azure_speech_services 1.0.5

A Flutter package for real-time Speech-to-Text, Text-To-Speech transcription, and Translation using Microsoft Azure Cognitive Services. Supports Mobile, Desktop, and Web platforms

1.0.5

• Reduce Text-to-Speech latency by reusing a single http.Client in AzureTextToSpeech instead of opening a fresh connection (TCP+TLS handshake) on every synthesize()/listVoices() call; add AzureTextToSpeech.dispose() to close it

1.0.4

• Fix PlatformException on dispose when the recorder was never started (e.g. when an external audio stream is supplied and the native recorder is never created)
• Fix Bad state: Cannot emit new states after calling close by stopping the STT session (closing the WebSocket and cancelling its subscription) before closing the transcription cubit on dispose, and by making TranscriptionCubit ignore emits that race past close() (websocket messages, the clear-text timer, or the explicit stop-listening call)

1.0.3

• Fix Text to speech sample
• Fix Web STT/Translation silence caused by record_web's broken resampler, by downsampling in software

1.0.2

• Fix building and running iOS and Android example

1.0.1

• Downgraded record library to 6.2.0 as the 7.x has incompatibilities with azure web services

1.0.0

• Initial Version of the library