In my recent assignment, I wanted to have a narrator for voice-over of the action. I've made videos before for webinars, and I have a wonderful Yeti Blue microphone for the purpose. Now where did I put that microphone…? Without the wonderful microphone, the recordings of my voice-over were faint and poor quality.
That's when I discovered that Vyond has a text-to-speech feature. I had a wondrous array of voices from which to choose: female or male; American, British, Indian or Australian. All you have to do is type in a text box the speech you want, and it is spoken–and lip-synced!–by the character.
Except. That's not the end of it. I chose a female Indian voice to begin with, as it fit the character of Donelda Futura. But there were instances where the voice was difficult to understand. When I tried out a few other voices, I realized that some are more distinct than others, as well as having different cadences. I changed to an American voice, since I wanted my classmates to understand her better.
The surprise to me was that the speech comes out differently depending on how you type it. For example, one phrase I used was, “You left your cell phone in your own time, of course.” Written that way, it came out as one sentence, with no pause. I changed it to two sentences, and it sounded much more natural: “You left your cell phone in your own time. Of course.” I changed several more sentences as well, and the final version was much improved.
The moral of the story is, when you're working with a different medium, the rules you're used to don't necessarily apply. I assumed the text would be spoken just as I typed it, with all my grammatical instincts in play. But when you're creating text-to-speech, flexibility is all-important. Be open to new ways of working, and don't try to force your expectations on the tool. See how it works best, and use its strengths to come up with a better result.
Hmm, does that apply to life as well…?