jit.movie~ only stereo??


    Oct 10 2016 | 3:24 pm
    Hi,
    do I understand right, that jit.movie~ is a new object sending the audio of a video to MSP directly instead of having to use spigot?
    Spigot was limited to stereo. Is this the case with jit.movie~ as well? Haven't found otherwise in the references. It really would be great to finally have a multichannel feature.
    best, mr.l

    • Oct 10 2016 | 11:44 pm
      Yes, unfortunately it remains only stereo. I have also issued a request to update the new object to support multichannel embedded audio, which will open up a lot of new doors, so it's great to see someone else asking for it.
      I think it would be relatively easy to do if they already have 2 channels being brought into MSP, so we'll just have to keep our fingers crossed!
    • Oct 11 2016 | 12:10 am
      PS Have a look MR. L over here, I did manage to find a way to maintain sync with multichannel audio in sfplay~, but it's not ideal: https://cycling74.com/forums/max-7-2-5-released
    • Oct 11 2016 | 3:35 pm
      hey guys, feature request noted. i'll update if / when i can accommodate this.
    • Oct 11 2016 | 6:37 pm
      Thanks so much Rob!!
    • Oct 11 2016 | 7:39 pm
      Rob - thanks, that would be great and a big time and programming-saver!
      @KCOUL - thanks for the link - I'll have a look. I remember having done some audio video synching using the audio-file as the master and compare the position of the video and audiofile. Once it went out of sync it slightly reduced or accelerated the rate of the movie to run back into synch. It worked for me back then, but really is a workaround I wouldn't want to built on, if there is a lot of other stuff going on. https://cycling74.com/forums/howto-multichannel-audio-jit-qt-movie-on-windows-7
      I guess for now, we'll just have to use what suites us best until Rob can implement the multichannel function.
      Best, rené
    • Oct 11 2016 | 7:41 pm
      oh, and I just noticed, that jit.movie~ doesn't playback movs with uncompressed audio. At least not on my mac. Only AAC and AppleLossless.
    • Oct 11 2016 | 10:45 pm
      Mr. L - did you try using the VIDDLL engine instead of QT?
      I should say that behavior is not ideal in my workaround, something I've seen in the past when separate player instances were responsible for video and audio, respectively.
      First of all, the behavior is not great when seeking, especially on systems using a spinning platter drive. It just sort of hangs, and then the audio will start playing after the delay, and the video will get stuck on a frame, and then eventually start playing (roughly) in sync with the audio.
      The other thing is that after seeking, the sync is ever so slightly off from then onwards, and there is no way to guarantee it is accurate down to the video frame/audio sample (that I know of).
      My question for your workaround - how do you know that the audio/video is out of sync in your patch? How do you know when it comes back into sync? I saw the patch when researching this problem and figure it would be easier just to ask you, as the patch seemed fairly complex to read!
    • Nov 01 2016 | 12:44 pm
      Dear KCOUL - it's been some time. Sorry for not answering, I was rather bussy and afterwards simply forgot...
      To answer your question about that workaround: the movie-object gives out the current position of the movie (sending a 'gettime' message with every bang it receives). The sfplay is also sending out its fileposition. Inside the subpatcher "compare" the audio-fileposition is recalculated into frames per seconds. Note that I use 25fps! (milliseconds /1000 *25) Then both framenumbers are subtracted. If the result is 0 (= same position) the movieframerate is set to 1. If it is different from 0 (unequal position) the framerate of the movie is set to eather 0.95 or 1.05 until they run back into sync.
      In the main patch you'll see a toggle saying "start". That sets both files to time 0 and starts both.
      That workaround did work for me if I start the files from 0. Not so good for jumping around in the movie...
      `
    • Nov 01 2016 | 8:55 pm
      that's too bad, but it's what I expected. Dealing with separate audio and video player instances which are meant to be synced together is never going to be perfect. For my situation, I need to support jumping around in the movie quite frequently, and have the audio wait until the movie is ready to start playing back, which with UHD can take quite a long time.
      At the very least, I wish that the movie object would send some sort of notification that it had finished rebuffering and was going to start playing back in the next millisecond. It is so easy to quickly cue audio compared to video, that would be a better workaround than the solution you found...
    • Nov 14 2017 | 4:06 pm
      hi there, any news on multichannel support ? would be greatly appreciated, thanks !
    • Nov 15 2017 | 9:41 am
      Yes, we are looking for a multichan solution as well!
    • Nov 15 2017 | 10:25 pm
      I'm not sure if it might help, but in case it does, perhaps we can try and provide a finite set of format layouts we are specifically hoping to get support for.
      For me, the big three are 5.1 surround, 7.1 surround, and 4-channel B-format (1st order ambisonic WXYZ).
      Nice to haves would be 9 channel 2nd order ambisonic, and 16 channel 3rd order, but by far the above three would be much more useful than these two.
      In an ideal world it would support any number of channels up to some reasonable maximum and make the user deal with whatever groupings might exist (multilingual audio tracks come to mind here, which could be quite useful for installations etc.)