Re: [iiwusynth-devel] iiwusynth: CVS update (noteon and noteoff)

[M. Nentwig]

Hello,

The motivation for the change was a stuck-note-problem I'm getting with
one particular keyboard (I'm quite sure that it's related to a 'sticky'
key contact, it seems to forget to send note-offs now and then on one
particular key). The problem is that with the old noteoff algorithm
there is absolutely no way to turn off a stuck key short of sending 'all
notes off' (or unplugging the MIDI connector, pushing the key, plugging
it in again, releasing the key).
The new noteoff implementation will turn off all voice processes
regardless of the ID, so simply retriggering the stuck key will release
both voices.

Now to me, the situation looks like this:
- A properly working keyboard will never transmit more than one noteon
without corresponding noteoff for each channel / keynumber.
- Same with a correct MIDI file (my guess, without any specs to refer
to), for the same reason: it's unphysical to hit the same key twice
without releasing it.
So in an ideal world, it wouldn't make a difference.
- However, stuck notes (for whatever reason) happen, and in my opinion a
foolproof 'stuck note' handling is very important. On stage, hanging
notes are a major embarassment (I had one, with an organ sound, and I
still remember it clearly after five or six years. The stuck-note
handling of my home-built MIDI router was not foolproof :-)

Now we could make the choice of the noteoff algorithm an option, and
leave the decision to the user. Maybe we'll have to, if the old
implementation is needed for some reason (or if the specs actually say
so). I still don't like the idea; in my opinion a MIDI data stream with
two noteons in a row is faulty, and we should handle that as safely as
possible.

Regards

Markus

> -----Original Message-----
> From: ext Peter Hanappe [mailto:address@hidden
> Sent: 21 July 2002 22:16
> To: M. Nentwig
> Cc: address@hidden
> Subject: Re: [iiwusynth-devel] iiwusynth: CVS update (noteon and
> noteoff)
>
>
> M. Nentwig wrote:
> > Hello all,
> >
> > My own changes to the
> > noteoff code are also included (now a stuck note can be released by
> > hitting the same key again).
> > The main reason that I'm getting stuck notes seems to come from the
> > mechanics of one particular keyboard - but that is beyond
> the control of
> > iiwusynth.
>
> There is some confusion about what notes should be turned off when a
> note off is received. Should it be 1) all the notes with the given
> channel-key pair or 2) only the oldest note with the given
> channel-key pair?
> I've been told only the oldest note should be turned off. That's why I

> use the ID of the notes to find the oldest note.
> Maybe we should make it an option rather than hard code one or the
> other solution.
>
> Cheers,
> Peter
>
> > Cheers
> >
> > Markus
> >
> >
> > _______________________________________________
> > iiwusynth-devel mailing list
> > address@hidden
> > http://mail.freesoftware.fsf.org/mailman/listinfo/iiwusynth-devel
> >
> >
> >
>
>
>
>
> _______________________________________________
> iiwusynth-devel mailing list
> address@hidden
> http://mail.freesoftware.fsf.org/mailman/listinfo/iiwusynth-devel
>