Beat time: Is it really beats, or quarter notes? about linkkit HOT 4 CLOSED

Yes. The duration of a beat corresponds to the one of a quarter note. If you're dealing with different signatures check out the section about Phase Synchronisation in Link's documentation.

from linkkit.

Thanks, although that does bring me to my next question ...

I'm using the current time signature's 'beats per bar' to determine the quantum. However, when the 'beats per bar' changes (as a result of a user interaction, or due to an event in the sequence), the beat time returned by LinkKit becomes discontinuous, and the timeline jumps to a different position. I've worked around this by adding a compensation offset when a quantum change occurs, as follows:

	// Has the quantum changed since the last call?
	if (mQuantum != zQuantum)
	{
		// Calculate the beat time using both the old and the new quantum
		f64 OldQuantumBeat = ABLLinkBeatAtTime(mpCapturedTimeline, HostTime, mQuantum);
		f64 NewQuantumBeat = ABLLinkBeatAtTime(mpCapturedTimeline, HostTime, zQuantum);

		// The difference between the two returned beat times must be added as an offset in all future calls
		// to prevent a timeline discontinuity
		mBeatOffset += (OldQuantumBeat - NewQuantumBeat);
		
		// This is now the new quantum
		mQuantum = zQuantum;
	}

	// Return the beat time, adding our offset to prevent discontinuities when the quantum changes
	return ABLLinkBeatAtTime(mpCapturedTimeline, HostTime, mQuantum) + mBeatOffset;

This seems to work, and the app's bar start (phase) still seems to align with the connected peers when the quanta are a multiple of each other but it still feels like a workaround. Have I handled this correctly?

from linkkit.

Yes. The beat time will be discontinuous when changing the quantum. For Ableton Live we're taking an alternative approach on handling this. We're saving the quantum when starting transport and use this for alignment with Link until transport is stopped. When transport is started again we update the quantum.

from linkkit.

OK. This is the same alternative approach I thought of last night, so I think I might change it!

Thanks once again for your help.

from linkkit.