摘要:
Overlay buffering scheme for multi-channel data in which one memory buffer content is overlayed over another as memory locations of an input buffer are freed when data is output from the input buffer. By overlaying the buffer content, only one input buffer is used, reducing the needed memory by half.
摘要:
A method of operating a multiple processor device. A first word of a sequence of words is received in a register. A target processor is determined from the first word and the target processor is interrupted. An input ready bit is set and first word from in the register is read with the target processor. A number of words in the sequence to follow the first word determined from the first word. A word counter is set and the input ready bit is cleared with the target processor. The target processor is returned to main code execution.
摘要:
A method of processing exponent data in an audio decoder. A first block of audio data is received including encoded exponent data. The encoded exponent data is packed into packed encoded words and stored in memory. Exponents are generated from the packed encoded words in memory for processing the first block of audio data. A second block of audio data is received. A determination is made as to whether a reuse flag has been set for the second block, and if the reuse flag has been set, exponents are generated from the packed encoded words memory for processing the second block of data.
摘要:
Digital tone controls 500 include a first path 502 including a digital filter 504 and a scaler 505 for controlling a level of a low frequency component of a received digital audio signal. A second 502 includes a digital filter 504 and a scaler 505 for controlling a level of a high frequency component of the received digital audio signal. A third path 503 includes a scaler 506 for controlling a level of an unfiltered component of the received audio signal. A summer 507 adds a contribution from each of the paths to generate a composite signal having a selected gain-frequency response.
摘要:
A method of exchanging messages between first and second processors. A pending flag in a first register is polled by the first processor and if the flag is in a first selected logic state, a message is written into a second register with the first processor. The pending flag is set to a second selected logic state with the first processor and an interrupt to the second processor is generated. The message is read from the second register with the second processor when the pending flag is in the second logic state. The pending flag set to the first logic state with the second processor.
摘要:
A method of concealing errors received in a stream of data by a multiprocessor system having a first digital signal processor for initial processing of incoming data and a second digital signal processor for processing data passed from the first processor. The method includes the steps of detecting an error in an incoming frame of the stream of data with the first processor, sending an error message from the first to the second processor, halting transmission of the remainder of the frame to the second processor, and processing a frame of dummy data with the second processor. The frame of dummy data is passed to a processing engine forming a portion of the second processor to maintain data pipelining. Interprocessor communications between the first and second processors is established by handshaking through an interprocessor communications register, wherein the first processor detects an error in a frame of data and sends a message to the second processor. In selected embodiments, the first and second processors are fabricated on a single integrated circuit chip along with shared memory for exchanging data between the processors and data memory associated with each processor.
摘要:
A method of managing multiple channels of audio data in an audio system having multiple speakers. A first channel signal is selectively passed through a software high-pass filter to selectively drive a first one of the speakers. A plurality of channel signals are selectively summed in software to generate a composite signal and the composite signal passed through a software low-pass filter to selectively drive a second one of the speakers.
摘要:
An audio decoder 100 for operating on a received compressed audio data stream compressed using an algorithm employing transform encoding and a bit allocation routine. A first processor 200 performs a first set of operations on the received compressed audio data stream including parsing the compressed audio data stream, recovering data fields within the compressed audio data stream, calculating a bit allocation, and passing frequency domain coefficients to shared memory. A second digital signal processor 100b performs a second set of operations on data passed from the first digital signal processor to shared memory including performing inverse transform operations on the data passed from the first digital signal processor.
摘要:
A method for transferring data bursts via a synchronous data link includes the step of receiving a burst of packets, each packet including a header and a frame of data compressed at a selected sampling rate and transmitted at a selected bit rate. At least one of the packets of the stream of packets is embedded into a carrier frame including a carrier frame header. The carrier frame is then transmitted via the synchronous link. The data frame is extracted from the carrier frame and decompressed at the sample rate.
摘要:
A method of operating shared memory in a multiple processor system. A token is by default maintained with a first processor, the token enabling access to shared memory. A determination is made that a second processor requires access to shared memory. A determination is also made as to whether the first processor is accessing to the shared memory. The token is transferred the second processor if the first processor is not accessing the shared memory. The second processor accesses the shared memory with the token.