Funny how if take time for longer conversations, ones that ramble, you often learn interesting things that you just wouldn’t learn any other way. My friend Don was in town for SQLSaturday #40 a few weeks back and we were talking about blogging. An interesting conversation because Don is a professional writer (professional = get’s paid for doing it), and so for him, the idea of casual writing is tough, he’s various conscious that if doesn’t write well (thought, spelling, grammar) even on a blog, that might translate to people assuming that his books are not well crafted either.
As we talked Don always reviews his work methodically, then has someone else do it, and then it goes to a real editor for review and further work. I asked how he manages his own reviews, since for me there is nothing harder than proof reading my own work. The answer was speech to text, he said he often caught things by listening that he didn’t think he would have found reading. Not just spelling or grammar, but the overall flow of the writing as well.
Interesting. I’ve ignored the voice features of Windows for, well, forever. As much as I’d love to be able to say in my best Picard voice “computer: playback current draft using voice of James Earl Jones”, it’s never been worth – to me – the pain of dealing with it. Time to revisit.
Don uses Final Draft ($249), a product that seems to focus on screen writing but serves his needs. I didn’t give it a try, figured I’d focus on things I already owned.
I started by searching the Windows help and that took me to Narrator (Programs | Accessories | Ease of Access) and running it (see next two images). Clunky. Voice is not great but understandable, had a hard time getting it to read text in an Office 2010 doc, worked a lot better when I pasted to Notepad. There are other voices in something called a SAPI5 format, here’s a link if you want to try some http://www.bytecool.com/voices.htm. I suspect a different voice doesn’t fix the clunky reading.
Next I looked in Office 2010 for a speech to text option, and finally found how step by step instructions. The short story is that you have to add the command to the toolbar (or ribbon) first, then you just click but speech button to start/stop play. Same voice, same API I’m guessing, worked smoother (but maybe that was just me).
Maybe a better product than Windows/Office would make it more interesting. I found Natural Reader with a quick search and they let you do a playback test of text you provide right on the web site. They have a free version, one for $49, one for $99. The $99 version lets you export to MP3, so I could see it might be interesting to drop it on a portable player or phone and do the review while commuting (though then you need a way to make notes about problems).
I’m not a convert yet, at least not for blog posts!