This week Microsoft announced the latest updates to its Cognitive Services Speech SDK, now in version 1.3. Included in the update is the selection of the input microphone by using the AudioConfig class, beta support for Unity and new sample code.
New to the SDK is the ability for it to use the AudioConfig class in order to support a selection of the input microphone. For developers this means they can now stream audio data to the service without being locked into the default microphone.