20 Oct
2008
20 Oct
'08
7:34 a.m.
Hello!
I just read through the JACK transport section in the JACK reference again.
If you want to be timebase-master you can (or must) supply a callback function
to update the time-info.
Does my program need audio thread (a process function?) Which timing source
is used to update the time? Does my program provide it or does JACK provide it
and my program only really kicks in when I want to relocate (jump to another
position)?
I am correct in assuming, that my code doesn't have to be multithreaded (at
least not look multithreaded) if I want to use the JACK transport interface
(control and timebase)?
Kindest regards
Julien
--------
Music was my first love and it will be my last (John Miles)
======== FIND MY WEB-PROJECT AT: ========
http://ltsb.sourceforge.net
the Linux TextBased Studio guide
======= AND MY PERSONAL PAGES AT: =======
http://www.juliencoder.de
20 Oct
20 Oct
7:43 a.m.
On Mon, 2008-10-20 at 13:33 +0200, Julien Claassen wrote:
Hello!
I just read through the JACK transport section in the JACK reference again.
If you want to be timebase-master you can (or must) supply a callback function
to update the time-info.
Does my program need audio thread (a process function?)
No. JACK will call your timebase callback at the right time.
Which timing source
is used to update the time?
that is up to you. it could be MTC, some other form of timecode, a
wallclock, NTP or whatever you want. keep in mind that JACK does not
support varispeed (a decision made long ago to avoid burdening every
JACK-transport app with very difficult implementation issues).
Does my program provide it or does JACK provide it
and my program only really kicks in when I want to relocate (jump to another
position)?
no, if your client elects to be time master then it is responsible for
updating the time every process cycle.
I am correct in assuming, that my code doesn't
have to be multithreaded (at
least not look multithreaded) if I want to use the JACK transport interface
(control and timebase)?
JACK will create a thread for you, you just supply the callback. The
callback will be called from the JACK-created thread, so you need to be
aware of this.
8:23 a.m.
Thanks you Paul!
One question: I can use MTC as a source for timebase. But the application
(client(s)) on the other end don't need to know about this. They can be you
average audio application?
Kindest regards
Julien
--------
Music was my first love and it will be my last (John Miles)
======== FIND MY WEB-PROJECT AT: ========
http://ltsb.sourceforge.net
the Linux TextBased Studio guide
======= AND MY PERSONAL PAGES AT: =======
http://www.juliencoder.de
8:26 a.m.
On Mon, 2008-10-20 at 14:23 +0200, Julien Claassen wrote:
Thanks you Paul!
One question: I can use MTC as a source for timebase. But the application
(client(s)) on the other end don't need to know about this. They can be you
average audio application?
if the MTC comes from a sample-clock-synced source (i.e. its sample
clock runs precisely in sync with the sample clock being used by JACK
(i.e. your audio interface)), then certainly.
but if the MTC is able to varispeed, then no. ardour specifically
prohibits varispeeding when synced to JACK for this very reason.
10:01 a.m.
Hi Paul!
And if the MTC is not varyspeeding, but simply taken from gettimeofday(2)? I
agree, this is not directly synced to my soundcards clock, but still as far as
my computer goes, it should be accurate.
Kindest regards
Julien
--------
Music was my first love and it will be my last (John Miles)
======== FIND MY WEB-PROJECT AT: ========
http://ltsb.sourceforge.net
the Linux TextBased Studio guide
======= AND MY PERSONAL PAGES AT: =======
http://www.juliencoder.de
10:14 a.m.
On Mon, 2008-10-20 at 16:01 +0200, Julien Claassen wrote:
Hi Paul!
And if the MTC is not varyspeeding, but simply taken from gettimeofday(2)? I
agree, this is not directly synced to my soundcards clock, but still as far as
my computer goes, it should be accurate.
what are you trying to do?
10:31 a.m.
Hello Paul!
I'd like to see midish and ecasound work together. What I need primarily:
they must start in sync. What would be convenient: They should both stop, if I
tell one of them to do so. In between: I don't care, I just want to record
sound from MIDI triggered devices/apps.
But I've heard, that it's necessary or at least clean, to have the MIDI
sequencer and the audio app in sync all the time.
So that's, what I'm researching. What would you advise?
Kindest regards
Julien
--------
Music was my first love and it will be my last (John Miles)
======== FIND MY WEB-PROJECT AT: ========
http://ltsb.sourceforge.net
the Linux TextBased Studio guide
======= AND MY PERSONAL PAGES AT: =======
http://www.juliencoder.de
11:12 a.m.
Paul, just to get it clear: You say only implement JACK transport control
(start, stop) and leave the rest for a later date.
Will this start midish and any other audio app always at the same time?
I once had a priliminary go at it and ecasound and midish didn't alwaysstart
in sync. I recorded seperate tracks of midish (midi instruments) at seperate
times and then tried to mix them in ecasound, only to discover, that they
weren't in sync. Was that probably my fault entirely or would you expect this
behaviour?
Kindest regards
Julien
--------
Music was my first love and it will be my last (John Miles)
======== FIND MY WEB-PROJECT AT: ========
http://ltsb.sourceforge.net
the Linux TextBased Studio guide
======= AND MY PERSONAL PAGES AT: =======
http://www.juliencoder.de
5672
days inactive
5672
days old
linux-audio-dev@lists.linuxaudio.org
8 comments
2 participants
participants (2)
-
Julien Claassen -
Paul Davis