With the new Ircam Forum website coming, the user groups will be migrated towards a new platform.
Please plan user groups downtime between Monday, March 25th and Tuesday, March 26th.

Group Admins

  • Profile picture of Karim
  • Profile picture of Jean


Public Group active 21 hours, 2 minutes ago

User group for OpenMusic and computer-aided composition. Visit the Forum for discussions.

Separating Large Clusters in chord-seq object

Author 3 Subscribed Users |
Profile photo of Steve97

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,


  1. Patch-Proof


January 9, 2019 at 14:40 #28759
Profile photo of Jean

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

January 9, 2019 at 20:16 #28764
Profile photo of Steve97

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.


January 10, 2019 at 19:21 #28778
Profile photo of Jean

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.


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.


January 12, 2019 at 10:29 #28796
Profile photo of michalszostalo

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.

January 12, 2019 at 18:50 #28813
Profile photo of michalszostalo

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.)

January 12, 2019 at 19:00 #28814

You must be logged in to reply to this topic.

Log in now