Before we start, make sure that a microphone is plugged into the correct socket on the PC. This will usually be a dedicated 3.5mm connector or USB port – check the computer’s manual for details. Most modern laptops have mics built in. Now click the Start button and choose Control Panel. Click the Ease Of Access link and then, at the next screen, click Speech Recognition (Vista calls this Speech Recognition Options). At the next window, ignore the other available options for now and click the Start Speech Recognition link to begin setting up your computer to recognise voice commands.
Assuming this is the first time the speech-recognition feature has been activated, Windows will start the Speech Recognition wizard to lead you through the setup process. At the first dialogue box, click Next. Choose the appropriate mic type from the list – a headset is best because they pick up less noise, but the other two kinds should work just as well in most situations. Select the mic that most closely resembles the one you are going to use and click Next.
Depending on the choice of mic, Windows will offer some advice in the next dialogue box about positioning it properly in order to get the best results. Click Next to continue and at the next screen you will be asked to read out a sentence so Windows can work out the best volume for the microphone: speak so the level meter at the bottom reaches the green bar but doesn’t extend into the red, when pick-up will be distorted. Click Next to continue.
When the mic is set up properly, click Next. Windows will now offer a chance to improve the accuracy of its speech recognition by allowing the PC to pore over the documents and emails stored in its search index: unless you have a specific reason for doing otherwise, click Enable Document Review and click the Next button. At the next screen, Windows 7 users can select ‘Use voice activation mode’ radio button – this will allow the speech-recognition feature to be turned on and off using voice commands (and that only seems appropriate). Then click Next.
Windows will give you the chance to look at a handy reference sheet of voice commands – click the button to open it in a new window and click the Print button at the top to create a hard copy of the information if you want. Regardless, we will explain how to open variations on this set of commands later on, using your voice. Close the window by clicking the little ‘x’ at the top right-hand corner. Then at the next screen click to place a tick next to ‘Run Speech Recognition at startup’ and click Next. Finally, click the Start Tutorial button.
This will launch the Windows Speech Recognition Tutorial, which is split into four main sections: Basics (the fundamentals of starting, stopping or pausing the service), Dictation (editing text, navigating round documents, opening menus, choosing options and so on), Commanding (things like starting and closing programs) and Working with Windows (which explains how to navigate round the operating system itself). As Speech Recognition is active, just say ‘Next’ to get started.
The tutorial takes about 30-40 minutes to work through and completing it will improve the accuracy of Windows’ speech recognition. However, if you are keen to get started straight away, close the tutorial (by clicking the red cross at the top right). When the speech-recognition tool is running and active, a small bar appears at the top of the screen: a grey/blue microphone button means it’s sleeping, while an orange button means it’s giving you feedback. Say ‘Start listening’ and the button will turn bright blue indicating that it is ready for action. To turn it off, say ‘Stop listening’ and it will go back to sleep.
With speech recognition on, say ‘Start Wordpad’. The basic word processor that’s included with Windows will launch, displaying an empty document. Let’s try some simple dictation. Here’s a good example of what can go wrong. In this sentence, Windows has entered ‘C’ instead of the word ‘see’. To fix it, we say ‘Select C’. When the word is highlighted we can then say the word ‘see’ and Windows displays a dialogue box showing suggested replacements. To correct the mistake we say the number associated with the right word (‘1’) and then say ‘OK’.
Having corrected the mistake we now need to move back to where the text cursor was and carry on dictating. To do this, we can say ‘go to end of sentence’ and the cursor will jump there. Then we’ll say ‘period’ to add a full stop and ‘new line’ to open a line space and start a new sentence. Then we can dictate a new sentence, correcting any mistakes as we did in the previous step. If Windows gets a phrase wrong, say ‘delete that’ and it will remove the last instance of dictation.
Say ‘What can I say?’ and a window will pop up that contains links to various types of voice-navigation commands. But hang on – how do you navigate between the different links in order to find the one you want and open it? There is a clever feature that helps with this. Make sure the Help window is open, as shown in our screenshot, and then say ‘show numbers’ – each link is briefly overlaid with a number.
We want to open the Controls section so first we say ‘18’ (the number associated with it) and when it’s highlighted, ‘OK’ to open it. If the desired information we want is further down the window, we can say ‘scroll down’ to move down a bit at a time. Once we have finished we can say ‘scroll up’ or ‘scroll to top’. Then we can close that section by saying ‘show numbers’, saying the relevant heading number again and saying ‘OK’. Once you have finished with the Help windows, just say ‘Close that’ and it will disappear.
So far, so straightforward. However, difficulties will arise from time to time. Imagine you have said a word that the speech-recognition feature has interpreted incorrectly. By uttering the ‘Select [word]’ command explained in Step 8, a list of replacements pops up – but the correct spelling isn’t there. In this case, say ‘spell it’ and read out the word, letter by letter. If you need an upper-case letter, say ‘capital M’, for example. If Windows doesn’t hear a letter correctly, say ‘undo’ and try again. When you have finished spelling the word, say ‘OK’ and it will be corrected.
Now try something else. With Wordpad open, say ‘show numbers’, choose 11 (by saying ‘11’) and then say ‘OK’. This will take Wordpad out of full-screen mode. Next say ‘start calculator’. When that application appears on the screen, say ‘start paint’. There are now three different programs open on the Windows Desktop. To move between them, say ‘switch to’, followed by the name of the program – ‘switch to calculator’, for instance. Similarly, to close a program, say ‘close’ followed by its name.
It’s even possible to use speech commands to move items around the Desktop. Try this. Using the mouse for a moment, right-click, choose View and ensure the ‘Auto arrange icons’ and ‘Align icons to grid’ options are not ticked. Then open the Pictures folder and drag a photo file into the centre of the screen. Say ‘mouse grid’. Windows will overlay a grid on the screen – the picture should be in the middle of the 5. Say ‘5 mark’ to highlight the picture and then ‘7 click’ to move it into the bottom left-hand square. Finally, if you wish to run through the speech-recognition tutorial now, repeat Step 1 and choose Take Speech Tutorial.