SpeechSynthesizer Class

A speech synthesizer.

Inheritance
builtins.object
SpeechSynthesizer

Constructor

SpeechSynthesizer(speech_config: ~azure.cognitiveservices.speech.SpeechConfig, audio_config: ~azure.cognitiveservices.speech.audio.AudioOutputConfig | None = <azure.cognitiveservices.speech.audio.AudioOutputConfig object>, auto_detect_source_language_config: ~azure.cognitiveservices.speech.languageconfig.AutoDetectSourceLanguageConfig = None)

Parameters

Name Description
speech_config
Required

The config for the speech synthesizer

audio_config

The config for the audio output. This parameter is optional. If it is not provided, the default speaker device will be used for audio output. If it is None, the output audio will be dropped. None can be used for scenarios like performance test.

auto_detect_source_language_config

The auto detection source language config

Default value: None

Methods

get_voices_async

Get the available voices, asynchronously.

speak

Performs synthesis on a speech synthesis request in a blocking (synchronous) mode.

This method is in preview and may be subject to change in future versions. Added in version 1.37.0.

speak_async

Performs synthesis on a speech synthesis request in a non-blocking (asynchronous) mode.

This method is in preview and may be subject to change in future versions. Added in version 1.37.0.

speak_ssml

Performs synthesis on ssml in a blocking (synchronous) mode.

speak_ssml_async

Performs synthesis on ssml in a non-blocking (asynchronous) mode.

speak_text

Performs synthesis on plain text in a blocking (synchronous) mode.

speak_text_async

Performs synthesis on plain text in a non-blocking (asynchronous) mode.

start_speaking

Starts synthesis on a speech synthesis request in a blocking (synchronous) mode.

This method is in preview and may be subject to change in future versions. Added in version 1.37.0.

start_speaking_ssml

Starts synthesis on ssml in a blocking (synchronous) mode.

start_speaking_ssml_async

Starts synthesis on ssml in a non-blocking (asynchronous) mode.

start_speaking_text

Starts synthesis on plain text in a blocking (synchronous) mode.

start_speaking_text_async

Starts synthesis on plain text in a non-blocking (asynchronous) mode.

stop_speaking

Synchronously terminates ongoing synthesis operation. This method will stop playback and clear unread data in PullAudioOutputStream.

stop_speaking_async

Asynchronously terminates ongoing synthesis operation. This method will stop playback and clear unread data in PullAudioOutputStream.

get_voices_async

Get the available voices, asynchronously.

get_voices_async(locale: str = '') -> ResultFuture

Parameters

Name Description
locale
Required

Specify the locale of voices, in BCP-47 format; or leave it empty to get all available voices.

Returns

Type Description

A task representing the asynchronous operation that gets the voices.

speak

Performs synthesis on a speech synthesis request in a blocking (synchronous) mode.

This method is in preview and may be subject to change in future versions. Added in version 1.37.0.

speak(request: SpeechSynthesisRequest) -> SpeechSynthesisResult

Parameters

Name Description
request
Required

The speech synthesis request.

Returns

Type Description

A SpeechSynthesisResult.

speak_async

Performs synthesis on a speech synthesis request in a non-blocking (asynchronous) mode.

This method is in preview and may be subject to change in future versions. Added in version 1.37.0.

speak_async(request: SpeechSynthesisRequest) -> ResultFuture

Parameters

Name Description
request
Required

The speech synthesis request.

Returns

Type Description

A future with SpeechSynthesisResult.

speak_ssml

Performs synthesis on ssml in a blocking (synchronous) mode.

speak_ssml(ssml: str) -> SpeechSynthesisResult

Parameters

Name Description
ssml
Required

Returns

Type Description

A SpeechSynthesisResult.

speak_ssml_async

Performs synthesis on ssml in a non-blocking (asynchronous) mode.

speak_ssml_async(ssml: str) -> ResultFuture

Parameters

Name Description
ssml
Required

Returns

Type Description

A future with SpeechSynthesisResult.

speak_text

Performs synthesis on plain text in a blocking (synchronous) mode.

speak_text(text: str) -> SpeechSynthesisResult

Parameters

Name Description
text
Required

Returns

Type Description

A SpeechSynthesisResult.

speak_text_async

Performs synthesis on plain text in a non-blocking (asynchronous) mode.

speak_text_async(text: str) -> ResultFuture

Parameters

Name Description
text
Required

Returns

Type Description

A future with SpeechSynthesisResult.

start_speaking

Starts synthesis on a speech synthesis request in a blocking (synchronous) mode.

This method is in preview and may be subject to change in future versions. Added in version 1.37.0.

start_speaking(request: SpeechSynthesisRequest) -> SpeechSynthesisResult

Parameters

Name Description
request
Required

The speech synthesis request.

Returns

Type Description

A SpeechSynthesisResult.

start_speaking_ssml

Starts synthesis on ssml in a blocking (synchronous) mode.

start_speaking_ssml(ssml: str) -> SpeechSynthesisResult

Parameters

Name Description
ssml
Required

Returns

Type Description

A SpeechSynthesisResult.

start_speaking_ssml_async

Starts synthesis on ssml in a non-blocking (asynchronous) mode.

start_speaking_ssml_async(ssml: str) -> ResultFuture

Parameters

Name Description
ssml
Required

Returns

Type Description

A future with SpeechSynthesisResult.

start_speaking_text

Starts synthesis on plain text in a blocking (synchronous) mode.

start_speaking_text(text: str) -> SpeechSynthesisResult

Parameters

Name Description
text
Required

Returns

Type Description

A SpeechSynthesisResult.

start_speaking_text_async

Starts synthesis on plain text in a non-blocking (asynchronous) mode.

start_speaking_text_async(text: str) -> ResultFuture

Parameters

Name Description
text
Required

Returns

Type Description

A future with SpeechSynthesisResult.

stop_speaking

Synchronously terminates ongoing synthesis operation. This method will stop playback and clear unread data in PullAudioOutputStream.

stop_speaking() -> None

stop_speaking_async

Asynchronously terminates ongoing synthesis operation. This method will stop playback and clear unread data in PullAudioOutputStream.

stop_speaking_async() -> ResultFuture

Returns

Type Description

A future that is fulfilled once synthesis has been stopped.

Attributes

authorization_token

The authorization token that will be used for connecting to the service.

Note

The caller needs to ensure that the authorization token is valid. Before the

authorization token expires, the caller needs to refresh it by calling this setter with a

new valid token. Otherwise, the synthesizer will encounter errors while speech synthesis.

bookmark_reached

Signal for events indicating a bookmark is reached.

Callbacks connected to this signal are called with a SpeechSynthesisBookmarkEventArgs instance as the single argument.

Note

Added in version 1.16.0.

properties

A collection of properties and their values defined for this SpeechSynthesizer.

synthesis_canceled

Signal for events indicating synthesis has been canceled.

Callbacks connected to this signal are called with a SpeechSynthesisEventArgs instance as the single argument.

synthesis_completed

Signal for events indicating synthesis has completed.

Callbacks connected to this signal are called with a SpeechSynthesisEventArgs instance as the single argument.

synthesis_started

Signal for events indicating synthesis has started.

Callbacks connected to this signal are called with a SpeechSynthesisEventArgs instance as the single argument.

synthesis_word_boundary

Signal for events indicating a word boundary.

Callbacks connected to this signal are called with a SpeechSynthesisWordBoundaryEventArgs instance as the single argument.

synthesizing

Signal for events indicating synthesis is ongoing.

Callbacks connected to this signal are called with a SpeechSynthesisEventArgs instance as the single argument.

viseme_received

Signal for events indicating a viseme is received.

Callbacks connected to this signal are called with a SpeechSynthesisVisemeEventArgs instance as the single argument.

Note

Added in version 1.16.0.