Music Assistant adds an optional Talk over the intro setting to AI Radio segments. The end of the host’s spoken break can play over the start of the next song, with the music lowered beneath the voice and the overlap ending before the first sung line.
The timing follows the version of the song that actually plays, using its audio analysis or synchronized lyrics. If there is no usable timing, or the intro or spoken break is too short, the whole break plays separately as before.
Marcel van der Veldt supplies the server work, alongside jpittelkow’s first contribution to the frontend. The setting starts off and is chosen separately for each segment.