Pulse or digital communications – Bandwidth reduction or expansion – Television or motion video signal
Reexamination Certificate
2000-07-25
2004-02-17
Kelley, Chris (Department: 2713)
Pulse or digital communications
Bandwidth reduction or expansion
Television or motion video signal
C375S240250
Reexamination Certificate
active
06693963
ABSTRACT:
BACKGROUND OF THE INVENTION
1. Technical Field of the Invention
The present invention relates generally to a subband encoding and a decoding system used in data compression and decompression which are capable of working at a decreased operation load.
2. Background Art
As typical subband encoding, the MPEG1 audio is known in the art.
FIG. 46
shows a conventional MPEG1 audio layer
1
encoding system. The bandwidth of digital signals
1101
sampled at a sampling frequency fs inputted to the encoding system is divided by a band splitter
101
into k subbands whose overall band is equivalent to the bandwidth of the Nyquist frequency (fs/2) of the inputted digital signals
1101
. The band splitter
101
outputs k subband signals
1102
to the encoding circuit
104
. Note that k is an integer. In the MPEG1 audio, the bandwidth of an input signal is divided into 32 regular subbands, but may alternatively be divided into a predetermined number of irregular subbands depending upon the design of filters.
In the MPEG1 audio layer, the subband signal in each subband is down-sampled to produce a baseband signal using a frequency modulation technique. Simultaneously, w of the input digital signals
1101
sampled at intervals of the reciprocal of the sampling frequency (1/fs) are time-frequency converted in the time-frequency converter
102
using the time-windowing in synchronism in time with the band splitter
101
to produce frequency information
1103
. The length w of a time window used in the time-frequency conversion is determined as
w=(
1
/fr
)/(1
/fs
)
where fr is a frequency resolution required in the frequency information
1103
.
In the MPEG1 audio, the time-frequency conversion is accomplished with the Fast Fourier Transform (FFT), so that the value of w is minimum two to the nth power meeting a desired frequency resolution fr, and each time window partially overlaps with preceding and following time windows for establishing the time continuity. The frequency analyzer
103
calculates the number of allocation bits in a time period excluding the portions overlapping with the preceding and following time windows using a known psychoacoustic masking technique in a psychoacoustic mode in each of the k subbands derived by the band splitter
101
and outputs bit allocation information
1104
to the encoding circuit
104
. The time period excluding the overlapping portions corresponds to a unit time length of frames.
The encoding circuit
104
determines a scale factor of each of the k subbands based on a maximum amplitude of one of the subband signals
1102
per unit frame length, normalizes the amplitude of each of the subband signals
1102
based on a corresponding one of the scale factors, re-quantizes it based on the bit allocation information
1104
, produces a bit stream based on the re-quantized samples, the bit allocation information
1104
, the scale factors, and frame synchronization information, and outputs the bit stream as an encoded output signal
1105
.
FIG. 47
shows a conventional MPEG1 audio layer decoding system.
The signal
1106
encoded by an encoding system such as the one shown in
FIG. 46
is inputted to the frame analyzer
105
. The frame analyzer
105
extracts from the signal
1106
a frame, the bit allocation information, and the scale factors and provides frame analyzed information
1107
to the decoding circuit
106
.
The decoding circuit
106
performs a decoding operation in each subband using the frame-analyzed information
1107
to produce subband signals
1108
. The subband signals
1108
are combined in the band combining circuit
107
and outputted as a decoded output signal
1109
. In order to decrease the deterioration of information with the encoding and decoding operations to reconstruct an input signal perfectly, the band combining circuit
107
needs to meet the perfect reconstruction requirements in relation to the band splitter
101
of the encoding system. To this end, a technique using QMFs is known in the art.
However, the conventional subband encoding used in the MPEG, as described in
FIG. 46
, performs the scale factor information producing operation, the bit allocation information producing operation, and the re-quantizing operation in each of the k subbands to construct a frame, thus encountering problems of increases in load of encoding operation and bit rate.
The subband encoding is required to perform the time-frequency conversion to analyze signals in a frequency domain for establishing the information compression based on the psychoacoustic model. The realization of high-efficiency compression without any deterioration of information requires keeping the frequency resolution completely. This also requires, when the frequency conversion is performed, performing a window function on a sample for an extended period of time.
Between the subband encoding and decoding operations, the length of a frame is determined based on the number of samples required for the windowing operation. This frame length is defined as a base unit to perform the encoding, decoding, and buffering operations. The time required for each of the operations equivalent to the frame length and the time delay by a subband split filter bank will lead to problems of increase in operation time delay with increases in sound quality and compression rate.
The subband encoding also has a problem of increase in total operation resulting from the frequency analyzing and bit allocation operations.
When the subband encoding is employed in radio transmission, the synchronization acquisition of clocks in a receiver system and synchronization of radio frame require production and detection of synchronization words. The decreasing errors occurring in a transmission path requires an additional error correcting operation, which will result in an increase in delay time in overall operations of the system resulting from buffering in each operation. The additional error correcting operation is usually performed regardless of characteristics of information produced in the subband encoding, thus resulting in an fatal error in each application even if a burst error and a bit error rate, as viewed in a long time unit, are not great.
SUMMARY OF THE INVENTION
It is therefore a principal object of the present invention to avoid the disadvantages of the prior art.
It is another object of the present invention to provide a subband encoding and a decoding system capable of decreasing a operation load and an encoding bit rate.
According to one aspect of the invention, there is provided a subband encoding apparatus which comprises: (a) a subband splitter dividing an input signal in a frequency band into subband signals; (b) a first scale factor information producing circuit measuring signal levels of the subband signals to determine scale factors and producing scale factor information indicative thereof; (c) a bit allocation information producing circuit producing bit allocation information based on the scale factor information; (d) a second scale factor information producing circuit producing scale factor flag information indicating the fact that the scale factor information has changed from that one frame earlier and updated scale factor information indicating the scale factor information which has changed from that one frame earlier; (e) a re-quantizing circuit re-quantizing the subband signals using the scale factor information and the bit allocation information to provide re-quantized output signals; (f) a frame constructing circuit constructing a frame made up of the re-quantized signal, the updated scale factor information, and the scale factor flag information and outputs the frame as an encoded output signal; and (g) a subband-limiting circuit limiting the number of subbands of the subband signals to be re-quantized by the re-quantizing circuit based on an upper limit frequency of an audible band.
According to the second aspect of the invention, there is provided a subband encoding apparatus for radio transmission which comprises: (a) a subband splitter dividing an input signal
Kelley Chris
Matsushita Electric - Industrial Co., Ltd.
Wong Allen
LandOfFree
Subband encoding and decoding system for data compression... does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Subband encoding and decoding system for data compression..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Subband encoding and decoding system for data compression... will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-3290999