Tag Archives: acapela TTS

Tips: how to use the speech dictionary of NVDA

Thanks to our user Michael Bayus for sending us this “how to” article, we are grateful to Michael and happy to share this tips with all our users.

We also invite all users to send us tips on mispronounced words and names by sending an email to contact@acapela-nvda.com or use the “report a pronunciation bug” field in the type&talk demo here on this page.

How to use the speech dictionary with the Acapela TTS for NVDA add-on

The speech dictionaries menu (found in the Preferences menu) contains dialogs that allow you to manage the way NVDA pronounces particular words or phrases. There are currently three different types of speech dictionaries. They are:

  • Default: rules in this dictionary affect all speech in NVDA.
  • Voice: rules in this dictionary affect speech for the synthesizer voice currently being used.
  • Temporary: rules in this dictionary affect all speech in NVDA, but only for the current session. These rules are temporary and will be lost if NVDA is restarted.

All dictionary dialogs contain a list of rules which will be used for processing the speech. The dialog also contains Add, Edit and Remove buttons.

To access the Speech dictionaries, press and hold the NVDA key, plus N to open the NVDA menu.  Then press the down arrow once to get to the preferences menu.  press the right arrow to access it.  Press the down arrow until you hear NVDA say “Speech dics”.

To add a new rule to the dictionary, press the Add button, and fill in the fields in the dialog box that appears and then press Ok. You will then see your new rule in the list of rules. However to make sure your rule is actually saved, make sure to press Ok to exit the dictionary dialog all together once you have finished adding/editing rules.

The rules for NVDA’s speech dictionaries allow you to change one string of characters into another.  For example, I like Ella, and she is the default voice for my screen reader.  She gets the word “peruse” wrong.  She says “parice”.  To teach her to pronounce the word correctly, I would go to the “voice” dictionary, and In the Add rule dialog, I would type the word “peruse” in the Pattern field, and I would type “per ruse” in the Replacement field. Now, Ella will say: “peruse” as she should. You may also want to type a description of the rule in the Comment field (something like: changes peruse to per ruse).

NVDA’s speech dictionaries however are much more powerful than simple word replacement. The Add rule dialog also contains a checkbox to say whether or not you want the rule to be case sensitive (meaning that NVDA should care whether the characters are uppercase or lowercase, NVDA ignores case by default).

Finally, a set of radio buttons allows you to tell NVDA whether your pattern should match anywhere, should only match if it is a complete word or should be treated as a “Regular expression”.