Why a Human Voice?
I was recently searching for an article to share about why to use a human voice rather than #ai. Everybody
is talking about artificial intelligence: businesses who make #corporatevideos are salivating and #voiceover actors are panicking.
So, why WOULD someone pay a #voiceactor to do something they can do for free by typing in a few words? Well, I finally took the big leap and listened to Speechify, a free online #ai voice generator to find out.
I decided to plug in the script from one of my #narration demo scripts to compare it to my own demo. I was a little nervous, if I'm honest. What if it really was just as good as a human? What if it was *gasp* better than me? Anxiously, I waited a few seconds for it to process the text.
I listened...
I'm gonna toot my own horn here and say that in the battle of Lisa-Marie vs. #AI, Lisa-Marie wins. (Really, though, any human with a heartbeat would win in this battle.)
The AI voice was "realistic" in that it had the timbre of a human voice, but that is where it ended. It was terrible! The voice was sort-of "upbeat," missing the point of the script, and wooden.
Imagine your #corporatevideo narration or #explainervideo contained the words "your health matters to us." Now imagine how you would like your patients and your customers to hear those words - you probably want the listener to feel as if you really cared, as if your organization really cared. What I heard from the AI voice was "YOUR...health matters to...us."
Is that how you want your customers to perceive your organization or your company? I didn't think so.
I tried switching to a different voice (there was a list of 10 choices: 5 American English and 5 British English). The second voice was better. The beginning of a sentence was fairly human, in that the voice modulated it's pitch. There is even a mimicked breath-sound somewhere in there. But I had typed in the script exactly as I had received it, with the same punctuation, and after a comma, this voice suddenly became robotic again.
I played with it - trying different voices, speeding up and slowing down, adding different "speaking tones," even one labeled "empathetic." (Can AI even have empathy??) And not only did it never sound anything like what I did in the booth, it never really sounded, well, human.
Now, I realize that this was the free version, that it is still fairly early days of AI, and that it isn't the only voice generator out there, but I was genuinely surprised at how inhuman it sounded.
I love the art and craft of working in the booth - trying different approaches until I get exactly the right tone and feeling you want to convey. This is what humans do. I feel relatively safe...for now.
Contact me for a custom demo. I'll even give you the human touch. ;-)