Data processing: speech signal processing – linguistics – language – Speech signal processing – For storage or transmission
Reexamination Certificate
1997-12-12
2001-05-15
Tsang, Fan (Department: 2645)
Data processing: speech signal processing, linguistics, language
Speech signal processing
For storage or transmission
Reexamination Certificate
active
06233554
ABSTRACT:
BACKGROUND OF THE INVENTION
I. Field of the Invention
The present invention relates to audio band encoders and decoders. More particularly, the present invention relates to a novel and improved audio band encoder/decoder (CODEC) with an internal amplifier which is gain controlled by a voice encoder/decoder (VOCODER).
II. Description of the Related Art
In the art of digital audio processing, an audio band encoder/decoder (CODEC) is typically used to convert analog audio signals to encoded digital signals, and vice-versa. For example, the CODEC may receive the analog output of a microphone, and convert the microphone-generated analog audio signal into a Pulse Code Modulation (PCM) encoded digital audio signal for further digital signal processing in a digital signal processor (DSP). Additionally, the CODEC may receive a PCM-encoded digital audio signal from the DSP and convert it to an analog audio signal for use by an audio speaker. Of course, the CODEC may use additional digital encoding techniques as are known in the art, such as A-law, &mgr;-law, or the like, or may merely use any other linear or non-linear encoding techniques.
A typical CODEC as known in the art may include an internal or more likely, an external audio amplifier of fixed or variable gain. The analog audio signal from the microphone is amplified by this amplifier prior to being digitally encoded and sent to the DSP. The gain of the amplifier is chosen so that it provides a linear gain over the expected operational range. However, in many applications, the analog audio signal generated by the microphone may exceed the dynamic range of the encoder In such a case, the encoded signal will saturate and become distorted. Clearly, this is undesirable because the audio signal will not be able to be acceptably reproduced due to the distortion.
This problem is apparent in digital telephone systems. In many digital telephone systems, the PCM-encoded output of the CODEC is further encoded by a vocoder to minimize the amount of digital data used to represent audio signals. The vocoder employs techniques to compress audio signals such as speech by extracting parameters that relate to a model of human speech. Vocoders are well known in the art. Examples of various vocoder designs may be found in U.S. Pat. No. 5,414,796, entitled “VARIABLE RATE VOCODER,” U.S. Pat. No. 5,341,456, entitled “METHOD FOR DETERMINING SPEECH ENCODING RATE IN A VARIABLE RATE VOCODER,” and U.S. Pat. No. 5,784,532, entitled “VOCODER ASIC.” Each of the above-named patents and applications is assigned to the assignee of the present invention and incorporated herein by reference. If the level of the analog audio signal input to the CODEC is high enough that it causes clipping, then the resultant PCM-encoded speech signal presented to the vocoder will be distorted and the vocoder will not be able to properly model the signal for proper reproduction when decoded. This is commonly called the “loud talker” problem.
In order to overcome the “loud talker” problem, it would be desireable to perform automatic gain control (AGC) on the audio amplifier, turning down the audio amplifier gain when the audio energy level is high, and turning up the audio amplifier gain when the audio energy level is low (i.e., for “soft talkers”). This would allow the CODEC to generate gain-controlled PCM-encoded audio data for use by the vocoder.
However, the vocoder typically estimates the level of background noise (i.e., non-speech components in the PCM-encoded output from the CODEC) and uses this estimated background noise level to perform background noise suppression. If the vocoder is presented with a gain-controlled input signal from the CODEC, then the background noise level will be changing according to how loudly the user is talking. This introduces inaccuracies into the background noise level estimation process.
Thus, what is needed is an automatic gain control circuit which overcomes the “loud talker” problem without introducing inaccuracies into the vocoder's background noise level estimation process.
SUMMARY OF THE INVENTION
The present invention is a novel and improved audio band processor and a method and circuit for controlling the gain of the audio band processor. In the preferred embodiment, the audio band processor comprises a first variable gain amplifier for receiving an analog audio signal; an analog to digital converter for converting said analog audio signal to a digital audio signal; and an interface circuit for encoding said digital audio signal. The interface circuit also arranges said digital audio signal into a sequence of encoded transmit frames, and receives a sequence of encoded receive frames, each receive frame comprising a predetermined number of encoded digital audio signal data bits and a predetermined number of digital gain control bits. The interface circuit also adjusts a gain level of said first variable gain amplifier in response to said digital gain control bits. In another embodiment, the audio band processor further comprises a second variable gain amplifier, and the interface circuit adjusts a gain level of said second variable gain amplifier in response to said digital gain control bits.
The novel audio band processor of the present invention may be used in conjunction with a circuit for controlling the gain of the audio band processor. The circuit comprises a first variable gain amplifier in said audio band processor for receiving an analog audio signal; an analog to digital converter in said audio band processor for converting said analog audio signal to a digital audio signal; an interface circuit for encoding said digital audio signal and for arranging said digital audio signal into a sequence of encoded transmit frames; and a vocoder for measuring an energy level of said sequence of encoded transmit frames. The vocoder also generates digital gain control bits in response to said measured energy level, and generates a sequence of encoded receive frames, each receive frame comprising a predetermined number of encoded digital audio signal data bits a predetermined number of padding bits. The vocoder futher multiplexes said digital gain control bits into said padding bits. The interface circuit adjusts a gain level of said first variable gain amplifier in response to said digital gain control bits. A method for using the present invention is also disclosed.
REFERENCES:
patent: 5117228 (1992-05-01), Fuchigama
patent: 5677962 (1997-10-01), Harrison et al.
“Voice-Band Audio Processors Application Report”; TCM320ACxx (5-V Versions), TLV320ACxx (3-V Versions); SLWA001, Jun. 1996; Texas Instruments.
“TCM320AC3x/4x Voice-Band Audio Processors—Application Report”; SPRA146, Oct. 1994; Greg Davis, Russ McDonald, Advanced Linear Applications—Semiconductor Group; Texas Instruments.
Heimbigner Wade L.
Hutchison, IV James A.
Oliveira Louis D.
Brown Charles D.
Opsasnick Michael N.
Qualcomm Incorporated
Tsang Fan
Wadsworth Philip R.
LandOfFree
Audio CODEC with AGC controlled by a VOCODER does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Audio CODEC with AGC controlled by a VOCODER, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Audio CODEC with AGC controlled by a VOCODER will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-2466205