Mac OS: Using Text to Speech and Speech Recognition

How do I use the Text to Speech and Speech Recognition functionality of the Mac OS? I have Mac OS 8 installed on my PowerPC Macintosh.
This article has been archived and is no longer updated by Apple.
Using Text to Speech

The TTS software running on the your Macintosh allows speech-enabled applications to speak selected text. The way this typically happens is you select some text (or an entire document), then select the application's Speak command. The TTS software in the System Folder then takes over and reads the text aloud. That's it!

Generally speaking, the Speech command will likely be in a menu or tool bar if the application supports TTS. The voice menu or list of available voices will usually be close by.

In SimpleText, use the Sound menu:

Other applications may or may not support speech. If youare not sure if your application support TTS, you should contact the developer of the application.

Mexican Spanish TTS uses the same menus and control panel as English TTS. To change from an English voice to a Mexican Spanish voice, simply select Carlos or Catalina from the list of available voices.

Using English Speech Recognition (SR)

This section explains how English SR works. As we progress through the process, remember SR can only recognize commands that are in the Speakable Items folder (with one exception: the OK/Cancel check box in the Speech control panel allows button names to be understood by SR).

Introduction to Speech

The first time the computer starts up after SR has been installed, the Apple Guide-based Introduction to Speech tutorial automatically opens:

This tutorial walks the SR user through the basics of turning on and configuring SR, how to speak to the computer, how to read the SR indicators, how Speakable Items works, etc. If novice SR customers call you with basic SR problems, it's a good idea to have them run through this tutorial. As a support rep for Apple, make sure you go through it when you work on the PlainTalk hands-on lab (a companion to this self-paced training).

So How Does It All Work, Then?

The Feedback Window

When your SR-enabled computer starts up, in addition to your normal desktop, you'll see a floating SR feedback window:

The face you see above corresponds to the feedback character previously selected (in the Speech control panel). In this example, "Tiki" is the SR name of the computer. "Speakable Items is ready" indicates the SR software and hardware have started up and are waiting for the user to speak a known command.

Speakable Items

So what counts as a valid command? Any item in the Speakable Items folder. This is a special folder installed by the SR installer in the Apple Menu. Any item in Speakable Items can be used as a command. For example, a user can put an alias of the Indigo beep sound in to the Speakable Items folder, issue the spoken command "Indigo" (or "Open Indigo"), and SR will play that sound, just as if the user had double-clicked that icon.

Simply put, any item in Speakable Items can be opened by speaking its name. That is how SR works.

The Speakable Items folder comes with a set of commands pre-installed. There is a More Speakable Items folder in the Apple Extras: Speech Technology folder. Not surprisingly, it contains additional commands you can drag into your Speakable Items folder.

What Happens When You Speak

When you speak, the PlainTalk microphone picks up your sound input and sends it to the SR hardware and software for recognition processing. How can you tell if your microphone is working and SR is listening? You'll see sound waves like these:

These sound waves to the left of the character indicate that computer can hear you. Sound waves are important indicators when troubleshooting SR. If you do not get any sound waves, you either don't have the mic connected, you don't have the right type of CPU, or the SR software is off or misconfigured.

To tell the computer you are ready to speak to it, speak its name. Look at the feedback character; if its eyes are open and you see sound waves, it is trying to interpret your speech. If its eyes are closed, SR is in standby, and nothing will happen until you wake up SR by using the keystroke set in the Speech control panel, above.

When the computer recognizes its name and is ready for your command, you will see its name ("Tiki" in this example) disappear from the feedback window. Whatever you say next will be interpreted. If it is correctly interpreted AND matches up with a speakable item, SR will open (execute) that item (command).

Adding Speakable Items

Simply put, adding icons of any kind to the Speakable Items folder adds commands to your SR system. As you may have noticed in the Speakable Items folder window, many commands are AppleScript macros. You can use the standard AppleScript Script Editor to create you own macros or edit existing ones.

Both English and Mexican Spanish TTS read aloud selected text when used with applications that support speech. Basically you simply highlights text, then select the application Speak command.

Speech Recognition is more complex. To speak commands and have SR execute them, you speak the computer's name, then the command. Commands are limited to what's in the Speakable Item folder. Add items to that folder to increase the number of commands you can have understood.
Published Date: Feb 18, 2012