26 Jul
2013
26 Jul
'13
4:40 a.m.
Hi all! I'm wondering why so many of my MIDI ports in Jack are named so similarly, and
if they can be changed? I don't remember this being a problem on my previous Linux
system, but there are a few differences. I'm avoiding GUIs and need to use CLI tools,
so no Qjackctl suggestions, please. I'm using an M-Audio MIDI sport Uno USB midi
interface, and the virMIDI device for internal patching, as well as MIDISH. In jack, all
of those, along with the main system ports like through are all listed as system midi
playback or capture and a number. I've tried both raw and seq jack MIDI modes. I can
see which is which with Jack_lsp with the -A or -t options set. But I was hoping to use
something like Jack Plumbing for reconnecting, and since the real client names don't
show up, I can't get jack plumbing to see them. Since the port numbers aren't
always the same, I can't depend on the suffix digits to identify the right port. Then,
using MIDISH, where the ports come and go when it starts and stops, its port numbers seem
to increment each time as well. Should I perhaps be using A2Jmidid instead of the jack
MIDI modes? Is there something else I'm missing? Or is this a difference between Jack
1 and Jack 2? Or differences from Ubuntu 9-10 and 12-10? Sorry to be so wordy, but wanted
to be clear. Thanks for any help.
Kevin
26 Jul
26 Jul
6:30 a.m.
Hello Kevin!
I'm not sure about your portnames, but with -Xseq (in JACK 1), my MIDI port
names look like this:
alsa_pcm:Emu10k1-WaveTable/midi_playback_4
This is very long, but for automated purposes this is OK. I could cut©
and paste it into any script.
But there's jack_alias, this could rename ports for you. Since your ports do
change names or numbers, this might not be the best solution neither. As far
as I know, jack_alias can rename a whole client, not a single port.
With j2amidi_bridge you could patch a JACK port thought to ALSA. You'd still
have to connect a softsynth port to it. But soft synth ports should have the
usual names. Other hardware ports can be connected directly through ALSA
sequencer. I can send you my example .midishrc off-list. There's more than
just a device setup, there are some convenience functions to connect new ports
on the fly.
I hope, this did answer your question, at least partly.
Kind regards
Julien
----------------------------------------
http://juliencoder.de/nama/music.html
12:07 p.m.
On Fri, Jul 26, 2013 at 12:40 AM, Kevin Utter <kvutter(a)frii.com> wrote:
stops, its port numbers seem to increment each time as well. Should I
perhaps be using A2Jmidid instead of the jack MIDI modes?
yes, absolutely. the only reason that -X raw and -X seq have not been
removed is that the code to replace them isn't ready. always use a2jmidid
(with -e to pick up external h/w ports).
1:04 p.m.
On Fri, Jul 26, 2013 at 08:07:51AM -0400, Paul Davis wrote:
yes, absolutely. the only reason that -X raw and -X
seq have not been
removed is that the code to replace them isn't ready. always use a2jmidid
(with -e to pick up external h/w ports).
This made me have a look at the sources.
Up to ten minutes ago I was believing that a2jmidid used timestamps etc.
to reduce jitter, and that this made it preferable over Jack's -X.
But apparently it doesn't, all events in the same Jack period are
just bunched together. There's even a comment in the source code:
/* -FIX- this should have the frame time of the event, instead of '0': */
So what is the advantage ?
Ciao,
--
FA
A world of exhaustive, reliable metadata would be an utopia.
It's also a pipe-dream, founded on self-delusion, nerd hubris
and hysterically inflated market opportunities. (Cory Doctorow)
1:53 p.m.
the last time i touched a2jmidi, it did precisely what you were looking for
- 1 period of latency, zero jitter.
however, i do not maintain a2jmidid (it is not my project) and i have no
idea what has happened to it since I last did any work on it.
On Fri, Jul 26, 2013 at 9:04 AM, Fons Adriaensen <fons(a)linuxaudio.org>wrote;wrote:
On Fri, Jul 26, 2013 at 08:07:51AM -0400, Paul Davis
wrote:
yes, absolutely. the only reason that -X raw and
-X seq have not been
removed is that the code to replace them isn't ready. always use a2jmidid
(with -e to pick up external h/w ports).
This made me have a look at the sources.
Up to ten minutes ago I was believing that a2jmidid used timestamps etc.
to reduce jitter, and that this made it preferable over Jack's -X.
But apparently it doesn't, all events in the same Jack period are
just bunched together. There's even a comment in the source code:
/* -FIX- this should have the frame time of the event, instead of '0': */
So what is the advantage ?
Ciao,
--
FA
A world of exhaustive, reliable metadata would be an utopia.
It's also a pipe-dream, founded on self-delusion, nerd hubris
and hysterically inflated market opportunities. (Cory Doctorow)
_______________________________________________
Linux-audio-user mailing list
Linux-audio-user(a)lists.linuxaudio.org
http://lists.linuxaudio.org/listinfo/linux-audio-user
2:44 p.m.
On Fri, Jul 26, 2013 at 09:53:06AM -0400, Paul Davis wrote:
the last time i touched a2jmidi, it did precisely what
you were looking for
- 1 period of latency, zero jitter.
however, i do not maintain a2jmidid (it is not my project) and i have no
idea what has happened to it since I last did any work on it.
The code in jack.c seems to do the right thing, the one in a2jmidi_bridge.c
and j2amidi_bridge.c (which is where I looked first) certainly doesn't.
Ciao,
--
FA
A world of exhaustive, reliable metadata would be an utopia.
It's also a pipe-dream, founded on self-delusion, nerd hubris
and hysterically inflated market opportunities. (Cory Doctorow)
3932
days inactive
3932
days old
linux-audio-user@lists.linuxaudio.org
5 comments
4 participants
participants (4)
-
Fons Adriaensen -
Julien Claassen -
Kevin Utter -
Paul Davis