[LAU] Sustain pedal issue with undesired repeated notes (m-audio keystation 88es + qjackctl + qsynth)

[David Santamauro]

On Tue, 08 Jun 2010 11:37:23 +0200
Nicolas BERCHER <nbercher at yahoo.fr> wrote:

> On 08.06.2010 11:21, David Santamauro wrote:
> >
> > Hi,
> >
> > On Tue, 08 Jun 2010 10:57:13 +0200
> > Nicolas BERCHER<nbercher at yahoo.fr>  wrote:
> >
> >    
> >> So, I checked what's going on:
> >> $ aseqdump --list
> >>    Port    Client name                      Port name
> >>     0:0    System                           Timer
> >>     0:1    System                           Announce
> >>    14:0    Midi Through                     Midi Through Port-0
> >>    20:0    SB Live! Value [CT4870]          EMU10K1 MPU-401 (UART)
> >>    24:0    USB Keystation 88es              USB Keystation 88es
> >> MIDI 1 $ aseqdump -p 24:0
> >> Waiting for data. Press Ctrl+C to end.
> >> Source_ Event_________________ Ch _Data__
> >>    24:0   Control change          0  64 127
> >>    24:0   Note on                 0  82  82
> >>    24:0   Note on                 0  82   0
> >>    24:0   Control change          0  64   0
> >>
> >> The above sequence dumped by aseqdump shows only one note event
> >> (lines 2+3: press+release) but I heard a delayed note as mentioned
> >> in my first mail.
> >>      
> > Actually, there are two note events and both are NOTE ON, one with
> > velocity of 82 and the other 0. The velocity 0 should be
> > interpreted as NOTE OFF.
> >
> > I had a similar problem and solved it by putting kmidimon between my
> > keyboard and sequencer.
> >
> > Could it be that qsynth doesn't handle NOTE ON with
> > velocity of 0 as NOTE OFF? Although I imagine that translation
> > should be happening at a lower level... just a guess.
> >    
> Yes it is, you're right, but it happens only for 4 notes among 88 I
> can play!
> If this putting kmidimon between my keyboard and qsynth can solve the 
> problem, I'd also like to understand why. Does it mean there exist 
> multiple midi standard to describe note on/off events?
> Nicolas.

Sorry, I re-read the initial post -- 4 notes, hrmmm. That points
to a hardware issue but aseqdump doesn't show additional events.

Although I haven't delved into the code, the translation of NOTE ON (0)
to NOTE OFF should be happening at a much lower level than user-space
software. There is one standard for midi note on/off and it is clear
that note on (0) must be interpreted as note off.

Have you tried playing it back through a different sound generator,
possibly external gear attached to your 20:0 port?

Just as a test, what is registered when you route your keyboard into
kmidimon and route kmidimon into qsynth (all defaults, simple piano
patch)
  - does kmidimon show 1 note-on/note-off only
  - does qsynth output correctly?

David