 |
Claims  |
|
|
We claim:
1. An apparatus for compressing a digital input signal to provide a
compressed digital output signal, the apparatus comprising:
a means for deriving plural spectral coefficients from the digital input
signal, and for grouping the spectral coefficients into bands, each band
having a band magnitude and a band frequency; and
an adaptive bit allocation means for adaptively allocating a number of
spectrum-dependent quantizing bits among the bands to allocate to each
band a number of spectrum-dependent quantizing bits for quantizing each
spectral coefficient in the band, the number of spectrum-dependent
quantizing bits allocated to each band being determined according to the
band magnitude, weighted depending on the band frequency.
2. The apparatus of claim 1, wherein the adaptive bit allocation means
allocates to each band a number of spectrum-dependent quantizing bits
determined according to the band magnitude, the band magnitude being
weighted depending on the band frequency such that the number of
spectrum-dependent quantizing bits allocated to lower frequency bands for
a given band magnitude is increased.
3. The apparatus of claim 1, wherein:
the apparatus additionally comprises:
a means for determining, for each band, the band magnitude in response to
the spectral coefficients in the band, and
a weighting means for weighting, for each band, the band magnitude
depending on the band frequency to provide a weighted band magnitude; and
the adaptive bit allocation means allocates to each band the number of
spectrum-dependent quantizing bits for quantizing each spectral
coefficient in the band in response to the weighted band magnitude for the
band.
4. The apparatus of claim 1, wherein the adaptive bit allocation means
allocates to each band a number of spectrum-dependent quantizing bits
determined according to the band magnitude weighted by a weighting
coefficient dependent on the digital input signal and the band frequency.
5. The apparatus of claim 1, wherein:
the adaptive bit allocation means allocates to each band a number of
spectrum-dependent quantizing bits determined according to the band
magnitude weighted depending on the band frequency according to a selected
frequency-dependent weighting pattern; and
the apparatus additionally comprises:
a means for storing plural frequency-dependent weighting patterns, and
a selecting means for selecting, in response to the digital input signal,
and for providing to the adaptive bit allocation means as the selected
frequency-dependent weighting pattern, an appropriate one of the plural
frequency-dependent weighting patterns.
6. The apparatus of claim 5, wherein:
the digital input signal has a level; and
the selecting means is for selecting the appropriate one of the plural
frequency-dependent weighting patterns in response to the level of the
digital input signal.
7. The apparatus of claim 5, wherein:
the digital input signal has a spectrum; and
the selecting means is for selecting the appropriate one of the plural
frequency-dependent weighting patterns in response to the spectrum of the
digital input signal.
8. The apparatus of claim 5, wherein the adaptive bit allocation means
allocates to each band a number of spectrum-dependent quantizing bits
determined according to the band magnitude, the band magnitude being
weighted depending on the band frequency according to the selected
frequency-dependent weighting pattern such that the number of
spectrum-dependent quantizing bits allocated to lower frequency bands for
a given band magnitude is increased.
9. The apparatus of claim 1, additionally comprising:
a level-dependent bit allocation means for allocating a number of
level-dependent quantizing bits among the bands according to a
predetermined bit allocation pattern defining, for each band, a number of
level-dependent quantizing bits allocated for quantizing each spectral
coefficient in the band; and
a quantizing means for quantizing each spectral coefficient in a band using
a number of quantizing bits equal to the sum of the number of
spectrum-dependent quantizing bits allocated for quantizing each spectral
coefficient in the band and the number of level-dependent quantizing bits
allocated for quantizing each spectral coefficient in the band.
10. The apparatus of claim 9, wherein the adaptive bit allocation means
allocates to each band a number of spectrum-dependent quantizing bits
determined according to the band magnitude, the band magnitude being
weighted depending on the band frequency such that the number of
spectrum-dependent quantizing bits allocated to lower frequency bands for
a given band magnitude is increased.
11. The apparatus of claim 9, wherein:
the apparatus additionally comprises:
a means for determining, for each band, the band magnitude in response to
the spectral coefficients in the band, and
a weighting means for weighting, for each band, the band magnitude
depending on the band frequency to provide a weighted band magnitude; and
the adaptive bit allocation means allocates to each band the number of
spectrum-dependent quantizing bits for quantizing each spectral
coefficient in the band in response to the weighted band magnitude.
12. The apparatus of claim 9, wherein the adaptive bit allocation means
allocates to each band a number of spectrum-dependent quantizing bits
determined according to the band magnitude weighted by a weighting
coefficient dependent on the digital input signal and the band frequency.
13. The apparatus of claim 9, wherein:
the adaptive bit allocation means allocates to each band a number of
spectrum-dependent quantizing bits determined according to the band
magnitude weighted depending on the band frequency according to a selected
frequency-dependent weighting pattern; and
the apparatus additionally comprises:
a means for storing plural frequency-dependent weighting patterns, and
a selecting means for selecting, in response to the digital input signal,
and for providing to the adaptive bit allocation means as the selected
frequency-dependent weighting pattern, an appropriate one of the plural
frequency-dependent weighting patterns.
14. The apparatus of claim 13, wherein the adaptive bit allocation means
allocates to each band a number of spectrum-dependent quantizing bits
determined according to the band magnitude weighted depending on the band
frequency such that the number of spectrum-dependent quantizing bits
allocated to lower frequency bands for a given band magnitude is
increased.
15. The apparatus of claim 9, wherein:
the level-dependent bit allocation means is for allocating level-dependent
quantizing bits among the bands according to a selected bit allocation
pattern; and
the apparatus additionally comprises:
a means for storing plural predetermined bit allocation patterns, and
an additional selecting means for selecting, in response to the digital
input signal, and for providing to the level-dependent bit allocating
means as the selected bit allocation pattern, an appropriate one of the
plural predetermined bit allocation patterns.
16. The apparatus of claim 9, wherein:
a total number of available quantizing bits is available for quantizing all
the spectral coefficients;
the digital input signal has a spectrum, and the spectrum has a smoothness;
the apparatus additionally comprises:
a division ratio determining means for determining a division of the total
number of available quantizing bits between the number of
spectrum-dependent quantizing bits and the number of level-dependent
quantizing bits in response to the smoothness of the spectrum of the
digital input signal, and for providing a division ratio between the
number of level-dependent quantizing bits and the total number of
available quantizing bits, and a complement of the division ratio,
a means for multiplying the number of spectrum-dependent quantizing bits
for each band by the complement of the division ratio, and
a means for multiplying the number of level-dependent quantizing bits for
each band by the division ratio; and
the quantizing means is for quantizing each spectral coefficient in a band
using a number of quantizing bits equal to the sum of the multiplied
number of spectrum-dependent quantizing bits allocated for quantizing each
spectral coefficient in the band and the multiplied number of
level-dependent quantizing bits allocated for quantizing each spectral
coefficient in the band.
17. The apparatus of claim 9, wherein:
a total number of available quantizing bits is available for quantizing all
the spectral coefficients;
the digital input signal has a spectrum, and the spectrum has a smoothness;
the apparatus additionally comprises a division ratio determining means for
determining a division of the total number of available quantizing bits
between the number of spectrum-dependent quantizing bits and the number of
level-dependent quantizing bits in response to the smoothness of the
spectrum of the digital input signal, and for providing a division ratio
between the number of level-dependent quantizing bits and the total number
of available quantizing bits, and a complement of the division ratio;
the number of spectrum-dependent quantizing bits allocated among the bands
by the adaptive bit allocation means is substantially proportional to the
complement of the division ratio; and
the number of level-dependent quantizing bits allocated among the bands by
the fixed bit allocation means is substantially proportional to the
division ratio.
18. The apparatus of claim 17, wherein:
the level-dependent bit allocation means is for allocating level-dependent
quantizing bits among the bands according to a selected bit allocation
pattern; and
the apparatus additionally comprises:
a means for storing plural predetermined bit allocation patterns, the
predetermined bit allocation patterns allocating numbers of
level-dependent quantizing bits corresponding to different values of the
division ratio, and
an additional selecting means for selecting, in response to the digital
input signal, and in response to the division ratio, and for providing to
the level-dependent bit allocating means as the selected bit allocation
pattern, an appropriate one of the plural predetermined bit allocation
patterns.
19. The apparatus of claim 1, wherein the means for deriving plural
spectral coefficients comprises:
a frequency dividing means for receiving the digital input signal and for
dividing the digital input signal into plural frequency ranges, and for
providing a frequency range signal in each frequency range;
a time dividing means for dividing in time a frequency range signal into
blocks; and
an orthogonal transform means for transforming a block of the frequency
range signal to provide the plural spectral coefficients.
20. A method for compressing a digital input signal to provide a compressed
digital output signal, the method comprising the steps of:
deriving plural spectral coefficients from the digital input signal, and
grouping the spectral coefficients into bands, each band having a band
magnitude and a band frequency; and
adaptively allocating a number of spectrum-dependent quantizing bits among
the bands to allocate to each band a number of spectrum-dependent
quantizing bits for quantizing each spectral coefficient in the band, the
number of spectrum-dependent quantizing bits allocated to each band being
determined according to the band magnitude, weighted depending on the band
frequency.
21. The method of claim 20, wherein, in the allocating step, the band
magnitude is weighted depending on the band frequency such that the number
of spectrum-dependent quantizing bits allocated to lower frequency bands
for a given band magnitude is increased.
22. The method of claim 20, wherein:
the method additionally comprises the steps of:
determining, for each band, the band magnitude in response to the spectral
coefficients in the band, and
weighting, for each band, the band magnitude depending on the band
frequency to provide a weighted band magnitude; and
in the allocating step, the number of spectrum-dependent quantizing bits
for quantizing each spectral coefficient in the band is allocated to each
band in response to the weighted band magnitude for the band.
23. The method of claim 20, wherein, in the allocating step, the band
magnitude for each band is weighted by a weighting coefficient dependent
on the digital input signal and the band frequency.
24. The method of claim 20, wherein:
in the allocating step, the band magnitude for each band is weighted
depending on the band frequency according to a selected
frequency-dependent weighting pattern; and
the method additionally comprises the steps of:
providing plural frequency-dependent weighting patterns, and
selecting in response to the digital input signal, and providing as the
selected frequency-dependent weighting pattern, an appropriate one of the
plural frequency-dependent weighting patterns.
25. The method of claim 24, for compressing a digital input signal having a
level, wherein, in the selecting step, the appropriate one of the plural
frequency-dependent weighting patterns is selected in response to the
level of the digital input signal.
26. The method of claim 24, for compressing a digital input signal having a
spectrum, wherein, in the selecting step, the appropriate one of the
plural frequency-dependent weighting patterns is selected in response to
the spectrum of the digital input signal.
27. The method of claim 24, wherein, in the allocating step, the band
magnitude is weighted such that the number of spectrum-dependent
quantizing bits allocated to lower frequency bands for a given band
magnitude is increased.
28. The method of claim 20, additionally comprising the steps of:
allocating a number of level-dependent quantizing bits among the bands
according to a predetermined bit allocation pattern defining, for each
band, a number of level-dependent quantizing bits allocated for quantizing
each spectral coefficient in the band; and
quantizing each spectral coefficient in a band using a number of quantizing
bits equal to the sum of the number of spectrum-dependent quantizing bits
allocated for quantizing each spectral coefficient in the band and the
number of level-dependent quantizing bits allocated for quantizing each
spectral coefficient in the band.
29. The method of claim 28, wherein in the step of allocating a number of
spectrum-dependent quantizing bits among the bands, the band magnitude is
weighted such that the number of spectrum-dependent quantizing bits
allocated to lower frequency bands for a given band magnitude is
increased.
30. The method of claim 28, wherein:
the method additionally comprises the steps of:
determining, for each band, the band magnitude in response to the spectral
coefficients in the band, and
weighting, for each band, the band magnitude depending on the band
frequency to provide a weighted band magnitude; and
in the step of allocating a number of spectrum-dependent quantizing bits
among the bands, the number of spectrum-dependent quantizing bits for
quantizing each spectral coefficient in the band is allocated to each band
in response to the weighted band magnitude.
31. The method of claim 28, wherein, in the step of allocating a number of
spectrum-dependent quantizing bits among the bands, the band magnitude for
each band is weighted by a weighting coefficient dependent on the digital
input signal and the band frequency.
32. The method of claim 28, wherein:
in the step of allocating a number of spectrum-dependent quantizing bits
among the bands, the band magnitude is weighted depending on the band
frequency according to a selected frequency-dependent weighting pattern;
and
the method additionally comprises the steps of:
providing plural frequency-dependent weighting patterns, and
selecting in response to the digital input signal, and providing as the
selected frequency-dependent weighting pattern, an appropriate one of the
plural frequency-dependent weighting patterns.
33. The method of claim 32, wherein, in the step of allocating a number of
spectrum-dependent quantizing bits among the bands, the band magnitude is
weighted such that the number of spectrum-dependent quantizing bits
allocated to lower frequency bands for a given band magnitude is
increased.
34. The method of claim 28, wherein:
in the step of allocating a number of level-dependent quantizing bits among
the bands, level-dependent quantizing bits are allocated among the bands
according to a selected bit allocation pattern; and
the method additionally comprises the steps of:
providing plural predetermined bit allocation patterns, and
selecting in response to the digital input signal, and providing as the
selected bit allocation pattern, an appropriate one of the plural
predetermined bit allocation patterns.
35. The method of claim 28, for compressing a digital input signal having a
spectrum, the spectrum having a smoothness, and using a total number of
available quantizing bits for quantizing all the spectral coefficients,
wherein:
the method additionally comprises:
determining a division of the total number of available quantizing bits
between the number of spectrum-dependent quantizing bits and the number of
level-dependent quantizing bits in response to the smoothness of the
spectrum of the digital input signal,
providing in response to the determined division a division ratio between
the number of level-dependent quantizing bits and the total number of
available quantizing bits, and a complement of the division ratio,
multiplying the number of spectrum-dependent quantizing bits for each band
by the complement of the division ratio, and
multiplying the number of level-dependent quantizing bits for each band by
the division ratio; and
in the quantizing step, each spectral coefficient in a band is quantized
using a number of quantizing bits equal to the sum of the multiplied
number of spectrum-dependent quantizing bits allocated for quantizing each
spectral coefficient in the band and the multiplied number of
level-dependent quantizing bits allocated for quantizing each spectral
coefficient in the band.
36. The method of claim 28, for compressing a digital input signal having a
spectrum, the spectrum having a smoothness, and using a total number of
available quantizing bits for quantizing all the spectral coefficients,
wherein:
the method additionally comprises:
determining a division of the total number of available quantizing bits
between the number of spectrum-dependent quantizing bits and the number of
level-dependent quantizing bits in response to the smoothness of the
spectrum of the digital input signal, and
providing in response to the determined division a division ratio between
the number of level-dependent quantizing bits and the total number of
available quantizing bits, and a complement of the division ratio;
in the step of allocating a number of spectrum-dependent quantizing bits
among the bands, the number of spectrum-dependent quantizing bits
allocated among the bands is substantially proportional to the complement
of the division ratio; and
in the step of allocating a number of level-dependent quantizing bits among
the bands, the number of level-dependent quantizing bits allocated among
the bands is substantially proportional to the division ratio.
37. The method of claim 36, wherein:
in the step of allocating a number of level-dependent quantizing bits among
the bands, level-dependent quantizing bits are allocated among the bands
according to a selected bit allocation pattern; and
the method additionally comprises the steps of:
providing plural predetermined bit allocation patterns, the predetermined
bit allocation patterns allocating numbers of level-dependent quantizing
bits corresponding to different values of the division ratio, and
selecting in response to the digital input signal, and in response to the
division ratio, and providing as the selected bit allocation pattern, an
appropriate one of the plural predetermined bit allocation patterns.
38. The method of claim 20, wherein the step of deriving plural spectral
coefficients from the digital input signal comprises the steps of:
dividing the digital input signal into plural frequency ranges, and
providing a frequency range signal in each frequency range;
dividing a frequency range signal in time into blocks; and
orthogonally transforming a block of the frequency range signal to provide
the plural spectral coefficients.
39. A medium whereon is stored a compressed digital signal derived from a
digital input signal by a data compressing method comprising the steps of:
deriving plural spectral coefficients from the digital input signal, and
grouping the spectral coefficients into bands, each band having a band
magnitude and a band frequency;
adaptively allocating a number of spectrum-dependent quantizing bits among
the bands to allocate to each band a number of spectrum-dependent
quantizing bits for quantizing each spectral coefficient in the band, the
number of spectrum-dependent quantizing bits allocated to each band being
determined according to the band magnitude, weighted depending on the band
frequency;
quantizing each spectral coefficient in each band using a number of
quantizing bits including the number of spectrum-dependent quantizing bits
allocated for quantizing each spectral coefficient in the band; and
including the quantized spectral coefficients in the compressed digital
signal.
40. The medium of claim 39, wherein, in the allocating step of the data
compressing method, the band magnitude is weighted depending on the band
frequency such that the number of spectrum-dependent quantizing bits
allocated to lower frequency bands for a given band magnitude is
increased.
41. The medium of claim 39, wherein, in the data compressing method:
the method additionally comprises the steps of:
determining, for each band, the band magnitude in response to the spectral
coefficients in the band, and
weighting, for each band, the band magnitude depending on the band
frequency to provide a weighted band magnitude; and
in the allocating step, the number of spectrum-dependent quantizing bits
for quantizing each spectral coefficient in the band is allocated to each
band in response to the weighted band magnitude for the band.
42. The medium of claim 39, wherein, in the allocating step of the data
compressing method, the band magnitude for each band is weighted by a
weighting coefficient dependent on the digital input signal and the band
frequency.
43. The medium of claim 39, wherein, in the data compressing method:
in the allocating step, the band magnitude for each band is weighted
depending on the band frequency according to a selected
frequency-dependent weighting pattern; and
the method additionally comprises the steps of:
providing plural frequency-dependent weighting patterns, and
selecting in response to the digital input signal, and providing as the
selected frequency-dependent weighting pattern, an appropriate one of the
plural frequency-dependent weighting patterns.
44. The medium of claim 39, wherein, in the data compressing method:
the method additionally comprises the step of allocating a number of
level-dependent quantizing bits among the bands according to a
predetermined bit allocation pattern defining, for each band, a number of
level-dependent quantizing bits allocated for quantizing each spectral
coefficient in the band; and
in the quantizing step, each spectral coefficient in each band is quantized
using a number of quantizing bits equal to the sum of the number of
spectrum-dependent quantizing bits allocated for quantizing each spectral
coefficient in the band and the number of level-dependent quantizing bits
allocated for quantizing each spectral coefficient in the band.
45. The medium of claim 44, wherein in the step in the data compressing
method of allocating a number of spectrum-dependent quantizing bits among
the bands, the band magnitude is weighted such that the number of
spectrum-dependent quantizing bits allocated to lower frequency bands for
a given band magnitude is increased.
46. The medium of claim 44, wherein, in the data compressing method:
the method additionally comprises the steps of:
determining, for each band, the band magnitude in response to the spectral
coefficients in the band, and
weighting, for each band, the band magnitude depending on the band
frequency to provide a weighted band magnitude; and
in the step of allocating a number of spectrum-dependent quantizing bits
among the bands, the number of spectrum-dependent quantizing bits for
quantizing each spectral coefficient in the band is allocated to each band
in response to the weighted band magnitude.
47. The medium of claim 44, wherein, in the step in the data compressing
method of allocating a number of spectrum-dependent quantizing bits among
the bands, the band magnitude for each band is weighted by a weighting
coefficient dependent on the digital input signal and the band frequency.
48. The medium of claim 44, wherein, in the data compressing method:
in the step of allocating a number of spectrum-dependent quantizing bits
among the bands, the band magnitude is weighted depending on the band
frequency according to a selected frequency-dependent weighting pattern;
and
the method additionally comprises the steps of:
providing plural frequency-dependent weighting patterns, and
selecting in response to the digital input signal, and providing as the
selected frequency-dependent weighting pattern, an appropriate one of the
plural frequency-dependent weighting patterns.
49. The medium of claim 44, wherein, in the data compressing method:
in the step of allocating a number of level-dependent quantizing bits among
the bands, level-dependent quantizing bits are allocated among the bands
according to a selected bit allocation pattern; and
the method additionally comprises the steps of:
providing plural predetermined bit allocation patterns, and
selecting in response to the digital input signal, and providing as the
selected bit allocation pattern, an appropriate one of the plural
predetermined bit allocation patterns.
50. The medium of claim 44, wherein the data compressing method is for
compressing a digital input signal having a spectrum, the spectrum having
a smoothness, and the method uses a total number of available quantizing
bits for quantizing all the spectral coefficients, wherein:
the method additionally comprises:
determining a division of the total number of available quantizing bits
between the number of spectrum-dependent quantizing bits and the number of
level-dependent quantizing bits in response to the smoothness of the
spectrum of the digital input signal,
providing in response to the determined division a division ratio between
the number of level-dependent quantizing bits and the total number of
available quantizing bits, and a complement of the division ratio,
multiplying the number of spectrum-dependent quantizing bits for each band
by the complement of the division ratio, and
multiplying the number of level-dependent quantizing bits for each band by
the division ratio; and
in the quantizing step, each spectral coefficient in a band is quantized
using a number of quantizing bits equal to the sum of the multiplied
number of spectrum-dependent quantizing bits allocated for quantizing each
spectral coefficient in the band and the multiplied number of
level-dependent quantizing bits allocated for quantizing each spectral
coefficient in the band.
51. The medium of claim 39, wherein, in the data compressing method, the
step of deriving plural spectral coefficients from the digital input
signal comprises the steps of:
dividing the digital input signal into plural frequency ranges, and
providing a frequency range signal in each frequency range;
dividing a frequency range signal in time into blocks; and
orthogonally transforming a block of the frequency range signal to provide
the plural spectral coefficients.
52. The medium of claim 39, wherein the medium is selected from a group
including an optical disc, a semiconductor memory, and a magnetic tape.
53. An expander for expanding a compressed digital signal derived from a
digital input signal by a data compressing method comprising the steps of:
deriving plural spectral coefficients from the digital input signal, and
grouping the spectral coefficients into bands, each band having a band
magnitude and a band frequency,
adaptively allocating a number of spectrum-dependent quantizing bits among
the bands to allocate to each band a number of spectrum-dependent
quantizing bits for quantizing each spectral coefficient in the band, the
number of spectrum-dependent quantizing bits allocated to each band being
determined according to the band magnitude, weighted depending on the band
frequency,
quantizing each spectral coefficient in each band using a number of
quantizing bits including the number of spectrum-dependent quantizing bits
allocated for quantizing each spectral coefficient in the band, and
including in the compressed digital signal, for each band, the quantized
spectral coefficients and data indicating the number of quantizing bits,
the expander including:
a means for separating from the compressed digital signal, for each band,
the quantized spectral coefficients and the data indicating the number of
quantizing bits;
a means for dequantizing, for each band, the quantized spectral
coefficients in response to the data indicating the number of quantizing
bits; and
a means for deriving the output signal from the dequantized spectral
coefficients.
54. The expander of claim 53, wherein, in the allocating step of the data
compressing method, the band magnitude is weighted depending on the band
frequency such that the number of spectrum-dependent quantizing bits
allocated to lower frequency bands for a given band magnitude is
increased.
55. The expander of claim 53, wherein, in the data compressing method:
the method additionally comprises the steps of:
determining, for each band, the band magnitude in response to the spectral
coefficients in the band, and
weighting, for each band, the band magnitude depending on the band
frequency to provide a weighted band magnitude; and
in the allocating step, the number of spectrum-dependent quantizing bits
for quantizing each spectral coefficient in the band is allocated to each
band in response to the weighted band magnitude for the band.
56. The expander of claim 53, wherein, in the allocating step of the data
compressing method, the band magnitude for each band is weighted by a
weighting coefficient dependent on the digital input signal and the band
frequency.
57. The expander of claim 53, wherein, in the data compressing method:
in the allocating step, the band magnitude for each band is weighted
depending on the band frequency according to a selected
frequency-dependent weighting pattern; and
the method additionally comprises the steps of:
providing plural frequency-dependent weighting patterns, and
selecting in response to the digital input signal, and providing as the
selected frequency-dependent weighting pattern, an appropriate one of the
plural frequency-dependent weighting patterns.
58. The expander of claim 53, wherein, in the data compressing method:
the method additionally comprises the step of allocating a number of
level-dependent quantizing bits among the bands according to a
predetermined bit allocation pattern defining, for each band, a number of
level-dependent quantizing bits allocated for quantizing each spectral
coefficient in the band; and
in the quantizing step, each spectral coefficient in each band is quantized
using a number of quantizing bits equal to the sum of the number of
spectrum-dependent quantizing bits allocated for quantizing each spectral
coefficient in the band and the number of level-dependent quantizing bits
allocated for quantizing each spectral coefficient in the band.
59. The expander of claim 58, wherein in the step in the data compressing
method of allocating a number of spectrum-dependent quantizing bits among
the bands, the band magnitude is weighted such that the number of
spectrum-dependent quantizing bits allocated to lower frequency bands for
a given band ma | | |
