< Back to IRCAM Forum

Separating Large Clusters in chord-seq object

OpenMusic
#1

Hi all,

I’ve imported an SDIF file of a sound analysis from SPEAR and converted it into a chord-seq object. My current patch is SDIFFile - sdif->chord-seq - chord-seq. My only issue is the chord sequence is a huge unreadable cluster, and even after increasing the zoom and font size I can’t make out many of the notes: I have attached a picture below of the cluster I am interested in and my current patch.

I tried to sort it following the advice in this thread: http://forumnet.ircam.fr/user-groups/openmusic/forum/topic/om-how-to-order-huge-cluster-chords-from-sdif-analysis/, adding first - chord then changing the chord to ArpUp, however after doing this everytime I press v on the chord it just wants me to open another SDIF file and nothing happens when I do.

Does anyone know how I can sort out this cluster, either into an arpeggio or more manageable chords etc?
Many Thanks,

Steve

Patch-Proof.png

#2

Hello Steve,

There’s various ways for you to filter or cleanup the data before or after converting it to a CHORD-SEQ object.
The AS->OM function available in the RepMus library might a first start: this function does pretty much like SDIF->chord-seq but with several kinds of approximation/clustering/filtering which usually make the resulting data more readable.

Otherwise, please let us know what more precisely you would like to do: is this just about vizualizing better your cluster of notes, or reduce it to fewer notes ?

everytime I press v on the chord it just wants me to open another SDIF file and nothing happens when I do.
=> this is because your SDIFFILE box is not "locked" : it reloads the data everytime, and as its input is not connected, asks you which file to load. Just lock it with 'b'.

Hope this helps
—Jean

#3

Hi Jean,

Thanks very much! Apologies, I am new to OpenMusic and still finding my feet.

I’ve now created a patch with as->om, however have noticed that by default it is filtering out a lot notes. Can you point me in the direction of any tutorials where I can learn how to use as->om to it’s full capabilities?

What I would like to do with this material is visualise it better, not reduce it to fewer notes but somehow expand it so I can see all the clusters and see which occur first etc.
Thanks,

Steve

#4

Unfortunately there is no recent documentation that I know of about as->om, but you will find here in the user-manual of PW-Repmus (the ancestor of Repmus) the doc for the previous as->pw, which is pretty much equivalent.

http://support.ircam.fr/docs/om-libraries/old/PW-Repmus.pdf

The “Zoom” controller in the CHORD-SEQ editor might allow you to stretch a little bit the display, otherwise, you could do someting like generating another chord-seq from the original one, multiplying all onsets by a given factor, etc.

Jean

#5

I have a few recommendations. The “explosion” tool from OMTristan will turn all of your notes into individual chords. That won’t solve your problem by itself, but you can then either use SPLIT-VOICES tool to get each note on an individual track (in a multiseq). I’m not sure if there’s a way to control how many tracks you end up with (you can always set the channels manually if you want), but when you’re done working you can always bounce it back down to a chord-seq.

#6

An even better option might be: dividing the cluster up into its constituent microtones: then you’ve turned one “track” into 4, rather than an indefinite number. A good tool for doing this r-hocket-multiseq, which will do exactly that: multiples of half steps on one track, 1/8, then 1/4, then 3/8. That should be more than enough to clear up your clutter. (If you are using OM Ruben, keep in mind that most of the boxes only work with multi-seqs for some reason.)