Handsfree Telephony

Why is acoustic processing necessary?

During an automotive handsfree call, the voice of the far-side speaker (the caller at the remote end of the conversation) is played out over the vehicle loudspeakers and is picked up on the near-side microphone along with the driver's voice. If they are not attenuated, the far-side speaker's words as well as background noise are audible to the far-side speaker as an echo repeating his or her voice and the noise. Acoustic echo cancellation is therefore required to maintain acceptable sound quality and intelligibility during a handsfree call.

The automobile can be an acoustically hostile environment for capturing speech signals. Road, traffic, and engine noises tend to mask speech signals, and requirements for microphone positioning can leave the voice almost hidden by noise. The result can be a telephone conversation that is tiring to listen to or difficult to understand. To further complicate matters, the noise in a car is dynamic, changing its loudness and frequency content due to many factors, including road surface, vehicle speed, open or closed windows, and so on. Acoustic noise reduction improves the subjective quality and the objective intelligibility of speech signals by removing unwanted noise and distortion and enhancing the speech.

In automotive environments, variation in loudspeaker and microphone placement can create a very complicated acoustic path. For instance:

• The echo may be several times louder than the near-side speaker's speech.
• The echo may traverse moving bodies before arriving at the microphone, making it difficult to prediction and cancel.

Acoustic echoes can also be a problem in more conventional handsfree speaker phones and handheld phones (including mobile phones), where the loudspeaker and microphone are in the same physical housing.