MIDI CLock - anyone have understanding?

[midimaestro]

Hi all!

New to the forums and yes I'm into using sequencers with music but am wondering if anyone here has any idea about how some certain aspects of MIDI function, in particular MIDI Beat Clock, ie how are time signature changes handled?

Of course I use cubase and would not consider anything else at any cost but what I'm trying to do is syncronize a hardware drum machine that supports SPP, System Realitime Messages such as start, stop & continue, as well as MIDI Clock or system clock as it is noted on the implementation chart.

From what I've read here:

http://www.blitter.com/~russtopia/MIDI/~jglatt/tech/midispec.htm

it would appear that there are 24 clocks per quarter note.

So how would that translate if a song changed from say 4/4 for example to 6/8 at some point in the song and back?

Hope someone can help.

Greetings cubendo users

[david pick]

It's a while since I've spent any time thinking about the weird little corners of the MIDI spec, so correct me if I'm wrong, but I think MIDI clock is a very simple MIDI sync protocol which effectively provides a series of timing pointers at a constant rate (24 per quarter note) and it doesn't have the sophistication to deal with time signature changes. It will simply carry on sending 24 pointers per quarter note, regardless. So if the receiving device changes time signature mid-tune, the tempo will change (ie the BPM).

I seem to remember that originally MIDI clock had no song position information, so if the master and slave were at different points in the song, they would just continue with the sync offset. Song Position Pointers (SPP) were introduced to address this. The master will send a SPP when it's started, followed by a stream of MIDI clocks. The SPP makes the slave jump to the correct position, then the MIDI clocks keep it there.

Just realised this answer comes a mere 5 months after the OP...