Forensic speaker recognition using traditional features comparing automatic and human-in-the-loop formant tracking