there's always two options rather directions if you wish:
1. add to current bloat on qtractor as an alternate model; yes, i've said bloat indeed ;)
2. do as a separate application or process, that possibly interfaces to qtractor via some kind of IPC; FYI. something quite similar has been proposed and implemented a few years ago, following jpmercury's freewheeling to which all live looping performance could get recorded via OSC messaging (as IPC); for all that meta/macro-recording kind of stuff there's--tada--the oscx branch still to play around.
now you touched the brimstone :)
there's always two options rather directions if you wish:
1. add to current bloat on qtractor as an alternate model; yes, i've said bloat indeed ;)
2. do as a separate application or process, that possibly interfaces to qtractor via some kind of IPC; FYI. something quite similar has been proposed and implemented a few years ago, following jpmercury's freewheeling to which all live looping performance could get recorded via OSC messaging (as IPC); for all that meta/macro-recording kind of stuff there's--tada--the oscx branch still to play around.
cheers