#16 startBristol fails when jackd -p is used to set port-max

closed
5
2008-09-17
2008-07-30
No

The startBristol script uses the value of the first -p flag on the jackd command line as the value of frames-per-period for the jackd ALSA back-end.

However, jackd also uses -p to set the maximum number of ports: the -p flag can thus appear twice on the jackd command line.

If maximum number of ports is set to a non-default value in QJackCtl, startBristol finds the wrong value for frames-per-period, and can thus fail to start with errors such as this:

Jack period count mismatch: try `bristol -jack -count 2048`)
You need to ensure that bristol uses the same period size
Bristol is exiting ungracefully, ^C or ^Z will help

I've attached a simple patch to startBristol.in, which makes this work for me by using the last rather than the first occurrence of -p on the jackd command line to find the frames-per-period which jackd was started with.

Discussion

  • Colin Fletcher

    Colin Fletcher - 2008-07-30
     
    Attachments
  • Nick Copeland

    Nick Copeland - 2008-08-09
    • labels: --> Audio Library
    • assigned_to: nobody --> ncopeland
     
  • Nick Copeland

    Nick Copeland - 2008-08-09

    Logged In: YES
    user_id=1426398
    Originator: NO

    Hi Colin,

    I will incorporate your changes into bristol but may have a quick look at them first - out of interest, why does the 'tail -n1' work? I dont' think the change will get into the 0.20.6 release that has been tarballed ready for distribution. The following release should have this change though.

    Regards, nick

     
  • Nick Copeland

    Nick Copeland - 2008-08-09

    Logged In: YES
    user_id=1426398
    Originator: NO

    Just a note, even with the existing script I would have thought that giving the count explicitly to startBristol would override any value found from Jack. In short, the script might find the wrong value but you can override that by using the -count option. Should work.

    Will still integrate your fix however there are some issues with out jack distributes these defaults - the only way to find them out in advance is to poll the daemon and that is a bit awkward with the Bristol initialisation procedures. It can be done however it would start and probably then stop the daemon again under many circumstances. Perhaps I should just put it in, I have the code to do so from another source.

     
  • Colin Fletcher

    Colin Fletcher - 2008-08-11

    Logged In: YES
    user_id=1367777
    Originator: YES

    Hello Nick,

    Thanks for your responses.

    The 'tail -n1' works because of grep -o, which prints only the matched (non-empty) parts of a matching line, with each such part on a separate output line.

    You're of course right that explicitly passing a count does work. It's just a little less convenient, especially when you're often switching between sound cards using different values for -p and you have to remember which is which.

     
  • Nick Copeland

    Nick Copeland - 2008-09-11

    This slipped past 0.20.7. The changes has gone into 0.20.8 and will be distributed with that release presently.

     
  • Nick Copeland

    Nick Copeland - 2008-09-11
    • status: open --> pending
     
  • Nick Copeland

    Nick Copeland - 2008-09-17
    • status: pending --> closed
     

Get latest updates about Open Source Projects, Conferences and News.

Sign up for the SourceForge newsletter:





No, thanks