This is all because people believe what they see on TV. You hear about “voice prints” as though they’re like fingerprints. With audio recording, especially cell phones, all you have is the equivalent of a “smudged fingerprint”...you have it but can’t really determine much because there’s too much distortion. By the time you go from having background noise, a crappy microphone, audio compression algorithms in the cell phone, etc. there’s no way one can determine, with accuracy, who it is from such a short recording.
I think it is (almost?) criminal for the prosecution to even attempt to use such “evidence”. It takes advantage of uneducated jurors who might believe anything an “expert” says. If the jury had heard this opinion it would have been a complete injustice and grounds for appeal. It really bothers me that this was even attempted. It is not science.
To say nothing about a voice screaming is not anywhere near the same voice talking. All the frequency peaks are shifted, nothing matches. Any subtle intonation is gone as the throat is wide open and everthing is under stress.