[LAD] midi beat clock

Ralf Mardorf ralf.mardorf at alice-dsl.net
Sat Feb 20 19:36:12 UTC 2010


Tim E. Real wrote:
> On February 20, 2010 11:33:04 am Ralf Mardorf wrote:
>   
>> Rui Nuno Capela wrote:
>>     
>>>> It also transmits them. It also recognizes tick but doesn't use it.
>>>> From what I understand, tick is nothing more than a periodic 'keep
>>>> alive' signal, not a sync signal. I could be wrong though...
>>>>         
>>> this "tick" signal (0xf9) must be (is it?) some leftover from ancient
>>> midi times, if ever used effectively. i fail to recognize its use but
>>> being just, uh,... _undesirable_ noise:)
>>>
>>> i'll turn off
>>>       
>> 0xf9 - "MIDI Tick" is a fixed and real sync signal, sync independently
>> from the tempo, the "keep alive" signal is 0xfe - "Active Sense", very
>> useful for a master keyboard during a live performance, assumed
>> receiving devices do know what to do, if "active sense" unexpected would
>> stay away.
>>     
> Oops. My mistake. Mixing up signals again.
> Tim.

I didn't know MIDI tick before, but I wondered that Rui was talking 
about noise, while you were talking about active sense. It makes me 
wonder, because active sense should be send every 300ms, but it also 
would be okay after active sense was send for the first time, that there 
is any signal within a second, e.g. active sense every 800ms instead of 
every 300ms, so this won't be a problem. I need to read about MIDI tick 
and yes, it has to be send exactly every 10ms. 100 times a second is 
more traffic :D. MIDI tick seems to be something relative new for MIDI, 
Pedro mentioned this before.



More information about the Linux-audio-dev mailing list