sound » speak(text)
Description
This command generates audio notification, just like other sound commands. However this one is different in
that it generates audio rendition of specified text via text-to-speech technology. By specifying text, this command
converts the audio equivalent, which can be purposed as an effective form of audio notification.
Setup
Current implementation limits text to 500 characters since this has direct impact of the underlying AWS usage
(and thus billing). Nexial utilizes
AWS Polly
as the underlying text-to-speech engine. To enable AWS Polly, the following configuration details must be provided:
| setting | default | explanation |
|---|---|---|
nexial.tts.region |
us-east-1 |
(Optional) Specify the AWS region to use for the text-to-speech service. |
nexial.tts.accessKey |
[REQUIRED] Specify the AWS access key to use for the text-to-speech service. | |
nexial.tts.secretKey |
[REQUIRED] Specify the AWS secret key to use for the text-to-speech service. | |
nexial.tts.sampleRate |
22050 | (Optional) The audio frequency specified in Hz. The valid values are 8000, 16000 and 22050. |
nexial.tts.language |
en-GB | (Optional) The language code of text to which to translate. Here are languages currently supported: |
The above settings can be configured via Nexial Setup,
via command line and via project.properties. For example,
- via command line:
set JAVA_OPT=-Dnexial.tts.accessKey=... -Dnexial.tts.secretKey=... nexial.cmd -script ... ... - via
project.properties:nexial.tts.accessKey=... nexial.tts.secretKey=... nexial.tts.language=en-US
Text Pattern
Nexial provides some “understanding” of the content of text. Here are some ways from which one can adjust the audio
rendering of the specified text in order to create a more natural appeal:
- pause - use
...or…in yourtextto create slight delay. For example,Well... I'm not sure what you meanwould be rendered as “Well (pause) I’m not sure what you mean”. - emphasis - use UPPERCASE in your
textto create vocal emphasis. For example,Found another error AGAINwould render: “Found another error (emphasis)AGAIN”.
Parameters
- text - the text to transition to speech.