Communication system with active noise cancellation

• US 5,033,082 A
• Filed: 07/31/1989
• Issued: 07/16/1991
• Est. Priority Date: 07/31/1989
• Status: Expired due to Term

• Current Assignee: Nelson Industries Incorporated (Cummins Incorporated)
• Original Assignee: Nelson Industries Incorporated (Cummins Incorporated)
• Inventors: Allie, Mark C., Eriksson, Larry J.
• Primary Examiner(s): Dwyer, James L.
• Assistant Examiner(s): Shehata, M.
• Application Number:
  US07/388,014
• Time in Patent Office: 715 Days
• Field of Search: 381/71, 381/94, 379/388, 379/392, 379/410, 379/411
• US Class Current: 379/406.08
• CPC Class Codes: G01R 33/3854 means for active and/or passive vibration damping or acoustical noise suppression in gradient magnet coil systems
• CPC Class Codes: G10K 11/17854 the filter being an adaptive filter
• CPC Class Codes: G10K 11/17857 Geometric disposition, e.g. placement of microphones
• CPC Class Codes: G10K 11/17881 the reference signal being an acoustic signal, e.g. recorded with a microphone
• CPC Class Codes: G10K 11/17885 additionally using a desired external signal, e.g. pass-through audio such as music or speech
• CPC Class Codes: G10K 2210/108 Communication systems, e.g. where useful sound is kept and noise is cancelled
• CPC Class Codes: G10K 2210/1082 Microphones, e.g. systems using "virtual" microphones
• CPC Class Codes: G10K 2210/1161 NMR or MRI
• CPC Class Codes: G10K 2210/1282 Automobiles
• CPC Class Codes: G10K 2210/3012 Algorithms
• CPC Class Codes: G10K 2210/3046 Multiple acoustic inputs, multiple acoustic outputs
• CPC Class Codes: G10K 2210/3224 Passive absorbers
• CPC Class Codes: G10K 2210/511 Narrow band, e.g. implementations for single frequency cancellation
• CPC Class Codes: G10K 2210/512 Wide band, e.g. non-recurring signals

Abstract

An active acoustic attenuation system (10) is provided with various adaptive filter models (40, 48, 56, 70, 84, 100) enabling communication between persons (26, 30) in spaced zones (12, 16) by selectively cancelling undesired noise and undesired speech, all on an on-line basis without dedicated off-line pretraining and also for both broadband and narrowband noise.

Claims

1. An active acoustic attenuation system comprising:

a zone subject to noise from a noise source;

a microphone sensing noise from said noise source;

a speaking location in said zone such that a person at said speaking location is subject to noise from said noise source;

a speaker introducing sound into said zone at said speaking location;

an error microphone sensing noise and speech at said speaking location;

an adaptive filter model adaptively modeling an acoustic path from said noise microphone to said speaking location, said adaptive filter model having a model input from said noise microphone, an error input from said error microphone, and outputting a correction signal to said speaker to introduce cancelling sound at said speaking location to cancel noise from said noise source at said speaking location, such that the output of said error microphone carries a speech signal from said person but not a noise signal from said noise source;

a second zone spaced from said first mentioned zone;

a second error microphone sensing speech by a second person at said second zone, the output of said second microphone being supplied to said speaker such that said first mentioned person can hear the speech of said second person.

2. The invention according to claim 1 comprising a second speaker introducing sound into said second zone, and wherein the output cf said first mentioned error microphone is supplied to said second speaker such that said second person at said second speaker can hear the speech of said first person, which speech is free of noise from said noise source.

3. The invention according to claim 2 comprising a second adaptive filter model adaptively cancelling speech from said second person in the output of said first error microphone, said second adaptive filter model having a model input from said second error microphone, an output correction signal summed with the output of said first error microphone to provide a sum, and an error input from said sum, said sum also being supplied to said second speaker such that said second person can hear the speech of said first person with the speech of said second person removed.

4. The invention according to claim 3 wherein said output correction signal of said second adaptive filter model is subtractively summed with the output of said first error microphone.

5. An active acoustic attenuation system comprising:

a zone subject to noise from a noise source;

a microphone sensing noise from said noise source;

a speaking location in said zone such that a person at said speaking location is subject to noise from said noise source;

a speaker introducing sound into said zone at said speaking location;

an error microphone sensing noise and speech at said speaking location;

a first adaptive filter model adaptively modeling an acoustic path from said noise microphone to said speaking location, said first adaptive filter model having a model input from said noise microphone, an error input from said error microphone, and outputting a correction signal to said speaker to introduce cancelling sound at said speaking location to cancel noise from said noise source at said speaking location;

a second adaptive filter model adaptively cancelling noise from said noise source in the output of said error microphone, said second adaptive filter model having a model input from said noise microphone, an output correction signal summed with the output of said error microphone to provide a sum, and an error input from said sum.

6. The invention according to claim 5 comprising:

a second zone spaced from said first mentioned zone;

a second error microphone sensing speech by a second person at said second zone, the output of said second error microphone being supplied to said speaker such that said first mentioned person can hear the speech of said second person.

7. The invention according to claim 6 comprising a second speaker introducing sound into said second zone, said sum also being supplied to said second speaker such that said second person at said second zone can hear the speech of said first person.

8. An active acoustic attenuation system comprising:

a first zone subject to noise from a noise source;

a second zone spaced from said first zone;

a microphone sensing noise from said noise source;

a speaking location in said first zone such that a first person at said speaking location is subject to noise from said noise source;

a first speaker introducing sound into said first zone at said speaking location;

a second speaker introducing sound into said second zone;

a first error microphone sensing noise and speech at said speaking location;

a second error microphone sensing speech of a second person at said second zone, the output of said second error microphone being supplied to said first speaker such that said first person can hear the speech of said second person;

a first adaptive filter model adaptively modeling an acoustic path from said noise microphone to said speaking location, said first adaptive filter model having a model input from said noise microphone, an error input from said first error microphone, and outputting a correction signal to said first speaker to introduce cancelling sound at said speaking location to cancel noise from said noise source at said speaking location;

a second adaptive filter model adaptively cancelling noise from said noise source in the output of said first error microphone, said second adaptive filter model having a model input from said noise microphone, an output correction signal summed with the output of said first error microphone to provide a first sum, and an error input from said first sum;

a third adaptive filter model adaptively cancelling speech from said second person in the output of said first error microphone, said third adaptive filter model having a model input from said second error microphone, an output correction signal summed with said first sum to provide a second sum, and an error input from said second sum, said second sum also being supplied to said second speaker such that said second person can hear the speech of said first person with the speech of said second person removed.

9. The invention according to claim 8 comprising a fourth adaptive filter model adaptively cancelling speech from said first person in the output of said second error microphone, said fourth adaptive filter model having a model input from said first error microphone, an output correction signal summed with the output of said second error microphone to provide a third sum, and an error input from said third sum, said third sum also being supplied to said first speaker such that said first person can hear the speech of said second person with the speech of said first person removed.

10. The invention according to claim 9 wherein said third sum is summed with the output correction signal of said first adaptive filter model to provide a fourth sum, said fourth sum being supplied to said first speaker.

11. The invention according to claim 10 wherein:

said third sum is additively summed with said output correction signal of said first adaptive filter model to provide said fourth sum;

said output correction signal of said second adaptive filter model is subtractively summed with said output of said first error microphone to provide said first sum;

said output correction signal of said third adaptive filter model is subtractively summed with said first sum to provide said second sum;

said output correction signal of said fourth adaptive filter model is subtractively summed with said output of said second error microphone to provide said third sum.

12. An active acoustic attenuation system comprising:

a first zone subject to noise from a noise source;

a second zone spaced from said first zone;

a microphone sensing noise from said noise source;

a speaking location in said first zone such that a first person at said speaking location is subject to noise from said noise source;

a first speaker introducing sound into said first zone at said speaking location;

a second speaker introducing sound into said second zone;

a first error microphone sensing noise and speech at said speaking location;

a second error microphone sensing speech of a second person at said second zone, the output of said second error microphone being supplied to said first speaker such that said first person can hear the speech of said second person;

a first adaptive filter model adaptively modeling an acoustic path from said noise microphone to said speaking location, said first adaptive filter model having a model input from said noise microphone, an error input from said first error microphone, and outputting a correction signal to said first speaker to introduce cancelling sound at said speaking location to cancel noise from said noise source at said speaking location;

a second adaptive filter model adaptively cancelling noise from said noise source in the output of said first error microphone, said second adaptive filter model having a model input from said noise microphone, an output correction signal summed with the output of said first error microphone to provide a first sum, and an error input from said first sum, said first sum also being supplied to said second speaker such that said second person can hear the speech of said first person;

a third adaptive filter model adaptively cancelling speech from said first person in the output of said second error microphone, said third adaptive filter model having a model input from said first error microphone, an output correction signal summed with the output of said second error microphone to provide a second sum, and an error input from said second sum, said second sum also being supplied to said first speaker such that said first person can hear the speech of said second person with the speech of said first person removed.

13. An active acoustic attenuation system comprising:

a first zone subject to noise from a first noise source;

a second zone spaced from said first zone and subject to noise from a second noise source;

a first noise microphone sensing noise from said first noise source;

a second noise microphone sensing noise from said second noise source;

a first speaking location in said first zone such that a first person at said first speaking location is subject to said noise from said first noise source;

a second speaking location in said second zone such that a second person at said second speaking location is subject to said noise from said second noise source;

a first speaker introducing sound into said first zone at said first speaking location;

a second speaker introducing sound into said second zone at said second speaking location;

a first error microphone sensing noise and speech at said first speaking location;

a second error microphone sensing noise and speech at said second speaking location;

a first adaptive filter model adaptively modeling a first acoustic path from said first noise microphone to said first speaking location, said first adaptive filter model having a model input from said first noise microphone, an error input from said first error microphone, and outputting a correction signal to said first speaker to introduce cancelling sound at said first speaking location to cancel noise from said first noise source at said first speaking location;

a second adaptive filter model adaptively cancelling noise from said second noise source in the output of said second error microphone, said second adaptive filter model having a model input from said second noise microphone, an output correction signal summed with the output of said second error microphone to provide a first sum, and an error input from said first sum;

said first sum also being supplied to said first speaker such that said first person can hear the speech of said second person.

14. The invention according to claim 13 wherein said first sum is summed with the correction signal output of said first adaptive filter model to provide a second sum which is supplied to said first speaker, said second sum being substantially free of noise from said second noise source as cancelled by said second adaptive filter model, such that said second sum to said first speaker is substantially free of noise from said second noise source but does contain speech from said second person, such that said first speaker cancels noise from said first noise source at said first speaking location and introduces substantially no noise from said second noise source and does introduce speech from said second person such that said first person can hear said second person substantially free of noise from said first and second noise sources.

15. The invention according to claim 14 comprising a third adaptive filter model adaptively cancelling speech from said second person in the output of said first error microphone, said third adaptive filter model having a model input from said second error microphone, an output correction signal summed with the output of said first error microphone to provide a third sum, and an error input from said third sum, said third sum also being supplied to said second speaker such that said second person can hear the speech of said first person with the speech of the second person removed.

16. The invention according to claim 14 comprising a third adaptive filter model adaptively cancelling speech from said first person in the output of said second error microphone, said third adaptive filter model having a model input from said first error microphone, an output correction signal summed with said first sum to provide a third sum, and an error input from said third sum, said third sum also being summed with said output correction signal of said first adaptive filter model to provide said second sum to said first speaker such that said first person can hear the speech of said second person with the speech of said first person removed.

17. The invention according to claim 14 comprising:

a third adaptive filter model adaptively cancelling speech from said second person in the output of said first error microphone, said third adaptive filter model having a model input from said second error microphone, an output correction signal summed with the output of said first error microphone to provide a third sum, and an error input from said third sum, said third sum also being supplied to said second speaker such that said second person can hear the speech of said first person with the speech of the second person removed;

a fourth adaptive filter model adaptively cancelling speech from said first person in the output of said second error microphone, said fourth adaptive filter model having a model input from said first error microphone, an output correction signal summed with said first sum to provide a fourth sum, and an error input from said fourth sum, said fourth sum also being summed with said output correction signal of said first adaptive filter model to provide said second sum input to said first speaker such that said first person can hear the speech of said second person with the speech of said first person removed.

18. An active acoustic attenuation system comprising:

a first zone subject to noise from a first noise source;

a second zone spaced from said first zone;

a noise microphone sensing noise from said first noise source;

a first speaking location in said first zone such that a first person at said first speaking location is subject to said noise from said first noise source;

a second speaking location in said second zone;

a first speaker introducing sound into said first zone at said first speaking location;

a second speaker introducing sound into said second zone at said second speaking location;

an error microphone sensing noise and speech at said first speaking location;

a speech microphone sensing speech by a second person at said second speaking location;

a first adaptive filter model adaptively modeling an acoustic path from said noise microphone to said first speaking location, said first adaptive filter model having a model input from said noise microphone, an error input from said error microphone, and outputting a correction signal to said first speaker to introduce cancelling sound at said first speaking location to cancel noise from said noise source at said first speaking location;

the output of said first error microphone also being supplied to said second speaker such that said second person can hear the speech of said first person;

the output of said speech microphone being supplied to said first speaker such that said first person can hear the speech of said second person;

a second adaptive filter model adaptively cancelling speech from said second person in the output of said error microphone, said second adaptive filter model having a model input from said speech microphone, an output correction signal summed with the output of said error microphone to provide a sum, and an error input from said sum, said sum also being supplied to said second speaker such that said second person can hear the speech of said first person with the speech of said second person removed.

19. The invention according to claim 18 comprising a third adaptive filter model adaptively cancelling speech from said first person in the output of said speech microphone, said third adaptive filter model having a model input from said error microphone, an output correction signal summed with the output of said speech microphone to provide a second sum, and an error input from said second sum, said second sum also being supplied to said first speaker such that said first person can hear the speech of said second person with the speech of said first person removed.

20. The invention according to claim 19 wherein said second sum is summed with the output correction signal of said first adaptive filter model to provide a third sum which is supplied to said first speaker.

21. The invention according to claim 20 wherein:

the model input of said second adaptive filter model is provided by said second sum;

the model input of said third adaptive filter model is provide by said first sum.

22. An active acoustic attenuation system comprising:

a first zone subject to noise from a first noise source;

a second zone spaced from said first zone;

a noise microphone sensing noise from said first noise source;

a first speaking location in said first zone such that a first person at said first speaking location is subject to said noise from said first noise source;

a second speaking location in said second zone;

a first speaker introducing sound into said first zone at said first speaking location;

a second speaker introducing sound into said second zone at said second speaking location;

an error microphone sensing noise and speech at said first speaking location;

a speech microphone sensing speech by a second person at said second speaking location;

a first adaptive filter model adaptively modeling an acoustic path from said noise microphone to said first speaking location, said first adaptive filter model having a model input from said noise microphone, an error input from said error microphone, and outputting a correction signal to said first speaker to introduce cancelling sound at said first speaking location to cancel noise from said noise source at said first speaking location;

the output of said error microphone also being supplied to said second speaker such that said second person can hear the speech of said first person;

the output of said speech microphone being supplied to said first speaker such that said first person can hear the speech of said second person;

a second adaptive filter model adaptively cancelling speech from said first person in the output of said speech microphone, said second adaptive filter model having a model input from said error microphone, an output correction signal summed with the output of said speech microphone to provide a sum, and an error input from said sum, said sum also being supplied to said first speaker such that said first person can hear the speech of said second person with the speech of said first person removed.

23. The invention according to claim 22 wherein said sum is summed with the output correction signal of said first adaptive filter model to provide a second sum which is supplied to said first speaker.

24. An active acoustic attenuation system comprising:

a first zone subject to noise from a first noise source;

a second zone spaced from said first zone and subject to noise from a second noise source;

a first noise microphone sensing noise from said first noise source;

a second noise microphone sensing noise from said second noise source;

a first speaking location in said first zone such that a first person at said first speaking location is subject to said noise from said first noise source;

a second speaking location in said second zone such that a second person at said second speaking location is subject to said noise from said second noise source;

a first speaker introducing sound into said first zone at said first speaking location;

a second speaker introducing sound into said second zone at said second speaking location;

a first error microphone sensing noise and speech at said first speaking location;

a second error microphone sensing noise and speech at said second speaking location;

a first adaptive filter model adaptively modeling a first acoustic path from said first noise microphone to said first speaking location, said first adaptive filter model having a model input from said first noise microphone, an error input from said first error microphone, and outputting a correction signal to said first speaker to introduce cancelling sound at said first speaking location to cancel noise from said first noise source at said first speaking location;

a second adaptive filter model adaptively modeling a second acoustic path from said second noise microphone to said second speaking location, said second adaptive filter model having a model input from said second noise microphone, an error input from said second error microphone, and outputting a correction signal to said second speaker to introduce cancelling sound at said second speaking location to cancel noise from said second noise source at said second speaking location;

the output of said first error microphone also being supplied to said second speaker such that said second person can hear the speech of said first person;

the output of said second error microphone also being supplied to said first speaker such that said first person can hear the speech of said second person.

25. The invention according to claim 24 wherein:

the output of said first error microphone is summed with the correction signal output of said second adaptive filter model to provide a first sum which is supplied to said second speaker, said output of said first error microphone being substantially free of noise from said first noise source as cancelled by said first adaptive filter model, such that said first sum supplied to said second speaker is substantially free of noise from said first noise source but does contain speech from said first person, such that said second speaker cancels noise from said second noise source at said second speaking location and introduces substantially no noise from said first noise source and does introduce speech from said first person such that said second person can hear said first person substantially free of noise from said first and second noise sources;

the output of said second error microphone is summed with the correction signal output of said first adaptive filter model to provide a second sum which is supplied to said first speaker, said output of said second error microphone being substantially free of noise from said second noise source as cancelled by said second adaptive filter model, such that said second sum supplied to said first speaker is substantially free of noise from said second noise source but does contain speech from said second person, such that said first speaker cancels noise from said first noise source at said first speaking location and introduces substantially no noise from said second noise source and does introduce speech from said second person such that said first person can hear said second person substantially free of noise from said first and second noise sources.

26. The invention according to claim 24 comprising a third adaptive filter model adaptively cancelling noise from said first noise source in the output of said first error microphone, said third adaptive filter model having a model input from said first noise microphone, an output correction signal summed with the output of said first error microphone to provide a sum, and an error input from said last mentioned sum.

27. The invention according to claim 24 comprising a third adaptive filter model adaptively cancelling noise from said second noise source in the output of said second error microphone, said third adaptive filter model having a model input from said second noise microphone, an output correction signal summed with the output of said second error microphone to provide a sum, and an error input from said last mentioned sum.

28. The invention according to claim 24 comprising a third adaptive filter model adaptively cancelling speech from said second person in the output of said first error microphone, said third adaptive filter model having a model input from said second error microphone, an output correction signal summed with the output of said first error microphone to provide a sum, and an error input from said last mentioned sum.

29. The invention according to claim 24 comprising a third adaptive filter model adaptively cancelling speech from said first person in the output of said second error microphone, said third adaptive filter model having a model input from said first error microphone, an output correction signal summed with the output of said second error microphone to provide a sum, and an error input from said last mentioned sum.

30. An active acoustic attenuation system comprising:

a first zone subject to noise from a first noise source;

a second zone spaced from said first zone and subject to noise from a second noise source;

a first noise microphone sensing noise from said first noise source;

a second noise microphone sensing noise from said second noise source;

a first speaking location in said first zone such that a first person at said first speaking location is subject to said noise from said first noise source;

a second speaking location in said second zone such that a second person at said second speaking location is subject to said noise from said second noise source;

a first speaker introducing sound into said first zone at said first speaking location;

a second speaker introducing sound into said second zone at said second speaking location;

a first error microphone sensing noise and speech at said first speaking location;

a second error microphone sensing noise and speech at said second speaking location;

a first adaptive filter model adaptively modeling a first acoustic path from said first noise microphone to said first speaking location, said first adaptive filter model having a model input from said first noise microphone, an error input from said first error microphone, and outputting a correction signal to said first speaker to introduce cancelling sound at said first speaking location to cancel noise from said first noise source at said first speaking location;

a second adaptive filter model adaptively modeling a second acoustic path from said second noise microphone to said second speaking location, said second adaptive filter model having a model input from said second noise microphone, an error input from said second error microphone, and outputting a correction signal to said second speaker to introduce cancelling sound at said second speaking location to cancel noise from said second noise source at said second speaking location;

a third adaptive filter model adaptively cancelling noise from said first noise source in the output of said first error microphone, said third adaptive filter model having a model input from said first noise microphone, an output correction signal summed with the output of said first error microphone to provide a first sum, and an error input from said first sum;

said first sum also being supplied to said second speaker such that said second person can hear the speech of said first person;

a fourth adaptive filter model adaptively cancelling noise from said second noise source in the output of said second error microphone, said fourth adaptive filter model having a model input from said second noise microphone, an output correction signal summed with the output of said second error microphone to provide a second sum, and an error input from said second sum;

said second sum also being supplied to said first speaker such that said first person can hear the speech of said second person.

31. The invention according to claim 30 wherein:

said first sum is summed with the correction signal output of said second adaptive filter model to provide a third sum which is supplied to said second speaker, said first sum being substantially free of noise from said first noise source as cancelled by said first and third adaptive filter models, such that said third sum to said second speaker is substantially free of noise from said first noise source but does contain speech from said first person, such that said second speaker cancels noise from said second noise source at said second speaking location and introduces substantially no noise from said first noise source and does introduce speech from said first person such that said second person can hear said first person substantially free of noise from said first and second noise sources;

said second sum is summed with the correction signal output of said first adaptive filter model to provide a fourth sum which is supplied to said first speaker, said fourth sum being substantially free of noise from said second noise source as cancelled by said second and fourth adaptive filter models, such that said fourth sum to said first speaker is substantially free of noise from said second noise source but does contain speech from said second person, such that said first speaker cancels noise from said first noise source at said first speaking location and introduces substantially no noise from said second noise source and does introduce speech from said second person such that said first person can hear said second person substantially free of noise from said first and second noise sources.

32. The invention according to claim 31 wherein:

the output correction signal from said second adaptive filter model is additively summed with said first sum to provide said third sum;

the output correction signal from said first second sum to provide said fourth sum;

the output correction signal from said third adaptive filter is subtractively summed with the output of said first error microphone to provide said first sum;

the output correction signal from said fourth adaptive filter model is subtractively summed with the output of said second error microphone to provide said second sum.

33. An active acoustic attenuation system comprising:

a first zone subject to noise from a first noise source;

a second zone spaced from said first zone and subject to noise from a second noise source;

a first noise microphone sensing noise from said first noise source;

a second noise microphone sensing noise from said second noise source;

a first speaking location in said first zone such that a first person at said first speaking location is subject to said noise from said first noise source;

a second speaking location in said second zone such that a second person at said second speaking location is subject to said noise from said second noise source;

a first speaker introducing sound into said first zone at said first speaking location;

a second speaker introducing sound into said second zone at said second speaking location;

a first error microphone sensing noise and speech at said first speaking location;

a second error microphone sensing noise and speech at said second speaking location;

a first adaptive filter model adaptively modeling a first acoustic path from said first noise microphone to said first speaking location, said first adaptive filter model having a model input from said first noise microphone, an error input from said first error microphone, and outputting a correction signal to said first speaker to introduce cancelling sound at said first speaking location to cancel noise from said first noise source at said first speaking location;

a second adaptive filter model adaptively modeling a second acoustic path from said second noise microphone to said second speaking location, said second adaptive filter model having a model input from said second noise microphone, an error input from said second error microphone, and outputting a correction signal to said second speaker to introduce cancelling sound at said second speaking location to cancel noise from said second noise source at said second speaking location;

a third adaptive filter model adaptively cancelling noise from said first noise source in the output of said first error microphone, said third adaptive filter model having a model input from said first noise microphone, an output correction signal summed with the output of said first error microphone to provide a first sum, and an error input from said first sum;

a fourth adaptive filter model adaptively cancelling noise from said second noise source in the output of said second error microphone, said fourth adaptive filter model having a model input from said second noise microphone, an output correction signal summed with the output of said second error microphone to provide a second sum, and an error input from said second sum;

a fifth adaptive filter model adaptively cancelling speech from said second person in the output of said first error microphone, said fifth adaptive filter model having a model input from said second error microphone, an output correction signal summed with said first sum to provide a third sum, and an error input from said third sum;

said third sum also being supplied to said second speaker such that said second person can hear the speech of said first person;

a sixth adaptive filter model adaptively cancelling speech from said first person in the output of said second error microphone, said sixth adaptive filter model having a model input from said first error microphone, an output correction signal summed with said second sum to provide a fourth sum, and an error input from said fourth sum;

said fourth sum also being supplied to said first speaker such that said first person can hear the speech of said second person.

38. An active acoustic attenuation system comprising:

a first zone subject to noise from a first noise source;

a second zone spaced from said first zone and subject to noise from a second noise source;

a first noise microphone sensing noise from said first noise source;

a second noise microphone sensing noise from said second noise source;

a first speaking location in said first zone such that a first person at said first speaking location is subject to said noise from said first noise source;

a second speaking location in said second zone such that a second person at said second speaking location is subject to said noise from said second noise source;

a first speaker introducing sound into said first zone at said first speaking location;

a second speaker introducing sound into said second zone at said second speaking location;

a first error microphone sensing noise and speech at said first speaking location;

a second error microphone sensing noise and speech at said second speaking location;

a first adaptive filter model adaptively modeling a first acoustic path from said first noise microphone to said first speaking location, said first adaptive filter model having a model input from said first noise microphone, an error input from said first error microphone, and outputting a correction signal to said first speaker to introduce cancelling sound at said first speaking location to cancel noise from said first noise source at said first speaking location;

a second adaptive filter model adaptively modeling a second acoustic path from said second noise microphone to said second speaking location, said second adaptive filter model having a model input from said second noise microphone, an error input from said second error microphone, and outputting a correction signal to said second speaker to introduce cancelling sound at said second speaking location to cancel noise from said second noise source at said second speaking location;

a third adaptive filter model adaptively cancelling speech from said second person in the output of said first error microphone, said third adaptive filter model having a model input from said second error microphone, an output correction signal summed with the output of said first error microphone to provide a first sum, and an error input from said first sum;

a fourth adaptive filter model adaptively cancelling speech from said first person in the output of said second error microphone, said fourth adaptive filter model having a model input from said first error microphone, an output correction signal summed with the output of said second error microphone to provide a second sum, and an error input from said second sum;

said first sum also being supplied to said second speaker such that said second person can hear the speech of said first person;

said second sum also being supplied to said first speaker such that said first person can hear the speech of said second person.

34. The invention according to claim 39 wherein:

the model input of said fifth adaptive filter model is provided by said fourth sum;

the model input of said sixth adaptive filter model is provided by said third sum.

35. The invention according to claim 34 wherein said third sum is summed with the correction signal output of said second adaptive filter model to provide a fifth sum which is supplied to said second speaker, said fifth sum being substantially free of noise from said first noise source as cancelled by said first and third adaptive filter models, and being substantially free of speech from said second person as cancelled by said fifth adaptive filter model, such that said fifth sum to said second speaker is substantially free of noise from said first noise source and speech from said second person but does contain speech from said first person, such that said second speaker cancels noise from said second noise source at said second speaking location and introduces substantially no noise from said first noise source and introduces substantially no speech from said second person and does introduce speech from said first person such that said second person can hear said first person substantially free of noise from said first and second noise sources and substantially free of his own speech;

said fourth sum is summed with the correction signal output of said first adaptive filter model to provide a sixth sum which is supplied to said first speaker, said sixth sum being substantially free of noise from said second noise source as cancelled by said second and fourth adaptive filter models, and being substantially free of speech from said first person as cancelled by said sixth adaptive filter model, such that said sixth sum to said first speaker is substantially free of noise from said second noise source but does contain speech from said second person, such that said first speaker cancels noise from said first noise source at said first speaking location and introduces substantially no noise from said second noise source and introduces substantially no speech from said first person and does introduce speech from said second person such that said first person can hear said second person substantially free of noise from said first and second noise sources and substantially free of his own speech.

36. The invention according to claim 35 wherein:

the output correction signal from said fifth adaptive filter model is subtractively summed with said first sum to provide said third sum;

the output correction signal from said sixth adaptive filter model is subtractively summed with said second sum to provide said fourth sum.

37. The invention according to claim 36 wherein:

the output correction signal from said third adaptive filter model is subtractively summed with the output of said first error microphone to provide said first sum;

the output correction signal from said fourth adaptive filter model is subtractively summed with the output of said second error microphone to provide said second sum.

39. The invention according to claim 38 wherein said first sum is summed with the correction signal output of said second adaptive filter model to provide a third sum which is supplied to said second speaker, said third sum being substantially free of noise from said first noise source as cancelled by said first adaptive filter model, such that said first sum to said second speaker is substantially free of noise from said first noise source but does contain speech from said first person, such that said second speaker cancels noise from said second noise source at said second speaking location and introduces substantially no noise from said first noise source and does introduce speech from said first person such that said second person can hear said first person substantially free of noise from said first and second noise sources;

said second sum is summed with the correction signal output of said first adaptive filter model to provide a fourth sum which is supplied to said first speaker, said fourth sum being substantially free of noise from said second noise source as cancelled by said second adaptive filter model, such that said fourth sum to said first speaker is substantially free of noise from said second noise source but does contain speech from said second person, such that said first speaker cancels noise from said first noise source at said first speaking location and introduces substantially no noise from said second noise source and does introduce speech from said second person such that said first person can hear said second person substantially free of noise from said first and second noise sources.