Analogue-sequencer

> Just done some more tests with Cubase/P3 and I've noticed 
> that its not drifting out of tempo its stepping out of tempo 
> i.e. it sounds like either Cubase isnt sending or P3 isnt 
> receiving a few MIDI clock pulses. 
> This is happening in a random fashion, sometimes sync is fine 
> for 60 bars, sometimes 20 sometimes 5..no real pattern to ot 
> as far as I can see.

Something I've discovered while developing the code for slaving Cirklon to
MIDI clock (Cirklon has a very different method of generating internal tempo
to P3), is that the MIDI clock output from a PC can have a remarkably
high-level of jitter.
You really have to smooth out any changes in tempo, as PC people seem to
think that a steady tempo of, for example 120bpm is the same as an *average*
tempo of 120bpm.
If you consider that 120bpm is 21ms per clock tick - if you have a clock
tick delayed by 15ms for whatever reason, the next tick occurs 6ms later.
Which amounts to a sudden drop to 69bpm (36ms per tick) followed by a leap
to 400mpm (6ms per tick).
That seems to be considered fair play for a PC sequencer.

It could be that there is a bug that causes Cubase to drop the odd MIDI
clock.
Or possibly an extreme case of having two clock pulses back-to-back might
cause a race condition in the P3 sync handler, which could happen if the
Cubase clock output was irregular enough.
I think the first thing to look for would be any other references to loss of
single ticks on devices slaved to Cubase under your Mac OS.

Cheers,
Colin