I've been working on a step sequencer in max4live and noticed that the output of a metro synced to Live's transport is quite inaccurate. Midi note triggered by the metro are delayed by a number of samples (10-20) compared to midi notes coming from a Live clip. I could live with it if the delay was constant but, there is aslo a substantial jitter.
I though that metro synced to Live's transport was supposed to be "tick accurate".
This kind of jitter makes max4live kind of useless for designing step sequencers, since it introduces random phasing problems
I've attached a max4live patch as well as an audio file demonstrating the issue. In the audio file, I have recorded outputs from two instances of Impulse. One driven by a midi clip the other by a metro synced to the beat. Both instances trigger a sample with an impulse. In the audio file (if you zoom in) you can see that the impulse with higher amplitude (which was triggered by metro) is delayed by a "random" number of samples in relation to the other impulse (which always happens exactly on the beat).
Has anyone encountered a similar problem?
I'd really appreciate some comments from the devs on this one.