5/6 5/3 5/2 10/3 25/6] (this is a 1/1 2/1 3/1 4/1 5/1 chord centered
randomly on 5/2, it is NOT played it is simply for working out the
permutation)
1/1 5/4 (5/2) (15/4) 5/1 (one of the possible permutations of [5/6 5/3 5/2
10/3 25/6] centered on for instance voice 3 or 4)
or alternatively (5/6) (5/3) 20/9 (10/3) (25/6) (see this one has no note in
common with 4/3 2/1 5/2 5/1 20/3)

Marcel

[Non-text portions of this message have been removed]

Marcel>"One thing to easily work out a permutation possiblity is like this:"

I at least think I get this part...you appear to achieve new chords by stacking the intervals in different orders using on note as the root.
In a way...this works in a similar way as in going from a major chord to minor chord with a root based on a note in the previous major chord (where only the order of the intervals between notes switches and nothing is inverted).

>"The voices can become higher or lower than it's neighbour voice previously
was, in a next chord, but then the neighbouring voice will drop aswell.
What I mean is that for instance the highest voice will allways stay the
highest voice and can't yet walk all the way down to become the lowest
voice."

Right, so again (it seems) you and Dimitri are on the same page; desiring an algorithm that aligns voices so the distance a note must travel between chords is minimal (as neighboring voices must stay in the same order).

I think I get a lot of it now and, for sure, your research could have many implications...not just for how scales can work ideally...but also how composition style can be derived ideally to match a scale. Hopefully sometime in the future your scale -> chord composition possibilities research can, say, be combined with Sethares scale -> timbre research to make something both more accessible and easier to compose with than anything the music world has seen before.

-Michael

________________________________
From: Marcel de Velde <m.develde@...>
To: MakeMicroMusic@yahoogroups.com
Sent: Thu, December 10, 2009 10:25:03 AM
Subject: Re: [MMM] My first composition.

Hi Michael,

>"Only with combined permutations does this sequence become possible, and
> for
> instance the D F A chord become 9/8 4/3 5/3."
>
> So let me get this right
> A) combined permutations result in chords with differing greatest common
> multiple denominators (IE your above example has 9/8 where 8 = 2^(a power)
> and 4/3 where 3 = 3^(a power) but not 2^(a power)?
>

I haven't yet figured out exactly how to use combined permutations.
But one can say a permutation is a 1/1 2/1 3/1 4/1 5/1 chord only with the
order of the intervals shuffled in some way.
All possible permutations of 1/1 2/1 3/1 4/1 5/1 give the scale 1/1 16/15
9/8 6/5 5/4 4/3 45/32 3/2 8/5 5/3 9/5 15/8 2/1 when reduced to one octave.
By combined permutation i mean a chord that has notes in it that belong to
different permutations.
So by defenition this means that this chord can only have notes from the 1/1
16/15 9/8 6/5 5/4 4/3 45/32 3/2 8/5 5/3 9/5 15/8 2/1 scale.
So for instance 9/8 4/3 5/3 becomes possible, where 4/3 and 5/3 belong to
one permuation and 9/8 belongs to another permutation (for instance the V
chord that comes after the 4/3 5/3 chord)

> B) F (8/3) A (10/3) D (40/9) (the lowest denominator version possible in
> your scale that has a common GCD across all fractions) must be re-arranged
> in a combined permutation to make the above (and again, I'm having a tough
> time figuring out what's different between a "permutation combination" and a
> simple inversion)?
>

Forget permutation combinations for now :)
Normal permutations are better to focus on first, I'm not bother myself with
permutation combinations yet.
An inversion is something completely different. It works by simply shifting
notes of a chord up or down by an octave.
One can do this aswell (octave equivalent shifting) with the notes that come
from the permutations (as long as one remembers and works with the original
permutation structure and doesn't reinterpret octave shifted notes as a
permutation chord which would often result in not allowed permutations after
that) although this will offcourse have some effect on voices, I haven't
worked out additional rules for this.
Btw I'm not minding gcd or anything like that at all, it doesn't have any
use really in working with my theory.

> > "And in my algorithm there's no crossing of the voices."
> So in other words each note in the second chord can't be higher than the
> note over that note in the first chord OR lower than the note under that
> note in the first chord (thus making the chord have to either "slide" either
> up or down or consendense/ expand to an extent, but never do both at the same
> time to a huge extent)?
>

The voices can become higher or lower than it's neighbour voice previously
was, in a next chord, but then the neighbouring voice will drop aswell.
What I mean is that for instance the highest voice will allways stay the
highest voice and can't yet walk all the way down to become the lowest
voice.

One thing to easily work out a permutation possiblity is like this:
Take the 1/1 2/1 3/1 4/1 5/1 chord.
it has the intervals 2/1,then 3/2 then 4/3 then 5/4
A permutation is to simply shuffle the order of these intervals, so first
3/2, then 5/4, then 2/1, then 4/3 for instance.
Now this new permutated chord can be relevant to the 1/1 (1/1 stays thesame)
or 2/1 (2/1 stays thesame) etc.
When relevant to 2/1 you get the following chord sequence:
1/1 2/1 3/1 4/1 5/1
4/3 2/1 5/2 5/1 20/3

Now the way my algorithm did the mp3 was simply by doing thesame permutation
thing again and certering it on any of the 5 tones (so keeping atleast one
voice in common)
Please do not that voice one can center on voice one, voice 2 can center on
voice 2, but voice 1 can't center on voice 2!

So for instance:
1/1 (2/1) 3/1 4/1 5/1
4/3 (2/1) (5/2) 5/1 20/3
1/1 5/4 (5/2) 15/4 5/1

(to avoid confusion, note that even though 5/1 is on each of the chords, it
is first voice 5 that is 5/1, then voice 4, then voice 5 again. so there is
no voice that plays 5/1 constantly)

The way my mp3 worked wasn't complete.
There are additional permutation possibilities which I'll explain now.
What I should have written (and will write later today) was this:

Instead of going directly from 4/3 2/1 5/2 5/1 20/3 to 1/1 5/4 5/2 15/4 5/1
by a permutation centered on voice 3: 5/2
One should really go from 4/3 2/1 5/2 5/1 20/3 to a 1/1 2/1 3/1 4/1 5/1
chord first (but not play it!) which is centered on any of the voices of the
previous chord. And THEN do a permutation of this 1/1 2/1 3/1 4/1 5/1 chord
(which IS played (and can offcourse beceme 1/1 2/1 3/1 4/1 5/1 again as
itself as one of the possiblities) .
This then is a correct purmutation, it includes all of the results of the
previous method, and adds several other possibilities where there is not a
note connected.
So for instance
1/1 (2/1) 3/1 4/1 5/1
4/3 (2/1) (5/2) 5/1 20/3
-> [15/16 15/8 5/2 15/4 75/16] (this is a 1/1 2/1 3/1 4/1 5/1 chord
centered randomly on 5/2, it is NOT played it is simply for working out the
permutation)
1/1 5/4 (5/2) (15/4) 5/1 (one of the possible permutations of [15/16 15/8
5/2 15/4 75/16] centered on for instance voice 3 or 4)
or alternatively (15/16) (15/8) 45/16 (15/4) (75/16) (see this one has no
note in common with 4/3 2/1 5/2 5/1 20/3)

Marcel

[Non-text portions of this message have been removed]

[Non-text portions of this message have been removed]

Marcel,
This is a great start. Keep at it. Wonderful sonorities and some really wonderful changes. The art will be to find out which ones work, and which ones are better left on the cutting room floor.

Prent Rodgers

--- In MakeMicroMusic@yahoogroups.com, Marcel de Velde <m.develde@...> wrote:
>
> I wrote an algorithm based on my harmonic permutation JI theory.
> Still very simple, I will add many more things to the algorithm over time.
> But I allready really enjoy it's output.
>
> 5-limit permutation just intonation sung by computer
> choir.<http://sites.google.com/site/develdenet/mp3/9-12-2009_5limit_harmonic-permutation-ji_choir.mp3?attredirects=0>
>
> Should the link not work (which sometimes seems to be the case with google
> pages direct links) you can find the mp3 at www.develde.net
>
> Hope you enjoy!
>
> Marcel
>
>
> [Non-text portions of this message have been removed]
>

Hi Prent,

> Marcel,
> This is a great start. Keep at it.
>
Thanks!

Wonderful sonorities and some really wonderful changes. The art will be to
> find out which ones work, and which ones are better left on the cutting room
> floor.
>
Yes agreed.
I think with a bit more work on the algorithm (allready improved it, it now
truly allows all permutation possibilities of a harmonic limit), bringing in
controllable (semi random) rhythm, use of octave inversion, something with a
fundamental bass, and allowing controllable combination permutations.
And then less completely random but more using random seeds at the start for
logical permutation sequences / subalgorithms which could also bring in good
repititions / logical structure.
And a way to cut and past from the output and also cut and paste only a
melody in certain parts for instance where the algorithm then generates
harmonies again.
When this all is done I think it would make the best computer assisted
composition / inspiration tool ever :)
One where one works with the algorithms on an as easy or deep a level as one
wants, and pick out things you like etc.
I had this vision allready some time ago but now see the path to make it
reality.

Marcel

[Non-text portions of this message have been removed]

--- In MakeMicroMusic@yahoogroups.com, Michael <djtrancendance@...> wrote:
>
> >"maybe because your music sounds better than your theories?
> Sorry Marcel, I could not resist!!!!

I have to confess a similar thought crossed my mind...

> :-)"
>
> Funny thing, I've noticed within the past month Marcel's theories
> have gotten notably better...so I doubt it's just the composition
> method here that makes things sound so smooth.
>

The theories as such did not even sound that bad to me - what did sound bad were the bold claims made around them, such as them being "the correct solution", others being "invalid" and so on. And the resulting flame wars that tended to develop.

But putting that aside: It really sounds good, I join the praise. I will be interested to hear what happens when more of the theory is put into the algorithm. And a join Mike's wish for the same in 7 limit.
--
Hans Straub

Hello Hans,

The theories as such did not even sound that bad to me - what did sound bad
> were the bold claims made around them, such as them being "the correct
> solution", others being "invalid" and so on. And the resulting flame wars
> that tended to develop.
>

Yes my apologies for my behaviour in the past.
I didn't mean any bad but I had to learn how to behave on mailing lists :)

>
> But putting that aside: It really sounds good, I join the praise. I will be
> interested to hear what happens when more of the theory is put into the
> algorithm. And a join Mike's wish for the same in 7 limit.
>

I'm working on expanding and refining the algorithm. But it may take a bit
longer than I thought it would.
Some ideas I had on the way forward I allready come back from. Have to put
more thinking and experimenting in it.
I'll refrain from details, as the past has taught me that 4 out of 5 idears
I have turn out to be wrong, I won't bother this list with them.

As for 7-limit. The way the algorithm works now 7-limit simply sounds
terrible.
Random 7-limit permutations sound like a random atonal mess, with very
little musical feeling / emotion in it.
Even 6-limit is too random and chromatic for me now. I'm working on putting
some 5-limit limitations in 6-limit, but don't know yet what would be a good
way to do such a thing with 7-limit so a good sounding 7-limit algorithm may
take a loong time before it becomes reality.

Marcel

[Non-text portions of this message have been removed]

> As for 7-limit. The way the algorithm works now 7-limit simply sounds
> terrible.
> Random 7-limit permutations sound like a random atonal mess, with very
> little musical feeling / emotion in it.
> Even 6-limit is too random and chromatic for me now. I'm working on putting
> some 5-limit limitations in 6-limit, but don't know yet what would be a good
> way to do such a thing with 7-limit so a good sounding 7-limit algorithm may
> take a loong time before it becomes reality.

I've uploaded a 7-limit output anyhow so you can hear for yourself.
Halfway through the midi I've switched off 4 of the 7 voices, but it doesn't
make much of a difference. It's all a random mess with some terrible out of
tune sound to it.

/makemicromusic/files/Marcel/7-limit_harmonic-permutation_mess.mid

This doesn't mean to me that 7-limit doesn't work. But it does need rules,
and I don't know these yet.

Marcel

[Non-text portions of this message have been removed]

--- In MakeMicroMusic@yahoogroups.com, Marcel de Velde <m.develde@...> wrote:
>
> I've uploaded a 7-limit output anyhow so you can hear for yourself.
> Halfway through the midi I've switched off 4 of the 7 voices, but
> it doesn't make much of a difference. It's all a random mess with
> some terrible out of tune sound to it.
>
> /makemicromusic/files/Marcel/7-limit_harmonic-permutation_mess.mid
>
> This doesn't mean to me that 7-limit doesn't work. But it does need
> rules, and I don't know these yet.
>

Hmm, to my ears, it doesn't even sound that bad! For which, of course, the reason may be that my ears have been exposed to a certain amount of 20tc century music...
--
Hans Straub