Data processing: speech signal processing – linguistics – language – Speech signal processing – For storage or transmission
Reexamination Certificate
2000-10-26
2004-01-06
Chawan, Vijay (Department: 2654)
Data processing: speech signal processing, linguistics, language
Speech signal processing
For storage or transmission
C704S201000, C704S278000, C704S209000, C704S205000, C381S097000, C381S098000, C381S101000, C381S103000, C084S622000, C084S626000, C084S627000
Reexamination Certificate
active
06675141
ABSTRACT:
BACKGROUND OF THE INVENTION
The present invention relates to an apparatus and method which convert the acoustic-signal reproducing speed by processing a small amount of data.
Various techniques of converting the speed of reproducing digital PCM acoustic signals from a given recording medium are known. Of these techniques, a method such as redundant addition wherein the motion of a pointer is controlled, PICOLA (Pointer Interval Controlled Over Lap and Add) is generally utilized.
A reproducing-speed converting apparatus will be described, which generates R-fold acoustic signals from source acoustic signals by means of redundant addition achieved by controlling the motion of a pointer (PICOLA system). R is a constant that represents the rate of converting the reproducing speed. R is greater than one (R>1) in the case of high-speed reproduction of acoustic signals. R is equal to or less than one in the case of low-speed reproduction of acoustic signals.
FIG. 1
is a block diagram showing the reproducing-speed converting apparatus.
The reproducing-speed converting apparatus comprises a data-recording section
1
, an input buffer section
2
, a pitch-calculating section
3
, a process control section
4
, a data-operating section
5
, and a data-accumulating section
6
. The data-recording section
1
records acoustic signals and holds the same. The input buffer section
2
receives an input acoustic signal s
1
from the data-recording section
1
. The signal s
1
(sampled for max. pitch cycle×2) has been generated from a process-start position P. The input buffer section
2
transfers an acoustic signal s
2
for finding a pitch, to the pitch-calculating section
3
. The pitch-calculating section
3
calculates a pitch cycle s
3
, which is supplied to the process control
4
. Under the control of the process control section
4
, the input buffer section
2
transfers a signal s
4
, to the data-operating section
5
. The data-operating section
5
performs a prescribed process on the signal s
4
to achieve high-speed reproduction or low-speed reproduction, thereby generating an operation process signal s
5
. The signal s
5
is supplied via the input buffer section
2
to the data-accumulating section
6
. In the meantime, the process control section
4
supplies a process control signal s
6
to the input buffer section
2
. Further, the process control section
4
supplies a data-read control signal s
7
to the data-recording section
1
.
How the conventional reproducing-speed converting apparatus, which is a PICOLA system, accomplishes high-speed reproduction and how-speed reproduction will be described below.
The high-speed reproduction will be first explained, with reference to
FIGS. 2
to
4
. First, of the acoustic signals held in the data-recording section
1
, an input acoustic signal s
1
(sampled for max. pitch cycle×2) is read from the process-start position P shown in
FIG. 2
, into the input buffer section
2
. The signal s
1
is transferred from the section
2
to the pitch-calculating section
3
.
The pitch-calculating section
3
calculates a pitch cycle s
3
. More specifically, the section
3
generates a pitch cycle s
3
(T
0
) that mininimizes the mean distortion d (T) defined by the following equation (1):
(Equation 1)
d
(
T
)
=
1
T
∑
i
=
0
T
-
1
{
x
(
i
)
-
x
(
i
+
T
)
}
2
,
T
min
≤
T
≤
T
max
(1)
The input buffer
2
transfers an acoustic signal, or a signal s
4
, to the data-operating section
5
. The signal s
4
is based on the pitch cycle s
3
(T
0
) the pitch-calculating section
3
has calculated in accordance with the equation (1). The signal s
4
lasts for
2
pitch cycles from the process-start position P.
The acoustic signal s
4
lasting for 2 pitch cycles (2×T
0
), read into the data-operating section
5
, is subjected to weight-adding process that is performed in accordance with the weight-window data shown in FIG.
3
. The section
5
generates a weight-added signal, or an operation process signal a
5
that lasts for 1 pitch cycle (T
0
sample).
Then, the process control section
4
calculates a length L of a reproduced signal (T
0
sample), in accordance with the rate R (R>1) of converting the reproducing speed. The length L is defined by the following equation (2):
(Equation 2)
L
=
T
0
×
1
R
-
1
(2)
The reproduction-signal length L calculated in accordance with the equation (2) may be longer than the pitch cycle T
0
(1<R<2). In this case, the acoustic signal (i.e., operation process signal s
5
) generated by the data-operating section
5
and lasting for one pitch cycle (i.e., T
0
sample) is transferred to the data-accumulating section
6
. Moreover, other input acoustic signals are transferred from the input buffer section
2
to the data-accumulating section
6
, so that all samples transferred to the data-accumulating section
6
defined the reproduction-signal length L.
The length defined by the input acoustic signals read into the input buffer section
2
may be shorter than the reproduction-signal length L. If so, other acoustic signals are read from the data-recording section
1
into the input buffer section
2
in accordance with a data-read control signal s
7
supplied from the process control section
4
. These signals, which are required to make the length equal to the reproducing-signal length L, are directly transferred to the data-accumulating section The reproducing-signal length L may be shorter than the pitch cycle T
0
(R>2) as is illustrated in FIG.
4
. In this case, the acoustic signals, which are L samples included in T
0
samples that define one pitch cycle calculated by the data-operating section
5
, are transferred to the data-accumulating section
6
.
The next process-start position P′ in the data-recording section
1
is updated in accordance with the following equation (3):
(Equation 3)
P
′
=
P
+
T
0
×
R
R
-
1
(3)
The low-speed reproduction will be now explained, with reference to
FIGS. 5
to
7
. First, of the acoustic signals held in the data-recording section
1
, an input acoustic signal s
1
(sampled for max. pitch cycle×2) is read from the process-start position P shown in
FIG. 5
, into the input buffer section
2
. The signal s
1
is transferred from the section
2
to the pitch-calculating section
3
. The pitch-calculating section
3
calculates a pitch cycle s
3
.
The input buffer
2
transfers an acoustic signal, or a signal s
4
, to the data-operating section
5
. The signal s
4
is based on the pitch cycle s
3
(T
0
) the pitch-calculating section
3
has calculated. The signal s
4
lasts for 2 pitch cycles from the process-start position P.
The acoustic signal s
4
lasting for 2 pitch cycles, read into the data-operating section
5
, is subjected to weight-adding process that is performed in accordance with the weight-window data shown in FIG.
6
. The section
5
generates a weight-added signal, or an operation process signal a
5
that lasts for 1 pitch cycle (T
0
sample).
Next, the process control section
4
calculates a length L of a reproduced signal [sample], in accordance with the rate R (0<R<1) of converting the reproducing speed. The length L is defined by the following equation (4):
(Equation 4)
L
=
T
0
×
1
1
-
R
(4)
The reproduced-signal length L calculated in accordance with the equation (2) may be longer than two pitch cycles (2×T
0
) and, hence (0.5<R<1). If so, the acoustic signal for one pitch cycle (T
0
sample) from the first signal held in the input buffer section
2
is transferred to the data-accumulating section
6
, along with the acoustic signal (i.e., operation process signal s
5
) generated by the data-operating section
5
and lasting for one pitch cycle (i.e., T
0
sample). Moreover, other input acoustic signals are transferred from the input buffer section
2
to the data-accumulating section
6
, so that all samples transferred to the data-accumulating section
6
define the reproduced-signal length L.
The length defined
Inoue Akira
Nishiguchi Masayuki
Chawan Vijay
Maioli Jay H.
Sony Corporation
LandOfFree
Apparatus for converting reproducing speed and method of... does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Apparatus for converting reproducing speed and method of..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Apparatus for converting reproducing speed and method of... will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-3232491