| US 7,477,735 B2 | ||
| System and method for enhanced stereo audio | ||
| Trygve Frederik Marton, Oslo (Norway); Ingvar Flaten Aarnes, Oslo (Norway); Bjørn Winsvold, Tranby (Norway); and Geir Ole Øverby, Oslo (Norway) | ||
| Assigned to Tandberg Telecom AS, Lysaker (Norway) | ||
| Filed on Dec. 22, 2004, as Appl. No. 11/22,293. | ||
| Claims priority of application No. 20035796 (NO), filed on Dec. 23, 2003. | ||
| Prior Publication US 2005/0157866 A1, Jul. 21, 2005 | ||
| Int. Cl. H04M 9/08 (2006.01) | ||
| U.S. Cl. 379—406.04 | 16 Claims |
| 1. An audio echo canceller adjusted to provide an echo attenuated output signal from an echo added input signal including
near-end audio in addition to far-end audio adding an echo contribution, comprising:
a mono echo compensator configured to generate a mono echo estimate by implementing a first adaptive model of acoustic echo,
converging to minimize a mono part of the echo contribution in the echo added input signal;
a first subtracting unit configured to subtract the mono echo estimate from the echo added input signal to generate a first
output signal;
one or more multi component audio detectors configured to detect whether multi component audio is present in the far-end audio;
a stereo echo compensator configured to generate a stereo echo estimate by implementing a second adaptive model of acoustic
echo; and
a second subtracting unit configured to subtract the stereo echo estimate from the first output signal to generate a second
output signal,
wherein the mono echo compensator is configured to suspend converging to minimize the mono part of the echo contribution in
the first adaptive model when said one or more multi component audio detectors detects multi component audio in the far-end
audio, and
the stereo echo compensator is configured to suspend converging in the second adaptive model when said one or more multi component
audio detectors does not detect multi component audio in the far-end audio.
|