Submit Hint Search The Forums LinksStatsPollsHeadlinesRSS
14,000 hints and counting!

10.7: Automator Services for multi-language speech System 10.7
OS X Lion now has many different languages available to install in your speech preferences. This is something I've wanted for a long time, because I have long aspired to learn Korean and Japanese. Being able to easily hear things spoken in those languages would really help out.

The problem is that Lion will only speak selected text using the voice that you set as default. I have Alex set as default and when I try to have it speak text using the built-in context menu item, it does nothing.

There is a way around this by using Automator.

First you need to go into the Speech preferences and install any languages that you want to use. Next, open up Automator and make a new Service that receives the selected text in any application. Then add the 'Speak Text' action to the workflow and choose the voice that you want. Finally, save the service with a name that tells you what language it can speak.

Next time you come across some text in that language that you want to hear, select it and right click. The service you made will be in the context menu and you can click it to hear it speak the text you selected.

[crarko adds: You'll need to make a separate Service for each voice you want to use, or write a script that lets you select a voice and use it in the Service instead.]
    •    
  • Currently 3.60 / 5
  You rated: 5 / 5 (5 votes cast)
 
[5,072 views]  

10.7: Automator Services for multi-language speech | 3 comments | Create New Account
Click here to return to the '10.7: Automator Services for multi-language speech' hint
The following comments are owned by whoever posted them. This site is not responsible for what they say.
10.7: Automator Services for multi-language speech
Authored by: scottbayes on Aug 10, '11 12:32:19PM

Wow! To me, this is absolutely astounding.

(I had written an involved discussion of Japanese pronunciation, but it's not appropriate here. Suffice it to say that the Japanese voice Kyoko appears to trigger the speech engine to do some reasonably deep analysis of sentence structure and "kanji" character combinations, and chooses mostly correct "readings", not an easy task in Japanese. This may be done by some internet service, though I don't know of such a service).

My wife listened to the Japanese voice Kyoko speaking some text from a Japanese website, and though it had some errors (even I caught some of them), she was pretty impressed at how much it got right, given the fact Japanese requires alternate readings of the same character, depending on context and sentence structure.

All this is hiding behind Apple's simple statement "VoiceOver in OS X Lion includes built-in voices that speak 22 languages…"

I wonder how well it does for other non-Indo-European languages.



[ Reply to This | # ]
10.7: Automator Services for multi-language speech
Authored by: dat5h on Aug 11, '11 12:54:47AM

I have to agree that the kyoko voice does a pretty good job at grabbing the appropriate pronunciation for kanji compounds. That being said, people should not rely on these voices to interpret proper-names that can be pronounced in more than one way since it may (and certainly does occasionally) guess wrong. Also, these voices have a tendency to neglect the proper intonations (rising and falling pitch), but the new voices are noticeably improved [not perfect] over the older voices from previous versions of OS X.

The integration of these voices is a huge improvement, but it makes me wonder which side of the "uncanny valley" are we approaching?



[ Reply to This | # ]
10.7: Automator Services for multi-language speech
Authored by: sugarquay on Aug 26, '11 03:56:38AM

Rather than making a separate Service for each voice or writing a script to enable you to choose a voice, simply add this simple step to the above hint:

After adding the 'Speak Text' action, click its 'Options' and check/select 'Show this action when the workflow runs'.

You may still select a particular voice from the list to be your default if you wish, but now each time this Service runs, you'll get a small dialog box asking you to choose a voice from all those installed on your system, and your selected text will be spoken using whichever voice you choose.



[ Reply to This | # ]