[LAU] Period Elongation NOW !

Frank Barknecht fbar at footils.org
Thu Jan 17 05:43:00 EST 2008

porl sheean hat gesagt: // porl sheean wrote:

> i don't know of any specifics, but in theory i can think of two ways to do
> it: one is to add a constant time offset to any 'note off' or 'note on with
> 0 velocity' event, and the other is to ignore those events and add them in
> manually by taking the normal note on event and adding a note off (or 0)
> event with a constant offset. don't know what things you can do it in
> though, maybe puredata?

Yes, Pd can do it and it's a common idiom to filter out note-offs if
you're not interested in them with the [stripnote] object. A
minimal note off filter patch would look like this: 

 |   |
 |   |

Make that with Pd, aconnect it as you like and that's it.

> On 17/01/2008, lanas <lanas at securenet.net> wrote:
> >   From Kmidimon, the Alesis Control Pad MIDI notes, when hit with drum
> > sticks, last typically 1 unit of time (whatever unit that is) eg.:
> >
> > Time   Event     Data1    Data2
> > 89985  Note on   63       0
> > 89984  Note on   63       55
> >
> >   Whereas the shortest note I can do on an Axiom 25 keyboard looks like:
> >
> > Time   Event     Data1    Data2
> > 41067  Note on   89       00
> > 41051  Note on   89       94
> >
> >   That's a whole 16 units of whatever time.  Same with the Axiom's
> > pads: they last a long time and thus, interact OK with Qsynth.
> >
> >   My theory about Qsynth (Fluidsynth...) is that the MIDI notes sent by
> > the MIDI percussion instrument are too short for it to pick them up.

Or maybe they get played, but turned off immediatly? At least that
would be the correct behaviour if you play a very short note. If you
don't want notes to turn off immediatly, then use a note off filter as

 Frank Barknecht                                     _ ______footils.org__

More information about the Linux-audio-user mailing list