Hearing my WIP read out loud is one of the easiest ways for me to “see” what’s wrong with a scene, but I dislike reading it myself and I double-dislike forcing my friends to do it. Especially when I’m editing at 3 a.m. on a Thursday. So I am a big fan of text-to-speech software, as painful as it is to listen to. If a passage sounds good in “mechanical voice,” it must be good.
Some writing software includes text-to-speech modules (including some early versions of MS Word), but I do it for free at sites like YAKiToMe.com (you can find others by googling “free text-to-speech” or start with this listing from U. of Texas ). I cut-and-paste a scene or chapter, name it and hit the “convert text to speech” button. A minute or so later, I have an audio file I can download or just listen to on-site. I listen on-site, then delete it so I can re-use my clever naming convention (“Chapter 3”) the next time I upload that chapter.
These voices have better intonation and pacing than 20 years ago (remember Stephen Hawking’s first, affectless voice?), but they are still far from perfect: For a regency WIP, I’m using the U.K. speaker (“Audrey”), which gets “lieutenant” right as “lef-tennint” but pronounces “lady” as “lah-dee.” Still, it’s screamingly clear when I’ve used “truly” the third, fourth, fifth time in a scene. Delete, delete, delete.
Next time you’re up at 3 a.m. jonesing for a little recitation, try the online option. And if you use headphones, no one will hear the errors scream.
Post a Comment