Recognizes a single finished Latin track as you actually hear it — its vocals, instruments, and overall sound.