Audio sessions (iOS)

Dragon Medical SpeechKit needs an initialized audio session to record audio data properly.

If your app doesn't initialize and configure an audio session because it doesn't use audio or uses the default settings as standard, Dragon Medical SpeechKit initializes and configures the audio session on its own when the speech recognition session is opened.

If your app does initialize an audio session, it's responsible for stopping recording in the speech recognition session (by calling stopRecording) if there are audio route changes and interruptions.

If your app changes the audio session category, after opening the speech recognition session, it must use an audio category that supports recording.

For more information on audio session programming, see the iOS Audio Session Programming Guide (http://developer.apple.com/library/ios/#documentation/Audio/Conceptual/AudioSessionProgrammingGuide).