search

FFMS / ffms2

An FFmpeg based source library and Avisynth/VapourSynth plugin for easy frame accurate access
Other
582 stars 105 forks source link

Creating an FFMS_AudioSource from multiple tracks possible? #195

Closed chartelt closed 9 years ago

chartelt commented 9 years ago

My case is a .mov container with separate wav tracks for a 5.1 sound layout.

Currently I create a FFMS_AudioSource for each audiotrack to get the samples one by one. I wonder if I can create an FFMS_AudioSource from more than one audiotrack in a container to avoid this step?

dwbuiten commented 9 years ago

This seems quite outside the scope of FFMS2. This is by all means, a corner case, and existing libraries exist to easily merge and/or mix them.

tgoyne commented 9 years ago

Is there any sort of metadata that indicates that the separate tracks should be combined? If so combining them automatically and reducing the number of stupid things users have to deal with would be reasonable.

dwbuiten commented 9 years ago

AFAIK in many situations it is just assumed, since it's part of some deliverables spec somewhere.

I may be wrong for other cases.

chartelt commented 9 years ago

The metadata (or channel layouts) are nicely signaled for each audio track (at least for mov), so I read those and handle the tracks accordingly. I was just wondering if there was an easier way to deal with this inside ffms, but I understand that this is out of scope.