Dictation Bridge Public Beta Now Available

The dictation bridge team is proud to announce the release of our first public beta. Over the past several months our team of excellent engineers, private beta testers, and technical writers, (the people responsible for creating documentation to guide you), have been working hard to bring you the world’s first NVDA add-on to allow screenreader users access to leading edge speach recognition software.

Our team is still working hard on bringing you access to JAWS and Window-eyes access and will be rolling that out shortly. In the meantime, please feel free to download the beta and try it with your favorite speech recognition package. This beta works with Microsoft speech recognition and Dragon NaturallySpeaking. Please be sure to let us know what you think and stay tuned for our next public beta and final products.

Please see below for your own copy of the beta and our product release notes. Thank you once again for your patience and support in our crowdfunding campaign and development of this free software.

Release notes for dictationBridge beta 1

Currently working features

  • Echo back of dictated text in Dragon and Windows Speech Recognition also known as WSR.
  • Speech only support of the WSR correction box.
  • Support to control NVDA from Dragon and WSR. At this time, only Dragon commands have been written.
  • A verbal notification of the microphone status when using Dragon. WSR has this feature built-in therefore no support needs to be created.
  • Command NVDA by voice from Dragon.

Feature Descriptions

Echo back of dictated text in Dragon and Windows Speech Recognition also known as WSR

When you dictate into any application with either Dragon or WSR running, dictationBridge will speak the text that has been recognized by either engine if the NVDA screen reader is active. There are no additional settings that have to be done. The functionality works out of the box. If you want to disable echo back, then disable the dictationBridge addon.

Speech only support of the WSR correction box

When you need to correct misrecognized text in WSR, do the following.
Position the cursor on the word you want to correct. Once your cursor is positioned, say “spell”. A floating window will come up on the screen. This window is invisible to NVDA and other screen readers. NVDA will read out the choices. You then say “choose” followed by the number of that choice; for example, if you wanted to select the fourth choice, you will say “choose 4”.

Support to control NVDA from Dragon and WSR

It is possible to execute any script in NVDA. This is done by sending the script name to dictationBridge which in turn sends it to NVDA. So, you can say “speak focus” to have NVDA say the contents of the keyboard focus. There are over 70 commands that have been added for Dragon. Commands for WSR have not yet been written. These commands, in the background execute NVDA scripts that have keyboard shortcuts assigned to them. These scripts are present in NVDA 2016.3. These script names should be the same with earlier versions of NVDA. To get this feature to work, you need to add the dictationBridge folder to the system path. Instructions for doing this will be included in the installation section.

A verbal notification of the microphone status when using Dragon

It is important to know if the microphone is active or not when using speech-recognition. This avoids the transcribing of unwanted text into your document. DictationBridge will now speak the status of the microphone as it is changed when using Dragon. WSR already indicates this status and there are distinct sounds mapped by default to microphone status changes.

Command NVDA by voice from Dragon

It is possible to control NVDA by speaking voice commands from dragon. The commands should have been imported into your user profile. If this has not happened, please see the section on importing them. These commands map to the majority of NVDA features giving you complete hands free control of this program. You can then use native Dragon commands as well to control other programs such as the Windows desktop, note pad and Microsoft Word.

Installing dictationBridge

This version of dictationBridge like the ones before is a NVDA addon. To install it, do the following.

  • Launch NVDA. Ensure that you are running an installed copy of NVDA. This addon has not been tested in a portable version of NVDA.
  • Download the Dictation Bridge addon
  • Once the addon has been downloaded, highlight the file and press enter on it.
  • Follow the prompts and allow NVDA to restart.
    The dictationBridge addon is now loaded.

Importing commands automatically

Once dictationBridge is installed, you will have an item in the tools menu called “Add Commands.” Activate this menu item and follow the prompts. It is crucial that you have already exported your commands that you want to import. These commands should be exported as XML files for this feature to work. If you have exported them as dat files, use the manual import feature from the Dragon command browser. You can do the same for XML commands.

Importing commands manually

  • Launch Dragon and ensure that your user profile is loaded.
  • Get to the Dragon bar and then to the tools menu.
  • Launch the command browser.
  • Hit alt+m to get to the Mode menu and arrow down to “Manage” and activate that option.
  • Hit ctrl+i, set the file type to XML and navigate to where you have saved the dictationBridgeCommands.xml file and highlight it.
  • Tab to the button labeled “Open” and activate it.
  • In Dragon version 15, you will be asked if you want to validate the commands. Agree to do this to ensure that the XML file is valid.
  • Tab through the resulting dialogue until you reach a button called “Import” and activate it.
    The commands should be imported.

Note:

After this procedure, it is crucial to have the following dynamic link libraries (DLLs) in the system path.
* DictationBridgeClient32.dll
* DictationBridgeClient64.dll

Making Dragon easier to use

There are a few options you need to configure in Dragon so that it works optimally when using a screen reader. You should make these changes immediately after creating your user profile.
Go to the Tools menu in the dragon bar and then to the Options menu item and activate it. This is a large multi-page dialogue and we need to change a few settings here. In Dragon 15, you will be unable to tab to the names of the property sheets in the dialogue. Use ctrl+tab to move forward to the next property sheet and ctrl+shift+tab to move backward to the previous property sheet.

On the Corrections tab, enable the options “Correct commands bring up Spelling Window” and “Spell commands bring up Spelling Window”

Navigate to the appearance tab and set the dragon bar to tray only mode.

Navigate to the miscellaneous tab and ensure that the option titled “Use menus that are compatible with screen readers” is checked.

Press “ok” to close the dialogue.

Known issues

The Dragon vocabulary editor is not usable as the list of words is not read by NVDA. We are waiting for the NVDA team to fix certain items before we can support this feature.

The training dialogue in Dragon is not fully supported. We are working on supporting this dialogue and hope to bring out this feature soon.

The word “New paragraph” may be heard in various situations. Please report when this occurs as we are actively tracking this issue.

Certain features such as the invoking of the WSR corrections dialogue may not work on versions of Windows that are set to a language other than English. This is because we have had to use English specific names to interact with portions of Dragon. We are working on enabling translations and hope to allow translators for your language to fix this problem soon.

Submitting bug reports

You should submit all bugs using the dictationBridge issue tracker. There are 3 components into which dictationBridge is split:
the NVDA add-on, the jaws scripts, and the dictationBridge core.

In order to file a bug report, you will first need to ensure that you have a Github account. When you navigate to one of the issue trackers above, Github will ask you to either sign in or create an account. If you have not done so already, create a Github account. Once this is done, you can sign in and file your bug report on the appropriate tracker.

One question that may arise is how to know where a bug is.
There is no accurate way to determine this without testing and seeing the code but here are a few general rules.
1. If you are using jaws and have a problem, you will file bugs in the jaws section.
2. If you are using NVDA, you will file bugs in the NVDA section.
3. If you are a developer and fancy cracking open a debugger, join the party and file bugs in the core.

Getting support and meeting other users

Please join the Discussion list for dictationBridge called dictation-discuss and talk to us.
list address:
dictation-discuss@3mousetech.com
subscription address:
dictation-discuss-subscribe@3mousetech.com
Note:
Send a message with the word “subscribe” without the quotes in the subject line to the subscription address.