State of the art for sample accurate tick triggers off master transport?


    Feb 22 2019 | 5:07 am
    Ok I apologize for posting something that I'm sure is in here elsewhere, but I'm having a really hard time finding out what the current (max 8) recommended practise is for making sample accurate timing triggers that are synced to the master clock, with the option of the master clock being sycned with, or driven by, the Live master clock. I've read the phasor~ delta~ edge~ stuff, I've read the sync~ help, but I'm not managing to come up with an elegant way of making ppq tick triggers for my step sequencer that stop and start off the master clock being started and stopped. Any help would be much appreciated, including pointers at examples. I'm wanting to be able:
    - have the clicks start and stop on master transport start/stop - have a way to receive the current location from the master transport to clean up my internal pointers (i.e. be able to respond to the rewind event and such) - have tight timing
    I'm not afraid of fancy hacking to get this right, I'm just getting frustrated trying to figure out what's current and hoping not to build up something and then find out I didn't know there's now some dead simple way... :-)
    thanks!

    • Feb 22 2019 | 5:26 am
      The new version of the phasor~ help seems to be saying (to me reading at least) that the if audio interrupt and overdrive are on, metro should be just as good as phasor now, is that correct?
    • Feb 22 2019 | 2:36 pm
      in theory there is always a risk that at data rate things can come late. for example as soon as you send 2x "hello" at the same logical time, they will of course be sent one after another. in practice sequencing music using metro works fine as long as the scheduler is not less accurate than one audio vector size is long. in max4live the vectorsize is 32, so when the scheduler runs at 1 ms this requirement is fullfilled (as long at the samplerate is not higher than 48khz.) if you drive, say a biquad object or a VST plug-in hosted in live, that target might also only update every vector even if you would send it a control signal at signal rate. :) imho phasor might be interesting for audio-only code like in a granulator, for the creation of musical events in time it is not. -110