Is there any easy way to highlight the current speaker automatically (rather than edit the video file manually) in a multi person podcast, similar to this for example:
1 Answer
Easy way: patch some client for voice conferences, save info about speakers. I did it with Mumble. Then write simple app for rendering such movies, read/write of images and footages can be done with ffmpeg, fftw3 can calculate spectrum for us, grab some filters for image scaling...
Hard way: write each speaker in separate channel (mumble can do this), then place all tracks in After Effects, use there tracks for scaling avatars and adding spectrograms. You can done it without writing code, but if you have many-many podcasts, you will suffer. Also, this not applicable if you want broadcast such video in realtime, for example, with questions and calls of guests.
