Microphone Array Basics

When you install the Microsoft Kinect SDK, the components required for speech recognition are automatically chain installed. The Kinect microphone array works on top of preexisting code libraries that have been around since Windows Vista. These preexisting components include the Voice Capture DirectX Media Object (DMO) and the Speech Recognition API (SAPI).

In C#, the Kinect SDK provides a wrapper that extends the Voice Capture DMO. The Voice Capture DMO is intended to provide an API for working with microphone arrays to provide functionality such as acoustic echo cancellation (AEC), automatic gain control (AGC), and noise suppression. This functionality can be found in the audio classes of the SDK. The Kinect SDK audio wrapper simplifies working with the Voice Capture DMO as well as optimizing DMO performance with the Kinect sensor.