User Manual

Acoustic Echo Canceller White Paper
Acoustic Echo Canceller
White Paper
June 2011
1. Introduction
Acoustic Echo
Communication technology has advanced to the point
where teleconferences that use communication lines to
connect remote conference rooms are taking place every
day.
Fig.1 Teleconference
In a teleconference, as in
Fig.1, sound is exchanged in both
directions through communication lines. When the voice of
the person speaking in room A is transferred to room B
through communication lines and reproduced by a
loudspeaker, that voice can be heard in room B. However
the microphone in room B also picks up the voice of the
person speaking in room A and sends that voice back to
room A. Because of this, in room A, the voice of the person
speaking is reproduced by the loudspeaker as an echo. This
is referred to as acoustic echo. In addition to hindering
conversation, it can cause feedback. To have a smooth
conference, it is necessary to reliably suppress acoustic
echo.
Acoustic Echo Canceller
One way to suppress acoustic echo is, for example, to
reduce the signal level of the microphone in room B when
a person is speaking in room A. This is referred to as echo
suppression. However, when echo suppression is used and
people in room A and room B talk at the same time (double
talk), their voices get cut off. This makes it difficult for
people to hear what was said and prevents smooth
conversation. To solve this problem, in 1966, M. M.
Sondhi
[1][2] proposed an acoustic echo canceller with an
ADF (adaptive digital filter). However, to actually create
an acoustic echo canceller with an ADF, one would need
highly advanced signal processing technology and a signal
1

Summary of content (4 pages)