Microsoft KB Archive/306537

From BetaArchive Wiki
Knowledge Base


Article ID: 306537

Article Last Modified on 5/7/2007



APPLIES TO

  • Microsoft Windows XP Professional
  • Microsoft Windows XP Home Edition



This article was previously published under Q306537

SUMMARY

This step-by-step article describes how to install and configure speech recognition.

The computer manufacturer may have installed the Microsoft speech recognition engine if you recently purchased a new computer. In that case, you do not have to install anything else. If you installed Microsoft Office XP, or if you purchased a new computer that has Office XP installed, the speech recognition engine is included, but the engine may not be installed.

For speech systems to work correctly and for optimal results, you must set up the components correctly. Speech capabilities are designed to work with system default settings. Except for the physical connection of speakers and microphones, all other aspects are intended to work automatically. Some computers are also equipped with built-in speakers and microphones. In such cases, you do not have to configure anything.

To make sure that the engine is installed correctly, follow the setup guidelines that this article describes. If the engine is not working correctly, see the "Troubleshooting" section.

For additional information about how to use speech in Windows XP, click the following article numbers to view the articles in the Microsoft Knowledge Base:

306901 How to use speech recognition in Windows XP


306899 How to use speech recognition profiles in Windows XP


306902 How to use text-to-speech in Windows XP


306993 How to use the Language Bar in Windows XP


For additional information, click the following article number to view the article in the Microsoft Knowledge Base:

278927 WD2002: Part 1: Speech and handwriting recognition frequently asked questions


For more detailed information about how to use Microsoft speech recognition, click the Help button on the Language bar.

For the most up-to-date information about speech recognition developments at Microsoft, visit the following Microsoft Web site:

Back to the top

Speech recognition requirements

To use speech recognition, you must have the following hardware and software components:

  • A high-quality close-talk (headset) microphone. A USB microphone is recommended.
  • A 400 megahertz (MHz) or faster computer with 128 megabytes (MB) of memory.
  • Microsoft Internet Explorer 5.0 or later.
  • The speech recognition engine must be installed. It is available with Office XP, but it may not have been installed.


Note Speech recognition engines are language-specific. The first three Microsoft speech engines that are available are Simplified Chinese, U.S. English, and Japanese. Engines for other languages will become available.

Back to the top

Setting up your hardware

This section describes how to set up your hardware.

Back to the top

How to set up a microphone

Microphones vary greatly in design and purpose. They will continue to develop and become more specialized. See the microphone manual for hardware and software specifics. However, most models install in a similar manner.

  1. Locate the sound connections, and then connect the microphone jack to the computer. Most computers use an internal sound card, and the connections are frequently in the back of the computer.

    These connections are a series of connections that are the same size and diameter as the microphone jack. To label the connection as the microphone connection, one connection has a small icon that looks similar to a microphone or a descriptive label.
  2. Plug the microphone into that connection.
  3. To test the connection, follow these steps:
    1. Click Start, click Control Panel, click Sounds, Speech, and Audio Devices, and then click Speech.
    2. Click the Speech Recognition tab.
    3. Speak directly into the microphone. The sound level should register in the Microphone Level indicator.
    If no sound is received after you connect the microphone, see the "Speech recognition problems" section for troubleshooting procedures.

Back to the top

How to set up speakers

Speakers vary greatly in design and purpose. They will continue to diverge and become more specialized. See the speaker manual for hardware and software specifics. However, most models install in a similar way.

To set up speakers, follow these steps:

  1. Locate the sound connections and connect the speaker jack to the computer. Because most computers use an internal sound card, the connections are frequently located at the back of the computer. These connections are a series of connections that are the same size and diameter as the speaker jack. In many cases, there are two sound out connections:
    • One connection is labeled as a line-out connection. Most speakers that require a separate power supply (such as an electrical adapter or batteries) use this connection. This connection is also used to export amplified sound to recording devices, including recordable CDs and tape cassette systems.
    • The other connection is for the non-powered speakers. Because the computer boosts the signal, powered speakers may be damaged if they are connected.
  2. Plug the speaker into the correct connection.
  3. To test the connection, follow these steps:
    1. Click Start, click Control Panel, click Sounds, Speech, and Audio Devices, and then click Speech.
    2. Click the Text-to-Speech tab, and then click Preview Voice to hear the currently selected voice. The text is spoken, and words are highlighted as they are spoken. If the speakers are working correctly, you hear the spoken words.
    If you do not hear sound after you connect the speakers, see the "Text-to-Speech problems" section for troubleshooting procedures.

Back to the top

How to configure the microphone

Microphone performance depends on a variety of factors, including the distance from mouth to microphone and the correct positioning of the microphone. Each manufacturer has specific requirements.

To configure the microphone, follow these steps:

  1. Click Start, click Control Panel, click Sounds, Speech, and Audio Devices, and then click Speech.
  2. Click the Speech Recognition tab, and then click Configure Microphone to start the Microphone Wizard. The features that are supported depend on the manufacturer of the engine.
  3. Follow the instructions that appear on your screen.


Note The Configure Microphone option is available only if the current engine supports microphone training.

Back to the top

How to set audio input device options

To set up microphone options, follow these steps:

  1. Click Start, click Control Panel, click Sounds, Speech, and Audio Devices, and then click Speech.
  2. Click the Speech Recognition tab, click Audio Input, and then click Properties.
  3. Click any one of the following options:
    • Use automatically chosen line: When you click this option, the input line is set to a default that is determined by the speech system. Because of differences in drivers, capabilities, and languages used, there may be variances in the selected option. The selected default may not work with all options. If the device line does not work correctly, manually select a new line. To do this, click Use this audio input line.
    • Use this audio input line: When you click this option, you can select a different line for audio input. The options include all audio lines for the computer. Not all audio lines are supported for speech.

Back to the top

How to select an audio input device

To select an audio input device, follow these steps:

  1. Click Start, click Control Panel, click Sounds, Speech, and Audio Devices, and then click Speech.
  2. Click the Speech Recognition tab, and then click Audio Input.
  3. Click one of the following options:
    • Use preferred audio input device: When you click this option, the input device is set to the default device for the computer. Click this option if you want your speech programs to use the same input device as all other sound devices for the computer. Frequently, computers have only one input device such as the headset microphone. This device is the default device in Speech properties. The default device is designated in the sounds or multimedia properties in Control Panel for each operating system. Additional information for the specific panel is available in the Help files.
    • Use this audio input device: When you click this option, you can select a different device for speech programs only. Click this option to change to a different input device. The drop-down list box is available only if other devices are available. In this drop-down list, click the device that you want to use. If you make a selection here, the default device is not changed for other audio programs. For example, you may want all speech input to use the headset microphone instead of an omnidirectional microphone.

Back to the top

How to verify the input sound levels

To verify the input sound levels, follow these steps:

  1. Click Start, click Control Panel, click Sounds, Speech, and Audio Devices, and then click Speech.
  2. Click the Speech Recognition tab. The Microphone box displays a linear register of the input sounds. If the microphone is accepting input, a bar that indicates the sound level appears. Ordinary speech should register slightly less than halfway on the graph. If no sound registers, make sure that the microphone is turned on. Some models have a mute switch. Also, verify the connections to the computer. Some computers have several locations to plug in a microphone.

Back to the top

How to change the input sound levels

To change the speech recognition volume input levels, follow these steps:

  1. Click Start, click Control Panel, click Sounds, Speech, and Audio Devices, and then click Speech.
  2. Click the Speech Recognition tab, click Audio Input, and then click Volume.
  3. A volume control mixer appears. Adjust the device to the level that you want.


Note Not all devices support this option in the same manner. Some devices do not support volume control and the Volume button is unavailable. Other engines may use their own interface. In those cases, follow the instructions that appear on the screen or that are documented separately with the engine.

Back to the top

How to select an audio output device

To select an audio output device, follow these steps:

  1. Click Start, click Control Panel, click Sounds, Speech, and Audio Devices, and then click Speech.
  2. Click the Text-to-Speech tab, and then click Audio Output.
  3. Click either of the following option buttons:
    • Use preferred audio device: When you click this option button, the output device is set as the default device for the computer. Click this option if you want to use the same output device for speech as all other sound for the system. It is also the default option for Speech properties. Frequently, computers have only one output device such as a pair of speakers. The default device is designated in the sounds or multimedia properties in Control Panel for each operating system. Additional information for the specific panel is available in the Help files.
    • Use this audio output device: When you click this option button, you can select a different device for speech programs only. The drop-down list box is available only if other devices are available. In this drop-down list, click the device that you want to use. When you do this, the default device for other audio programs is not changed. For example, you may want all speech output to go through your headset instead of the speakers.

Back to the top

How to set audio output device options

By default, this option is disabled. However, other speech engines may include advanced properties for audio line out options. In these cases, Audio Output is available. Follow the instructions that appear on the screen or in the documentation for the specific engine.

To set up audio output device options, follow these steps:

  1. Click Start, click Control Panel, click Sounds, Speech, and Audio Devices, and then click Speech.
  2. Click the Text-to-Speech tab, and then click Audio Output.
  3. Follow the instructions that appear on the screen.

Back to the top

How to install speech recognition

This section describes how to install speech recognition.

Back to the top

How to determine if the speech recognition engine is installed

To determine if speech recognition is installed on your computer, follow these steps:

  1. Click Start, click Control Panel, click Sounds, Speech, and Audio Devices, and then click Speech.
  2. Look for the Speech Recognition tab. If the Speech Recognition tab is available in Speech Properties, the speech recognition engine is installed. If it is not available, the engine is not installed.

Back to the top

How to install speech recognition from Microsoft Word 2002

  1. Start Microsoft Word 2002.
  2. On the Tools menu, click Speech. Speech recognition is now available in all Office programs and other programs that it is enabled for such as Microsoft Internet Explorer.

Back to the top

How to install speech recognition by using Add or Remove Programs

  1. Click Start, click Control Panel, and then click Add or Remove Programs.
  2. Click Change or Remove Programs, click Microsoft Office XP, and then click Change.
  3. Click Add or Remove Features, and then click Next.
  4. Under Features to install, double-click Office Shared Features.
  5. Double-click Alternative User Input, click Speech, click the down arrow, and then click Run from My Computer.
  6. Click Update. Speech recognition is now available in all Office programs and other programs where it is enabled such as Internet Explorer.

Back to the top

How to add speech recognition as a text service

Generally, speech recognition is added automatically as a text service when it is installed. Follow these steps only if you removed speech recognition as a text service and want to re-add it.

To add speech recognition as a text service, follow these steps:

  1. Click Start, click Control Panel, click Date, Time, Language, and Regional and Language Options, and then click Regional and Language Options.
  2. Click the Languages tab, and then click Details under Text services and input languages.
  3. Under Installed services, click Add.
  4. In the Input language list, click a language.
  5. Select the Speech check box, and then click an option in the list.

Back to the top

How to remove speech recognition

To remove speech recognition as a text service, follow these steps:

  1. Click Start, click Control Panel, click Date, Time, Language, and Regional and Language Options, and then click Regional and Language Options.
  2. On the Languages tab, click Details under Text services and input languages.
  3. Under Installed services, click Speech Recognition under the language that you are using, and then click Remove.


Note If you remove speech recognition, it is not deleted from your computer. Instead, speech recognition is no longer available as a text service and it is prevented from loading into memory. You can add it again later.

To remove the speech recognition engine from your hard disk, follow the instructions in the program that it was installed with, or use the Add or Remove Programs utility in Control Panel.

Back to the top

Speech recognition engines

This section describes speech recognition engines.

Back to the top

How to determine the selected speech recognition engine

To determine the selected speech recognition engine, follow these steps:

  1. Click Start, click Control Panel, click Sounds, Speech, and Audio Devices, and then click Speech.
  2. Click the Speech Recognition tab. The name that appears in the Language box is the active engine.


Note You may not be able to determine the language that an engine supports by the name of the engine. See the specific user's guide for detailed information about the engine. This information includes the language supported and the lexicon purpose. The lexicon purpose indicates whether the language is a general grammar or is jargon-specific to a profession such as the legal profession or the medical profession.

Back to the top

How to change speech recognition engines

To change speech recognition engines, follow these steps:

  1. Click Start, click Control Panel, click Sounds, Speech, and Audio Devices, and then click Speech.
  2. Click the Speech Recognition tab. The name that appears in the Language box is the active engine.
  3. Click the active engine or use the arrow to display a list of available engines.
  4. Click a new engine.
  5. Click OK or Apply to accept the new engine.
  6. Stop all programs that use speech, and then restart the program that you want to use to start the different speech engine.


Note You may not be able to determine the language that an engine supports by the name of the engine. See the specific user's guide for detailed information about the engine. This information includes the language supported and the lexicon purpose. The lexicon purpose indicates whether the language is a general grammar or is jargon-specific to a profession such as the legal profession or the medical profession.

Back to the top

How to change speech engine settings

Some speech engines have additional and specialized features. However, special features are not required, and not all engines support them. If the currently selected speech engine does not support custom features, the Settings button in the Language section is unavailable. If it is available, you can use these procedures.

To change speech engine settings, follow these steps:

  1. Click Start, click Control Panel, click Sounds, Speech, and Audio Devices, and then click Speech.
  2. On the Speech Recognition tab, click the Settings button in the Language section, and then follow the instructions in the dialog box or the instructions in the engine vendor's documentation. Special features depend on the engine's manufacturer and are documented separately.

Back to the top

How to train the speech recognition engine

To train the speech recognition engine, follow these steps:

  1. Click Start, click Control Panel, click Sounds, Speech, and Audio Devices, and then click Speech.
  2. Click the Speech Recognition tab, and then click the speech recognition engine that you want to use in the Language box.
  3. Click the profile that you want to use from the Recognition Profile group. Training is specific to an engine and profile. Therefore, if you train one engine or profile set, this training does not affect any other engine or profile set.
  4. Click Train Profile, and then follow the directions in the Voice Training Wizard. Not all engines support training. If your engine does not support training, the Train Profile button is unavailable.

    Note It is good practice spend at least 15 minutes training the computer. The more training you do, the higher your recognition accuracy will be.

    For additional information about how to use speech recognition in Windows XP, click the following article number to view the article in the Microsoft Knowledge Base:

    306901 How to use speech recognition in Windows XP

Back to the top

How to use speech properties

You can use either the Speech Recognition tab or the Text-to-Speech tab in the Speech properties to set up and to customize speech-enabled programs. These settings control general attributes such as input and output devices, the language used, the playback voice, and the accuracy of word recognition. Individual programs may implement speech differently. See the user's manuals for particular information. However, some attributes are shared by speech engines or by computers across different software programs. These shared features are set in the Speech properties.

For additional information about how to use text-to-speech in Windows XP, click the following article number to view the article in the Microsoft Knowledge Base:

306902 How to Use text-to-speech in Windows XP


You can select from several preferential option settings and customize the speech profile according to your requirements. However, many options are proprietary and may vary between computer systems. Therefore, some of the buttons and screens may not act the same under all circumstances. See the instructions for your computer, speech software, or device hardware.

You may also create a new profile or set a profile to fit your speaking style. It is good practice to create a new Recognizer Profile if you move offices, the noise level permanently changes, or more people are frequently present. To do this, click New on the Speech Recognition panel. With Recognizer Profiles, different users can share the same computer without interfering with other speech configurations.

For additional information about speech recognition profiles, click the following article number to view the article in the Microsoft Knowledge Base:

306899 How to use speech recognition profiles in Windows XP


With training, the speech recognizer adapts to the sound of your voice, word pronunciation, accent, speaking manner, and even new or idiomatic words. Training uses the Voice Training Wizard to implement this functionality. With ten minutes of training, you can significantly improve speech recognition. The computer continually adapts to your speech, and recognition increases over time.

For additional information about training the speech recognizer, click the following article number to view the article in the Microsoft Knowledge Base:

306901 How to use speech recognition in Windows XP


Back to the top

How to configure speech recognition

The options that are described in this section control the methods that the speech recognition engine uses to process voice inputs.

To change speech recognition options, follow these steps:

  1. Click Start, click Control Panel, click Sounds, Speech, and Audio Devices, and then click Speech.
  2. On the Speech Recognition tab, click Settings in the Recognition Profiles section.
  3. Click any one of the following option buttons:
    • Pronunciation Sensitivity: When you click this option button, you can set the level of confidence that is required for the computer to respond to your command. The High sensitivity setting indicates that you want the computer to reject any command when it is not confident of what you said. When you use this setting, the computer makes fewer recognition errors but more frequently rejects your commands. Therefore, you may have to enunciate more slowly and clearly. The Low sensitivity setting indicates that you want the computer to respond to your command when it has almost no confidence that it has correctly recognized what you said. When you use this setting, the computer recognizes fewer commands but rarely rejects a command. This option affects only command and control programs.
    • Accuracy vs. Recognition Response Time: When you click this option button, you can set the trade-off between the accuracy of speech recognition and the processing time that the computer requires to generate the recognized speech. With the Low/Fast setting, the computer performs limited processing. The recognized text appears on the screen quickly but with low accuracy. With the High/Slow setting, the computer processes more for higher accuracy but produces the dictated text more slowly. This option affects command-and-control programs and dictation programs.
    • Background Adaptation: When you click this option button, you can set whether the computer adapts to your voice and manner of speech. With this option, the computer learns what your voice sounds like (in a particular environment). This information is stored in your profile and contributes to an improvement in recognition accuracy. It is a best practice to select this option.
  4. To accept a new option, click OK. To ignore any changes that are made in the dialog box and to retain the current options, click Cancel. If you click Restore Defaults, all changes are overridden, and the options are set to a default state.


Note These features are unique to Microsoft speech recognition engines. Some speech engines have additional and specialized features. However, special features are not required, and not all speech engines support them. If the currently selected speech engine does not support custom features, options in the Recognition Profiles group are unavailable.

Back to the top

Troubleshooting

This section describes how troubleshoot issues that may occur with speech recognition.

Back to the top

How to troubleshoot speech recognition

To determine if speech recognition is working correctly, use the following test:

  1. Click Start, click Control Panel, click Sounds, Speech, and Audio Devices, and then click Speech.
  2. Speak into the microphone. When you are using a working microphone, the input volume appears in the display area. The level increases and decreases as the speaking volume changes. If no level appears, see the "Speech recognition problems" section for troubleshooting solutions.

Back to the top

Speech recognition problems

If no sound is detected after you test the system, one of the following conditions may be true:

  • The microphone is muted. Some microphones have an external control for muting. Make sure that the muting is turned off.
  • The microphone input level is set too low. To change this setting, click the Speech Recognition tab, click Configure Microphone, and then follow the directions in the Microphone wizard to set the input level.
  • The microphone may not be selected as the current input device. To change this setting, click the Speech Recognition tab, and then click Audio Input to make sure that the microphone is selected.
  • The microphone connections may not be correctly secured. Make sure that the microphone wires are intact and unbroken. The connections to the computer may be loose. Unplug the wires, and then reconnect them to make sure that they fit snugly. See the "Set Up a Microphone" section for additional connection information.
  • See the documentation for the microphone's hardware for information that is specific to the manufacturer. This information may include instructions to make sure that the sound card for the computer is correctly seated and installed and that the correct drivers are available.
  • The speech recognition engine may be corrupted. To test, switch to a different speech engine if you have one. See the "Change Speech recognition engines" section for more information. If the other engine is working correctly, reinstall the specific engine from the original source. If no engine is working correctly, reinstall the speech recognition engine from the original source or CD.

Back to the top

How to troubleshoot Text-to-Speech

To determine if text-to-speech is working correctly, use the following test:

  1. Click Start, click Control Panel, click Sounds, Speech, and Audio Devices, and then click Speech.
  2. Click the Text-to-Speech tab, and then click Preview Voice. The text in the Preview Voice box should be spoken audibly with each word highlighted in turn. If this behavior occurs, text-to-speech and the speakers are working.


If you cannot hear the preview voice or see words highlighted as they are spoken, see the "Text-to-Speech problems" section for troubleshooting solutions.

Back to the top

Text-to-Speech problems

If you cannot hear speech after you test the computer, one of the following conditions may be true:

  • The speaker volume is not turned up or is muted. Some speakers have external controls for volume and muting. Make sure that the volume is turned up sufficiently or that muting is off.
  • The speakers may not be selected as the current output device. To change this setting, click the Text-to-Speech tab, and then click Audio Output to make sure that the speakers are selected.
  • The speakers may not be connected correctly. See the speaker hardware documentation for additional information. Make sure that the sound card for the computer is correctly seated and installed and that the correct drivers are available. See the "Set Up Speakers" section for additional connection information.
  • The text-to-speech engine may be corrupted. To test, switch to another engine. See the "Change Speech recognition engines" section for more information. If another speech engine is working correctly, reinstall the specific speech engine from the original source. If no engine is working correctly, reinstall the speech recognition engine from the original source or CD.

For additional information about how to use text-to-speech in Windows XP, click the following article number to view the article in the Microsoft Knowledge Base:

306902 How to use Text-to-Speech in Windows XP


Back to the top

Additional troubleshooting

After you install speech recognition, your computer's performance may be reduced. Speech recognition requires computer memory and can affect your computer's performance. To work around this issue, use any one of the following methods:

  • Temporarily remove speech recognition as a text service while you are not using it. See the "How to remove speech recognition" section for more information about how to remove speech recognition as a text service.


Note If you remove speech recognition as a text service, it is not deleted from your computer. When you do this, speech recognition is no longer an available text service and it is prevented from loading into memory. You can add it again later.

  • Upgrade your computer processor.
  • Add additional RAM.

Back to the top

Keywords: kbenv kbhowto kbhowtomaster KB306537