Data processing: speech signal processing – linguistics – language – Speech signal processing – Application
Reexamination Certificate
2001-11-27
2004-11-23
McFadden, Susan (Department: 2655)
Data processing: speech signal processing, linguistics, language
Speech signal processing
Application
C704S500000, C381S017000
Reexamination Certificate
active
06823310
ABSTRACT:
BACKGROUND OF THE INVENTION
1. Field of the Invention
The present invention relates to an audio decoding apparatus used in AV (audio visual) equipment for decoding an encoded bit stream into PCM data. The present invention also relates to a signal processing device, a sound image localization device, a sound image control method, an audio signal processing device, and an audio signal high-rate reproduction method also used in AV equipment.
2. Description of the Related Art
A conventional audio decoding apparatus
550
will be described with reference to
FIGS. 6
,
7
and
8
.
FIG. 6
is a block diagram illustrating a structure of the conventional audio decoding apparatus
550
. The audio decoding apparatus
550
includes an integrated semiconductor device
508
. The integrated semiconductor device
508
includes an input bit stream syntax analyzer
501
, an exponential section decoder
502
, a mantissa data bit allocator
503
, a mantissa section decoder
504
, an IMDCT
505
, a down-mix operator
506
, and an internal memory device
507
. The integrated semiconductor device
508
exchanges data with an external memory device
500
.
A bit stream is first stored in the external memory device
500
and then input to the input bit stream syntax analyzer
501
. The input bit stream syntax analyzer
501
analyzes the syntax of the bit stream and extracts data required for decoding. Such data is sent to the exponential section decoder
502
. The exponential section decoder
502
forms exponential data for a frequency domain from the data required for decoding, and output the exponential data to the mantissa data bit allocator
503
and the IMDCT
505
. The mantissa data bit allocator
503
calculates a mantissa data bit allocation amount from the exponential data for the frequency domain and the data stored in the external memory device
500
, and outputs the mantissa data bit allocation amount to the mantissa section decoder
504
. The mantissa section decoder
504
forms mantissa data for the frequency domain from the mantissa data bit allocation amount and outputs the mantissa data to the IMDCT (inverted modified discrete cosine transformer)
505
. The IMDCT
505
forms decoded audio data in a time domain from the exponential data and the mantissa data for the frequency domain, and stores the decoded audio data in the external memory device
500
. The down-mix operator
506
forms PCM data from the decoded audio data stored in the external memory device
500
, performs interleaving and then stores n the resultant data in the external memory device
500
. The PCM data is then output from the external memory device
500
.
FIG. 7
is a memory map of the audio decoding apparatus
550
shown in FIG.
6
. The memory map shown in
FIG. 7
includes an area
600
for storing one-block PCM data, an area
601
for storing one-block decoded audio data for channel
0
, an area
602
for storing one-block decoded audio data for channel
1
, an area
603
for storing one-block decoded audio data for channel
2
, an area
604
for storing one-block decoded audio data for channel
3
, an area
605
for storing one-block decoded audio data for channel
4
, and an area
606
for storing one-block decoded audio data for channel
5
.
FIG. 8
is a flowchart illustrating a method for decoding one-block encoded audio data for each channel.
In step S
11
, a register (not shown), the internal memory device
507
(FIG.
6
), and an external memory device
500
are initialized. In step S
12
, the bit stream stored in the external memory device
500
is input to the integrated semiconductor device
508
(receipt of encoded data).
Then, in step S
13
, the syntax of the bit stream is analyzed, and data required for decoding is extracted (bit stream analysis). In step S
14
, exponential data for a frequency domain is formed using the extracted data. In step S
15
, a mantissa data bit allocation amount is calculated using the exponential data for the frequency domain. In step S
16
, mantissa data for the frequency domain is formed using the mantissa data bit allocation amount. In step S
17
, decoded audio data is formed using the mantissa data for the frequency domain and the exponential data for the frequency domain. In step S
18
, the resultant decoded audio data is stored in the external memory device
500
.
The above-described steps are executed for the number of channel included in one block until it is confirmed in step S
19
that the steps are repeated for the required times. As a result, the number of pieces of decoded audio data corresponding to the number of channels included in one block are formed and stored in the external memory device
500
.
In step S
20
, one-block decoded audio data for each channel in the external memory device
500
is input to the integrated semiconductor device
508
. In step S
21
, the one-block decoded audio data for each channel is converted into one-block PCM data (down-mix calculation). In step S
22
, the one-block PCM data is output to the external memory device
500
.
In the conventional audio decoder
600
, one-block PCM data is calculated in one down-mix calculation. Accordingly, the amount of data transferred for inputting the decoded audio data to the external memory device
500
before the down-mix calculation and for writing the PCM data to the external memory device
500
after the down-mix calculation is sufficiently large to occupy a significant part of the memory bus. Such an occupation has an adverse effect on other processing performed by the external memory device
500
.
A conventional signal processing device will be described. A part of the encoded data of a plurality of channels can be commonly shared by the channels. For example, high frequency band encoded data which is included in at least one of the plurality of channels and shared by the plurality of channels is decoded to form high frequency band decoded data. Low frequency band encoded data for each channel is decoded to form low-frequency band decoded data. The low-frequency band decoded data is coupled with the high-frequency band decoded data to form decoded data for each channel.
Such decoding will be described with reference to
FIGS. 19
,
20
and
21
.
FIG. 20
is a block diagram of a conventional signal processor
1350
for performing the above-described signal decoding. As shown in
FIG. 20
, the bit stream is temporarily stored in an internal memory device
1301
, and analyzed by a bit stream syntax analyzer
1300
. Thus, required data is extracted. Exponential data for a frequency domain is formed by an exponential section decoder
1302
based on the extracted data. A mantissa data bit allocation amount is determined by a mantissa data bit allocator
1303
based on the exponential data for the frequency domain. Mantissa data is formed by a mantissa section decoder
1304
based on the mantissa data bit allocation amount. Frequency domain data is formed by a frequency domain data forming device
1305
based on the data formed by the exponential section decoder
1302
and the mantissa section decoder
1304
.
The frequency domain data forming device
1305
decodes encoded data for an arbitrary channel in the following rule. High frequency encoded data which is included in at least one of a plurality of channels and shared by the plurality of channels is decoded to obtain high frequency band decoded data, and the high frequency band decoded data is multiplied by the ratio of the signal power of a prescribed channel obtained by an encoder with respect to the signal power of an arbitrary channel. The result is coupled with the low frequency decoded data for an arbitrary channel. Thus, decoded data for the arbitrary channel is obtained.
The obtained frequency domain decoded data is converted into time domain decoded data by a frequency domain-time domain converter
1306
, and the result is converted into PCM data, which is output.
FIG. 21
schematically shows decoding of encoded data for an arbitrary channel.
In step
141
, data in a prescribed channel
1400
is decoded to form a low frequ
Abe Kazutaka
Fujita Takeshi
Ishito Tsukuru
Katayama Takashi
Kawamura Akihisa
McFadden Susan
Renner , Otto, Boisselle & Sklar, LLP
LandOfFree
Audio signal processing device and audio signal high-rate... 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 signal processing device and audio signal high-rate..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Audio signal processing device and audio signal high-rate... will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-3340765