Re: [Jack-Devel] Sync problems with jack outputting midi clock

Yes  I would say that if the MIDI message correctly circulate inside JACK MIDI (so using jack_midi_dump would allow to test that part...) then the problem is probably in the  CoreMIDI and JACK  bridge.

Iain can you test with  jack_midi_dump?

Thanks.

Stéphane 


Le 12 juil. 2012 à 13:29, Paul Davis a écrit :

> or the bridging between CoreMIDI and JACK ....
> 
> you might try using jack_midi_dump too.
> 
> On 7/12/12, Gabriel M. Beddingfield <[hidden]> wrote:
>> 
>> Hello,
>> 
>> On 07/11/2012 08:47 PM, Iain Duncan wrote:
>>> Hi folks, I've made a midi clock test app to help track down the issues
>>> I've been having on OS X, but am getting some strange behaviour. Every
>>> once in a while, I'm getting two clock messages 1 or 2 ms apart. I'm no
>>> jack or C expert so it's quite possible I've made a stupid math error or
>>> something. If anyone has the time to take a look at it, that would be
>> 
>> I did a quick program to check the math... and it seems fine with all
>> events about 21ms apart.
>> 
>>> really helpful and would help us figure out whether there is an issue
>>> with the OSX midi driver. In the app I deliberately aligned the tempo to
>>> frame buffers to keep out as many sources of error as I could, it would
>>> be great to know whether this looks like it should work.
>> 
>> Nothing jumps out at me from the code.  How are you monitoring it?
>> 
>> Ideally, if you hook up jack client to jack client and check the
>> behaviour.  If it behaves fine then that points to the driver.
>> 
>> -gabriel
>> 
>> Jack-Devel mailing list
>> [hidden]
>> http://lists.jackaudio.org/listinfo.cgi/jack-devel-jackaudio.org
>> 
> 
> Jack-Devel mailing list
> [hidden]
> http://lists.jackaudio.org/listinfo.cgi/jack-devel-jackaudio.org

1342094294.12238_0.ltw:2,a <0C44249D-847B-441E-9FBA-4174AA5A4533 at grame dot fr>