Text to Audiobook free script
The idea of being able to select text and have it automatically converted into an audiobook using text-to-speech isn't a new one, but it is a good one. I mean, who doesn't love the idea of being able to take longer text on the road with you and listen to it in the car or on the bus?
Well, if you loved the idea as much as I did, but didn't think you'd be using it enough to justify coughing up some dough for it, you're in luck. Over at Mac OS X Hints, user miketyson has created a script that will add a service to your Mac's Services menu. The new service, which is appropriately called Speak to iTunes Audiobook, allows you to select a body of text and automatically have it converted into an audiobook, then have it imported straight into iTunes.
In testing the script works as advertised. My only quibble is that the clipped audiobooks don't get any sort of meaningful name given to them. They are simply named Speech 1, Speech 2, etc. If the script was smart enough to look at the title of the document the text was being clipped from and generate a title from that, this would be a killer utility. But that's really only a minor complaint considering the heavy lifting this little script actually accomplishes. And heck, if you find yourself using it, you might consider going back and giving GhostReader a second look.

![TUAW [Cafepress]](http://www.blogsmithmedia.com/www.tuaw.com/media/tuaw-cafepress-promo.png)


Reader Comments (Page 1 of 1)
Alex said 4:39PM on 6-08-2008
I did the same thing (though probably not in quite as slick a fashion). I converted the Constitution of the United States to audio format using Automator. It took an hour or so because I manually copied and pasted each section of each article, and then manually renamed each part, and then manually set some track settings in iTunes, etc.
The end result was approximately 25 audio tracks in iTunes, covering 28 or so minutes, taking up 75.4 MB.
Reply
Tony Barnhill said 4:47PM on 6-08-2008
Apparently you guys don't know that there is a built in automator action to do this since 10.4.
Reply
cajebo said 4:55PM on 6-08-2008
hmm, can't seem to find the script...
Reply
toypaj said 5:17PM on 6-08-2008
http://www.macosxhints.com/article.php?story=2008050623341520&query=audiobook
Reply
David Sinclair said 6:01PM on 6-08-2008
If you want to make an audiobook from text with synthesized speech, that features multiple voices (for multiple characters in a story or participants in an interview), check out Dejal Narrator 2, which requires Leopard.
http://www.dejal.com/narrator/
Reply
matt said 4:37AM on 6-09-2008
Yeah, a story without a link to the script isn't terribly useful.
Fortunately, it's easy to find. Alternatively, you can just use Automator, like another poster mentioned.
Reply
Jason Clarke said 8:31PM on 6-09-2008
Thanks Matt - I hadn't realized that I forgot to link to the script in the body of the post, though we always provide a link at the bottom of our post titled "Read".
In any event, the link is there now.
Dave said 10:25AM on 6-09-2008
Even the best sounding Leopard voice isn't up to the task of reading anything like a novel. No emotion, no context recognition, and some words just don't sound right, even when spoking correctly with Leopard's voice to text.
I've been playing with Alex for a while for these kind of tasks, and disappointed so far. Cepstral has some supposedly superior voices for this kind of stuff, but the samples I played with have the same issues, no lack of comprehension of what is being read, which is what is required to not sound like a computer.
When they finally port over the software that some of these big businesses are running on their voicemail systems, we might get some natural sounding voices for audio books created like this, until then the experience will never match a professional reader reading the text to you.
For emails and the like, maybe this is fine, but the expectations are lowered for that. Here's to a future where your emails are read to you in the voice of the person who sent it! ;-0
Reply