Digital speech processing system

• US 6,420,986 B1
• Filed: 08/02/2000
• Issued: 07/16/2002
• Est. Priority Date: 10/20/1999
• Status: Active Grant

• Current Assignee: Motorola Solutions, Inc.
• Original Assignee: Motorola, Inc. (Motorola Solutions, Inc.)
• Inventors: Cohen, Aviad, Shahaf, Mark, Avayu, Izak
• Primary Examiner(s): Young, Brian
• Assistant Examiner(s): NGUYEN, JOHN B
• Application Number:
  US09/630,755
• Time in Patent Office: 713 Days
• Field of Search: 341/139, 704/225, 704/233, 704/221, 704/205, 704/230, 704/226, 704/231, 381/46, 381/50, 381/42, 381/41
• US Class Current: 341/139
• CPC Class Codes: H03M 1/183 the feedback signal controlling the gain of an amplifier or attenuator preceding the analogue/digital converter

Abstract

In a speech processing system (10) characterized by a finite range of audio levels, the speech processing system (10) receiving an incoming audio signal, the speech processing system amplifying (12) the incoming audio signal by an audio gain factor, the speech processing system (10) representing the amplified audio signal by the finite range of audio levels, a method for adjusting the audio gain factor, including the steps of: decreasing the audio gain factor when detecting clipping of the amplified audio signal, maintaining the audio gain factor for a hold time period, and increasing the gain factor when detecting that the result of amplification of the incoming sound levels by the audio gain factor, is lower than the highest level of the finite range of audio levels.

Claims

1. A method of operation in a speech processing system, including receiving an incoming audio frequency signal by the speech processing system, amplifying the incoming audio frequency signal by an audio gain factor, representing the amplified audio signal by a finite range of audio levels, and adjusting the audio gain factor by applying the following steps:

in an attack mode initiated when a speech signal is detected to be present, analyzing one or more speech blocks of the amplified audio frequency signal to determine whether clipping of the amplified audio frequency signal has taken place;

in a hold mode applied whilst no gain adjustment is being applied and when it is detected in the attack mode that clipping has taken place, analyzing the energy level of one or more blocks of the amplified audio frequency signal to determine what required adjustment should be applied to the audio gain factor; and

in a release mode applied following the hold mode, adjusting the audio gain factor according to the required adjustment determined in the hold mode.

2. A method according to claim 1 wherein a rate for applying the gain adjustment in the release mode is calculated in the hold mode.

3. A method according to claim 1 wherein the gain adjustment in the release mode is applied in steps.

4. A method according to claim 3 wherein the size of the steps applied is calculated in the hold mode.

5. A method according to claim 1 wherein from an energy level measurement of one or more speech blocks a corresponding average block peak value is calculated by which determination of required gain adjustment is made.

6. A method according to claim 1 wherein clipping is determined to have taken place when the result of amplification of said incoming audio frequency signals by the audio gain factor exceeds the highest level of said finite range of amplified audio levels.

7. A method according to claim 1 wherein clipping is determined to have taken place when the result of amplification of the average of the incoming sound levels by the audio gain factor exceeds the highest level of the finite range of amplified audio levels.

8. A method according to claim 1 wherein clipping is determined to have taken place when the result of amplification of RMS energy value of said incoming sound levels by the audio gain factor exceeds the highest level of the finite range of amplified audio levels.

9. A method according to claim 1 wherein clipping is determined to have taken place when a mapped value of the result of amplification of an RMS value of the incoming sound levels by said audio gain factor exceeds the highest level of the finite range of amplified audio levels.

10. A method according to claim 1 wherein the step of adjusting the audio gain is performed in the presence of speech.

11. A method according to claim 1 which includes the step of detecting speech in said incoming audio frequency signal by a voice activity detector.

12. A method according to claim 1 wherein the hold mode is applied for a time period which is predetermined.

13. A method according to claim 1 wherein the hold mode is applied for a time period which is dynamically variable.

14. A method according to claim 1 wherein the amplified incoming audio frequency signal is further processed as an analogue signal.

15. A method according to claim 1 wherein the amplified incoming audio frequency signal is further processed as a digital signal.