mp3 - when ffmpeg Import, use higher quality float decoder

for Slightly better quality (which ? more important when editing than playback)
use -c:a mp3float option (before -i )
so that it decodes the mp3 (since it uses some frequency-based encoding)
into a 32-bit float format,

by default it decodes the mp3 into a 16-bit integer

By default, Audacity imports MP3 files with libmad, not FFmpeg. Libmad performs all calculations with 32-bit fixed-point integer representation.