Methods for processing audio information in a multiple processor audio decoder

• US 6,253,293 B1
• Filed: 01/14/2000
• Issued: 06/26/2001
• Est. Priority Date: 11/14/1997
• Status: Expired due to Term

• Current Assignee: Cirrus Logic Incorporated
• Original Assignee: Cirrus Logic Incorporated
• Inventors: Luo, Zheng, Rao, Raghunath, Dokic, Miroslav, Divine, James, Niehaus, Jeffrey
• Primary Examiner(s): BRAGDON, REGINALD GLENWOOD
• Application Number:
  US09/483,290
• Time in Patent Office: 529 Days
• Field of Search: 704/500, 709/213, 709/201, 709/208, 709/248, 711/147, 711/149, 700/94
• US Class Current: 711/147
• CPC Class Codes: G10L 19/16 Vocoder architecture

Abstract

A method of processing a stream of audio information received by a multiple processor audio decoder. Processing operations are performed by a first processor on the stream of audio information to produce at set of results. The first processor writes the set of results into a shared memory and a flag is set indicating that the results are ready. In response to the flag, a second processor reads the results from shared memory. When the results have been read from shared memory, the second processor sends a command to the first processor. The first processor then clears the flag.

Claims

1. A method of processing audio information in a multiple processor audio decoder comprising the steps of:

receiving a stream of audio information;

performing processing operations on the stream of audio information with a first processor to produce a set of results;

writing the set of results into a shared memory with the first processor;

setting a flag indicating that the set of results are ready;

reading the set of results from the shared memory with a second processor in response to said step of setting a flag;

sending a command from the second processor to the first processor indicating that the second processor has read the set of results from the shared memory; and

clearing the flag with the first processor.

2. The method of claim 1 wherein the audio information received by the first processor comprises audio information encoded using transform encoding and said step of performing processing operations comprises the step of producing a set of frequency domain transform coefficients.

3. The method of claim 2 wherein said step of performing processing operations with the first processor further comprises the substep of recovering data fields from the stream of audio information.

4. The method of claim 1 and further comprising the step of performing processing operations on a second set of results previously read from the shared memory with the second processor substantially concurrently with said step of performing operations with the first processor.

5. The method of claim 4 wherein said step of performing processing operations on a second set of results with the second processor comprises the step of performing inverse transform operations on the data fields recovered by the first processor using the transform coefficients produced by the first processor.

6. The method of claim 1 and further comprising the step of interrupting the first processor after said step of setting the flag.

7. A method of decompressing a stream of audio data using first and second digital signal processors and shared memory for exchanging data therebetween comprising the steps of:

receiving a stream of compressed audio data at an input to an audio decoder;

generating a set of frequency domain coefficients from the stream of compressed audio data using the first digital signal processor;

loading the frequency domain coefficients into the shared memory and setting a flag with the first processor;

reading the frequency domain coefficients from the shared memory with the second processor;

clearing the flag with the first processor after performing said step of reading;

performing inverse transform operations with the second processor on data fields recovered from the stream of compressed audio data; and

substantially concurrent with said step of performing inverse transform operations with the second processor generating a second set of frequency domain coefficients from the stream of compressed audio data with the first processor.

8. The method of claim 7 wherein said first and second processors are fabricated together on a chip.

9. The method of claim 7 and further comprising the step of transmitting a signal from the second processor to the first processor upon completion of said step of reading.

10. The method of claim 7 and further comprising the step of recovering the data fields from the stream of compressed audio data with the first processor.

11. The method of claim 7 wherein the stream of compressed audio data comprises a stream of AC3 encoded audio data.

12. The method of claim 7 and further comprising the step of setting a bit in a shared register with the second processor after completion of said step of reading.