Telephonic communications – Echo cancellation or suppression
Reexamination Certificate
1998-05-26
2002-08-27
Isen, Forester W. (Department: 2644)
Telephonic communications
Echo cancellation or suppression
C379S406040
Reexamination Certificate
active
06442272
ABSTRACT:
FIELD OF THE INVENTION
The present invention relates to an audio communication terminal for conducting a hands-free telephonic conference. In particular the present invention relates to such a terminal configured to permit conference participants at the location of the terminal to be physically separated by an acoustic distance greater than the tail length of an echo canceller provided in the terminal.
BACKGROUND
Audio conferencing systems are popular tools for conducting telephonic conversations among multiple participants. Early generations of speakerphones included a microphone for picking up near-end speech, a loudspeaker for reproducing far-end speech, and a voice switch or gain switch. The function of the voice switch is to disable the microphone during reception of far-end speech, in order to prevent the microphone from picking up the loudspeaker and re-transmitting the far-end speech as echo. Such an arrangement was unsatisfactory to many users because the switch would often cause undesirable clipping of speech signals and prevented simultaneous, or full-duplex, conversation. The relatively recent availability of inexpensive digital signal processing technology has rendered it economically feasible to provide full-duplex conferencing terminals by employing an adaptive echo canceller instead of a voice switch. The echo canceller estimates the echo signal that will result from a received far-end signal and subtracts the estimated echo signal from the microphone signal in order to cancel the echo component of the microphone signal prior to transmission to the far-end. Such full-duplex terminals have been developed for compatible use with traditional telephone networks and for conducting voice conversations via digital data transmission systems.
An important design parameter of an echo canceller is the period of time in which the echo canceller is operative to detect that a microphone signal includes an echo component. In order to detect echo, and to develop a model of the impulse response of the echo path, the echo canceller maintains a register containing stored samples of the received far-end communication signal. The size of this register establishes a time period, known as the “tail length” of the canceller, over which the echo canceller will develop a model of the echo path and apply echo cancellation.
If an echo canceller is deployed in a room wherein a microphone is positioned at an acoustic distance from the loudspeaker that exceeds the tail length of the canceller, then the canceller will fail to recognize signals from that microphone as comprising an echo component. Consequently, the echo path model may become distorted as the echo canceller attempts to adapt the model on the basis of signals that are not correctly identified as echo signals. The potential distortive effect of deploying the conferencing system in a large room will be exacerbated by local amplification of the near-end speech signals, which is desirable in large rooms in order to ensure that the near-end conference participants are able to clearly able to hear one another. Hence, the combined factors of increased room size and near-end speech amplification can combine to cause poor echo canceller performance. Although an echo canceller having a long tail length could be provided, at added expense, it would be desirable to provide an audio conferencing system that could readily configured to function in large and small rooms, and which would provide for local sound amplification without significantly detracting from echo canceller stability.
SUMMARY OF THE INVENTION
In accordance with the present invention, an audio conferencing system includes a first loudspeaker for reproducing received far-end audio signals and a second loudspeaker for reproducing amplified near-end audio signals. Near-end audio signals are provided to the audio conferencing system by at least two microphones, to permit conference participation by a plurality of near-end participants. The first microphone, or a first group of microphones, is positioned relative to the first loudspeaker such that the acoustic time delay for signals transmitted from the first loudspeaker to the first microphone is less than the tail length of the echo canceller. A second microphone, or a second group of microphones, is positioned relative to the first loudspeaker such that the acoustic time delay for signals from the first loudspeaker to the second microphone is longer than the tail length of the echo canceller. Signals from the first and second microphones are combined by an audio mixer to provide a near-end input signal to an echo canceller, and as an input signal to the second loudspeaker. The second microphone signal is selectively gated to the mixer, such that second microphone signal is provided to the mixer only during periods of near-end speech. Hence, adaptation of the echo canceller is determined by the signal from the first microphone, which is positioned within a distance from the first loudspeaker defined by the tail length of the echo canceller.
REFERENCES:
patent: 3992586 (1976-11-01), Jaffe
patent: 4636586 (1987-01-01), Schiff
patent: 5027393 (1991-06-01), Yamamura et al.
patent: 5297210 (1994-03-01), Julstrom
patent: 5404397 (1995-04-01), Janse et al.
patent: 5668794 (1997-09-01), McCaslin et al.
patent: 5848146 (1998-12-01), Slattery
Berryhill John B.
Dann Dorfman Herrell & Skillman
Harold Jefferey
Isen Forester W.
Tellabs Inc.
LandOfFree
Voice conferencing system having local sound amplification does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Voice conferencing system having local sound amplification, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Voice conferencing system having local sound amplification will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-2911931