A layperson’s exploration of all things voice

April 2, 2019

How People Prefer Human to Synthetic Voices

It shouldn’t come as a shock that humans prefer to hear human voices over synthetic. That’s logical. Yet, when building out new skills, some companies are using synthetic because it’s faster & easier to put together. Most synthetic voices use a technology called text-to-speech (TTS) – so you can essentially just use text to create your synthetic voice. That’s much easier than recording a bunch of audio files.

Hiring real “voice talent” – people trained in the “voiceover” field – can be costly. Using “normal” humans to serve as the voice of your skill can sound bad. So you’ll have to run a cost-benefit analysis and see what works best for you. Check out Susan Westwater’s discussion of this topic at the 32:00 mark of this video (ie. depends upon the brand’s personality, etc.)…

One last thought – sometimes a synthetic voice may be preferable over a human one. That works in short bursts of time. For example, in this Voicebot podcast with Drivetime’s CEO Niko Vuori, Niko describes how his customers love the synthetic scorekeeper – whom they have named “Miles” – in his trivia game. The scorekeeper says 3-4 words at particular intervals in the game. To me, that is the perfect length of time for a synthetic voice…