record 2nd track into QT movie for synchronised data storage

    Beshmeee
    Jan 21 2012 | 2:31 pm
    I'm looking to record a second track into an existing video, and use it to carry control/automation data that related to objects in the video (manual input via mouse etc). Objective is to keep all the data in sync with the frame where/when it was crated. I also need to be able to get it back out again of course!!
    If this is possible, can it be done with multiple new tracks for more data feeds. Each stream of data is quite small in size, typically no more than 6 floats.
    I've been doing this thus far with coll and a timecode track in the movie, but I'd like to keep all the data in one place (in the movie file) if possible.
    Any tips would be very much appreciated.
    I've always struggled with Jitter, and simply getting data of this kind in and out of a matrix I find challenging!
    Cheers, B.
    • Parmerud
      Jan 24 2012 | 12:05 pm
      The main problem is (I think) that as far as I know, QT multitrack movies can only output one video track at the time. Basically used originally (again I think) to create virtual presentations where you can switch for instance the angle of a filmed object. So probably the only way to pack automation data would be to use the audio track. Since Jitter can convert from image to sound (I presume you found a way to encode your control data into a matrix format, you could record the automation data as sound and then convert it back when you replay the movie. Don´t know if it´s a decent solution but the only one I can think of at the moment.
    • Beshmeee
      Jan 24 2012 | 5:16 pm
      Thanks for the reply Parmerud.
      This may work, but I imagine I'd be limited with the track count for the multiples of the automation data (short of doing some for of channel coding). I was hoping to use a single QT track to store all the data, there will only be a relatively small amount per automation channel so imagined I'd be able to pack a heck of a lot into each matrix.
      Still, the initial project will only require one set of data for the tests so I'll give it a go. Thanks again for the help here.
      Cheers, B.