Dump audio files of table to single file

[frankenjoe]

It might be useful to dump all audio files of a filewise to a single audio file (or array). If timestamps are provided (e.g. as column in the table) the function should be able to respect the temporal alignment and insert silence between adjacent segments. The case of overlapping segments needs to be handled, e.g. by mixing.

[frankenjoe]

Maybe it's also sufficient that we introduce a generic utility function that takes as input a list of audio files and timestamps and mixes them to a single audio file.

[hagenw]

This is related to audformat.utils.to_filewise_index(), which is more or less the opposite. The main difference you are proposing here is that you would like to create only a single file in the end, correct?

[frankenjoe]

The idea for such a function came up when we were logging user and system turns, where we ended up with a list of audio chunks and timestamps. From that we wanted to create a segmented table, which required us to concatenate the audio chunks to a single audio file.