Currently the Audio gets encoded twice for no real reason, so depending on how complex your videos audio is you might lose some detail (or entire channels...).
I would suggest to use -c:a copy to create the temp audio file rather than mp3 and put it into an MKA container.
Then put it together the same way as before, so you end up with unchanged original audio.
I know what i need to tell ffmpeg to do this, but the code itself relies on the audio file being called voice_tmp.mp3
So changes to a couple of files would be necessary. I have never done a pull request before, but i will try to create one with all the changes needed.
In some rare cases this could cause problems when a audio codec is not supported by the mka container. The only thing that comes to my mind right now tho is DSD audio formats, but i have never seen one in a video file...
Currently the Audio gets encoded twice for no real reason, so depending on how complex your videos audio is you might lose some detail (or entire channels...).
I would suggest to use
-c:a copyto create the temp audio file rather than mp3 and put it into an MKA container. Then put it together the same way as before, so you end up with unchanged original audio.I know what i need to tell ffmpeg to do this, but the code itself relies on the audio file being called
voice_tmp.mp3So changes to a couple of files would be necessary. I have never done a pull request before, but i will try to create one with all the changes needed.
In some rare cases this could cause problems when a audio codec is not supported by the mka container. The only thing that comes to my mind right now tho is DSD audio formats, but i have never seen one in a video file...