When it comes to PCs, we often think the only input methods are the keyboard and mouse. Yet, your voice can be surprisingly effective and fast. This is especially true when you dictate documents using Office Speech Recognition with Microsoft Word. Voice input can also be valuable if you run into issues where typing is difficult because of illness or injury.

When using your voice as an input method for Microsoft Word, there are several points to remember. Don’t expect your voice to fully replace your computer keyboard. Even though it is your voice, the recognition isn’t perfect. You’ll still need to make corrections and edits as before. Just like humans, the software at times misinterprets your speech.

Getting Started with Speech Recognition

The first step is to find a good microphone. For example, don’t use the cheap one that came with your PC and plugs into the audio jack of your PC or sound card. It probably won’t work. Instead, find a quality microphone that plugs into a USB jack. For example, I use the same headset that I wear when making VoIP phone calls. I think I paid $30 for this Logitech headset.

The second step is to make sure you have Office Speech Recognition installed in Microsoft Word. The feature has been around since Word 2002, but isn’t available with all configurations. Specifically, if you use Windows XP and Office 2007, you do not have this functionality. You either need to use the voice recognition program built into Microsoft Vista or use an older version of Microsoft Office. (Note: Windows Vista users might want to read Adam Pash's Lifehacker article,"Control Your PC with Your Voice.")

You can go to the Tools menu and select Speech. If the feature is installed, you will see a small dialogue box like the one below asking you to train the speech recognition system.

The 15 minute training process steps you through each area. At the conclusion, the program will open Internet Explorer and go to a Microsoft web page with more details. There is also a nice collection of Word 2003 Speech Recognition “how to”s on the Microsoft site.

During this training there are three main points to remember:

1. You need to be an environment with little background noise.

2. It’s best to speak at a constant and easy pace. The system is better at understanding words in context rather than individual syllables.

3. Don’t worry about making mistakes. If the system doesn’t understand your words during training, you can reread the selection.

Using the Language Bar

After you complete your voice training, you’ll see a small toolbar appear. This is the Language toolbar that provides commands and feedback. The buttons change based on your state. In the screen snap below, the toolbar shows all the buttons, but if I turn off the microphone, some buttons collapse.

One confusing aspect of the Language bar is that it shows in programs where it doesn’t work such as Firefox. There are five main buttons on this toolbar.

Toolbar Button	Purpose
Microphone	Toggles the microphone “on” or “off”.
Dictation	Allows you to speak your content.
Voice Commands	Allows you to control the program such as using menu commands.
Speak	Reads back your content.
Tools	Provides a menu with additional settings.

Before speaking, I would suggest printing some of the help topics from the toolbar found in the Speech commands book. These topics cover “Basic Input Commands” which you need to control punctuation and formatting. Many commands are obvious such as speaking “comma”,“period” or “new paragraph”. Others are not as intuitive such as when you wish to display a number instead of the word.

Apart from Dictation mode where the software interprets your voice and types the words, you can also control Microsoft Word by using Voice Commands. These commands vary based on which Microsoft Office program you’re using. For example, Word has a voice command for “select sentence”, but that command wouldn’t work with Microsoft Excel. When you speak a voice command, you’ll see the interpretation in the language bar.

Document Corrections and Read Back

One of the expectations people have with voice recognition is that it will work perfectly like it does in the movies. It doesn’t. Even if it did, there are times when you want to edit your words.

I’m not at the stage where I’ve fully mastered the voice commands, but I routinely use the dictation mode to write drafts and capture ideas. Just as I edit documents I type, I need to make changes to these documents as well. Sometimes, the corrections are specific to the translation. As example, when I spoke “USB jack” above, the software capitalized “Jack”. I’ve also goofed at times when using web addresses by using the phrase “forward slash” for “/”. The software responds with “forward /”. After a bit, you learn how to adjust to the system.

One hidden benefit that I didn’t realize was I could use the software to read back my documents. My first thoughts were I would hear the article read in a computer voice. I was shocked to hear my own voice. Portions do have the synthesized voice where I used the keyboard instead of my voice. When you hear your voice, you also hear the input commands such as “new line” or “comma” read back. This doesn’t happen with the computer read version.

Your voice file doesn’t get saved with your document. If I close Word and opened the document I would hear just the synthesized voice. Unlike the initial read back which includes your voice and may sound disjointed at times based on your edits, the saved version is more fluid.

Office Voice recognition isn’t something that will change your life from day 1. However, I think you’ll be shocked at how well it can translate your words into type. I consider myself a good typist, but I can’t beat my voice. The accuracy is better than I expected although maybe not as good as some commercial programs. I would certainly encourage you to give it a shot.

Related Microsoft Word Tutorial

How to create Microsoft Word Keyboard shortcuts