I have a spoken interview (2 males) recorded as a sound file. I wonder if there is a way to use Audition (or perhaps some other audio software) to, first, establish a ‘footprint’ of each of the two voices based on their individual timbres; and second, highlight the waveform to show speaker identity, i.e. who speaks when, for instance use red-tinted highlight for speaker A, and green for speaker B.
This may well not be doable with good-enough accuracy, but it would be helpful to know what I could try. Thanks in advance for any tips!
Update: an acceptable solution I found for now is to view the recording not in the Waveform Display but in the Spectral Frequency Display. At least for the present two voices, the way each is rendered visually with the spectrum colours is sufficiently distinct that one can recognise speaker identity based on the spectrum view – but for more similar voices, a more advanced solution might be needed, if anyone would still like to chip in.