< Back to IRCAM Forum

Chord-seq to poly

OpenMusic
#1

Hi there,

I’m kind of a newbie to OM, so the solution to my problem might be very simple.

I’m generating melodies (with rythm) in a Chord-seq and then use the diamanter of the OMTristan, it results in a second chord-seq with a few chords of harmonics. I’d like to bring them together in one poly object in order to export it in musicxml.

I could figure out how to do it using different voice objects and bringing them together in the second input of the poly object. But the chords I got with the diamanter are then reduced to one single note because going through the voice object. May someone help me please ?

Thanks a lot !

Best

#2

Hi,
Can you send an example patch? it will be easier to figure out what you’re doing wrong.

BEst
K

#3

Yes sure !

Here’s the last part of the patch…

Thanks

Colgen2.omp (39.3 KB)

#4

Ok

Nothing’s wrong with your patching. It’s unfortunately a bug.
I am sending you a workaround.
As you could notice also , that in your transformation all chords are of 1000ms which is also transcribed int the corrected patch. Now if you need a legato i think you need to specify it in the diamanter function . Not so sure, 'cause i didn’t load it.

colgen2-workaround.omp (73.1 KB)

#5

Happy to know it’s a bug and not me screwing it up ! Thanks a lot for your corrected version, I learned about a new object :slight_smile: ! I’ll build it in my main patch, it should be working !

#6

Hi Paddys, Karim

Actually, there is no bug. From function diamanter’s documentation: “The harmonic[s’] attacks are spread out between the attack of the original chord and the attack of the next. Diamanter is a term coined by Olivier Messiaen. As an example, Messiaen explained in his orchestration class how the high register of the piano may add brilliance to orchestral sound textures: le piano diamante lÕorchestre.”

And duration is specified as an input in diamanter (1000 by default).

All the best,
António

PS: I’ve attached a slightly modified version of your patch which does what you want. Hope this helps.

Colgen2-solution.omp (34.1 KB)

#7

Dear Antonio,

Sorry to insist but htere is a bug, and it is not at all related to the tristan library,
The bug comes from the chord-seq to voice connection and it’s autmatic method of heritance. It skips the chords and just places single notes. I hope this is clear to you now.

Best
K

#8

I vote for Antonio
Chord-seq to voice keeps the chords together when there is no note offsets.
Here, the diamanter function adds a (huge) offset to each note inside a chord, and the voice conversion places each note on a different pulse, so that the voice sounds approximately like the original chord-seq
All best

#9

ok
my bad
sorry