You are in:

Apple Mac

There are two main options for speech recognition on the Mac:

  • the Dictation system built in to recent versions of the macOS
  • and the commercial Dragon program from Nuance.

These both allow you to dictate sentences and have your speech converted into text. In addition, the Siri voice interface used on iPhones and iPads was added to macOS Sierra in late 2016. The macOS version of Siri can be used for sending very short messages and interacting with your computer, but it cannot be used for dictating passages of text.

Setting up the Microphone

You have to set up the microphone before using any speech recognition program on your Mac. We recommend using a good USB microphone, such as the Andrea NC 181VM microphone (£29.95 from Iansyst).

Make sure that your Mac is using the correct microphone by clicking on the Apple icon in the top-left of your screen, then selecting System Preferences, followed by Sound. Click on Input and make sure that the correct microphone is selected - make sure that the Input level indicator moves significantly when you speak.

 

Dictation - Apple's Free Option

The Dictation facility on the Mac can be used wherever you can type text as an option to using the keyboard. To activate it, click on System Preferences, followed by Keyboard. You will then be given an option to choose keyboard preferences, including an option to use Dictation.

Click on Dictation, then turn Dictation On to make it available. It is a good idea to select the Use Enhanced Dictation option, otherwise you will need to have a good wifi connection to use the program. Enhanced Dictation is also considerably faster to use as you don't have to wait so long for your words to be converted into text. It is good practice to set a keyboard shortcut, e.g. "Press Right Command Key Twice" to use speech recognition. You are now ready to try the Mac's Dictation option.

Open a word processor, or another program into which you can type, then wake Dictation up by clicking on Edit, then Start Dictation, or using the keyboard shortcut you set up earlier. Now dictate a few sentences of text. Remember to add punctuation by voice, e.g. saying "full stop" at the end of a sentence. The Apple Support page on Dictate includes a list of punctuation you can use.

In addition to dictating text, there are a number of Dictation Commands that you can use to navigate within your document, make basic edits and move between programs. You can also use voice to access a range of program-specific editing commands that will change depending on the program you are using, e.g. you can use commands such as "Open New Word Document" and "Align Text Left" in Word. You can see the dictation commands available by clicking on the microphone icon at the top of the screen, followed by Show Dictation Commands.

What can Dictation do?

  • Recognise your voice and convert your speech into text without any need to train the system.
  • Dictate directly into any application into which you can type, including text boxes in PDF files.
  • Adapt to your voice, though you can't train the system to recognise how you pronounce individual words.
  • Open, close and move between different applications, allowing you to dictate required commands within each.

What can't Dictation do?

  • Correct mis-recognised words by voice.
  • Play back a recording of what you dictated.
  • Read back the text that you just dictated, as interpreted by the system (though you can use the Mac's built-in text-to-speech system to read the text out loud).

Is Dictation any good?

People who have previous experience of speech recognition will find it very easy to use, with no need to train it to your voice, and no need to fiddle around with settings. A total novice may find it a little harder to get to grips with as there are no clues to getting started on the screen.

It is pretty accurate, making only 8 mistakes in a 100-word test passage. These included the addition of one small word, the omission of another and failing to add a couple of apostrophes.

 

Dragon - The Commercial Option from Nuance

Dragon Professional Individual, v6 costs £249.99 and is available from the Nuance website. An Education version is also available, costing £124.99. It can be purchased either as a download, or as a boxed DVD. Note that a microphone is no longer supplied.

When you open Dragon, you will get a mini toolbar with a microphone icon which turns from red to green when the microphone is on, and icons that let you choose dictating mode and whether you show available commands on the screen. The interface is much simpler and more basic than on the Windows version. You may be asked to do a Microphone Setup, dictating a few sentences to set input level on the microphone, but after that you are free to dictate into most applications.

Unlike the Windows version, there are relatively few settings that you can change in Preferences, though you may want to set a key to act as a shortcut to turn the microphone on and off.

What can Dragon do?

  • Accurately convert your speech into text without any need to train the system, though there is an option to train a voice profile which will improve recognition if you have a distinctive voice.
  • Dictate directly into any application into which you can type, including text boxes in PDF files.
  • Correct mis-recognised words by voice
  • Transcribe an audio file of speech into text, provided that there is only one speaker.
  • Adapt to your voice, also allowing you to train the system to recognise how you pronounce individual words.
  • Open, close and move between different applications, allowing you to dictate required commands within each.
  • Control the mouse, either by using a MouseGrid command to position the mouse, or movement or action commands, e.g. "Move mouse left", "Double-click".

What can't Dragon do?

  • Play back a recording of what you dictated.
  • Read back the text that you just dictated, as interpreted by the system (though you can use the Mac's built-in text-to-speech system to read the text out loud).

Is Dragon any good?

Dragon on the Mac has fewer options for 'fine tuning' than the Windows version, but this may actually make it easier to use. Unfortunately, the lack of the Play Back and Read Back in the Windows version makes this less useful for people with reading and writing difficulties.

It is very accurate in general use, making no mistakes in a 100 word test passage. When it eventually made mistakes, the correction by voice system was more basic than the Windows equivalent and would be frustrating for somebody relying on voice for making corrections.