Bustikated

Lately, I’ve been using the built in speech synthesis on OS X to read wikipedia articles out loud while I do other stuff. It’s cool, but requires either a little patience or “lower standards”.

If you’ve tried this, or you do try it, you’ll see what I mean. The layout of the articles isn’t designed to be read this way. The speech service wasn’t designed to parse HTML documents, but rather flat text.

Staring with the directions found here after a quick google search, I set up an automator action to request text, and then read it to the desktop as an aiff file.

I chose the San Francisco article as a good example, and copied the text into TextMate. I made some quick edits, adding periods after all of the headings, for instance, to force “Alex” to pause properly, and removed all of the footnotes.

The reading went well, after a few minutes I was presented with a 100+mb aiff. After manually adding it to iTunes & playing around with some quality settings, I settled on the AAC podcast setting. The convert ran quickly, and reduced the file size by almost 90%.

Finally, I went back in and added this action to my automator application, so that the iTunes import and conversion happens automagically. The important thing here is to remember to set iTunes’ conversion preferences before running the automator app.

Don’t forget to change the quality back to your normal music ripping standard!

Links

• Here’s the Disk Image containing my Automator Actions.
• The sample text file
• The compressed result