Note: Descriptions are shown in the official language in which they were submitted.
- ~r ~oi~ -~
CA 02376816 2001-12-10
Specification
Speech Switching Apparatus
Technical Field
The present invention relates to a speech
encoding/decoding apparatus and, more particularly, to a
speech switching apparatus for switching one of a
plurality of speech signals.
Background Art
Conventionally, speech is transmitted on a
transmission path on which the bit rate changes by using
an encoding method of adjusting the quality of a
reconstructed speech signal by adapting an encoding bit
rate to the transmission path bit rate by
increasing/decreasing the bandwidth of the speech signal
in accordance with the transmission path bit rate. The
present inventor has already proposed a speech
encoding/decoding apparatus in Japanese Patent Laid-Open
No. 9-202475, as a speech encoding apparatus for
generating N + 1 signals by changing the sampling
frequency of an input speech signal, in hierarchically
encoding the speech signal, and simultaneously
multiplexing N-level indexes representing linear
predictive coefficients, pitches, multipath signals, and
gains which are obtained by sequentially encoding the
input speech signal and the signals obtained by changing
the sampling frequency in increasing order of sampling
- 1 -
' CA 02376816 2001-12-10
frequency, and a speech decoding apparatus for
hierarchically changing the sampling frequency of a
reconstructed signal in accordance with the bit rate at
which decoding is performed. In this apparatus, a first
CELP (Code Excited Linear Prediction) encoding circuit
for receiving the signal obtained by down-sampling an
input signal using a down-sampling circuit outputs an
encoded output to a second CELP encoding circuit, the
second CELP encoding circuit encodes the input signal on
the basis of the encoded output from the first CELP
encoding circuit, a multiplexer outputs the encoded
outputs from the first and second CELP encoding circuits
in the form of a bit stream, a demultiplexer outputs the
encoded output obtained by the first CELP encoding
circuit from the bit stream to a first CELP decoding
circuit when a control signal has a low bit rate, and
extracts part of the output obtained by the first CELP
encoding circuit and the output obtained by the second
CELP encoding circuit from the bit stream, when the
control signal has a high bit rate, to output them to a
second CELP decoding circuit so as to output them
through a switching circuit.
On the decoding side, the bandwidth of a
reconstructed speech signal, i.e., the sampling
frequency of a decoded speech signal, changes in
accordance with the bit rate at the time of reception.
When a user is to hear a sampled speech signal, a
- 2 -
CA 02376816 2001-12-10
sampling frequency must be set for conversion processing
from a digital signal to an analog signal. In this case,
in order to switch and reconstruct speech signals having
different sampling frequencies, sampling frequencies
must be set/changed. During this sampling frequency
setting/changing processing, interruptions tend to occur
in reconstructed speech.
The operation of a conventional speech
switching apparatus will be described with reference to
Fig. 7. The speech switching apparatus receives two
types of speech signals (first and second digital speech
signals) sampled with two different sampling frequencies
(e. g., 8 kHz and 16 kHz), together with a control signal,
and switching and reconstructing the first and second
speech signals in accordance with the control signal.
In this case, the control signal is a signal
for giving an instruction to reconstruct a specific one
of the two types of speech signals.
A switching circuit 103 receives first and
second speech signals and control signal, and switches
and outputs the two types of speech signals to a D/A
conversion circuit 112 at the switching timing
designated by the control signal.
The D/A conversion circuit 112 sets the
sampling frequency of the speech signal designated by
the control signal, converts the input digital signal
into an analog signal, and outputs it.
- 3 -
CA 02376816 2001-12-10
In the above conventional speech switching
apparatus, in switching and reconstructing speech
signals having different sampling frequencies, the
sampling frequency in the D/A conversion circuit must be
set/changed. During the setting/changing processing,
interruptions occur in the reconstructed speech.
The present invention has therefore been made
in consideration of the above problems, and has as its
object to provide a speech switching apparatus which can
reduce strange sounds produced when a plurality of
different speech signals are reconstructed/switched.
Disclosure of Invention
In order to achieve the above object,
according to the present invention, there is provided a
speech switching apparatus for receiving a plurality of
input signal sampled with a plurality of different
sampling frequencies and a control signal for
designating a signal of the plurality of input signals
which is to be reconstructed, and selecting and
outputting one of the plurality of input signals in
accordance with the control signal, characterized by
comprising at least one sampling frequency conversion
circuit for converting a sampling frequency of at least
one of the plurality of input signals, a delay
adjustment circuit for adjusting a phase of the input
signal of the plurality of input signals, whose sampling
frequency is converted by the sampling frequency
- 4 -
CA 02376816 2001-12-10
conversion circuit and a phase of the remaining input
signal, and outputting the signals, and a switching
circuit for selecting one of a plurality of output
signals from the delay adjustment circuit in accordance
with the control signal.
In this case, the delay adjustment circuit may
make an adjustment to match the phase of the signal
whose sampling frequency is converted to the phase of
the remaining input signal.
The switching circuit may switch outputs at a
timing set in consideration of a delay time in the delay
adjustment circuit with respect to a switching timing
designated by the control signal.
According to the present invention, there is
provided a speech switching apparatus for receiving a
plurality of input signal sampled with a plurality of
different sampling frequencies and a control signal for
designating a signal of the plurality of input signals
which is to be reconstructed, and selecting and
outputting one of the plurality of input signals in
accordance with the control signal, characterized by
comprising a plurality of sampling frequency conversion
circuits for converting sampling frequencies of the
plurality of input signals to a predetermined frequency,
a delay adjustment circuit for adjusting phases between
output signals from the plurality of sampling frequency
conversion circuits and outputting the signals, and a
- 5 -
CA 02376816 2001-12-10
switching circuit for selecting one signal from a
plurality of output signals from the delay adjustment
circuit in accordance with the control signal.
In this case, the delay adjustment circuit may
make an adjustment to match the phase of the signal
whose sampling frequency is converted to the phase of
the remaining input signal.
In addition, the switching circuit may switch
outputs at a timing set in consideration of a delay time
in the delay adjustment circuit with respect to a
switching timing designated by the control signal.
According to the present invention, there is
provided a speech switching apparatus for receiving a
plurality of input signal sampled with a plurality of
different sampling frequencies and a control signal for
designating a signal of the plurality of input signals
which is to be reconstructed, and selecting and
outputting one of the plurality of input signals in
accordance with the control signal, characterized by
comprising at least one sampling frequency conversion
circuit for converting a sampling frequency of at least
one of the plurality of input signals, a delay
adjustment circuit for adjusting a phase of the input
signal of the plurality of input signals, whose sampling
frequency is converted by the sampling frequency
conversion circuit and a phase of the remaining input
signal, and outputting the signals, an addition circuit
- 6 -
CA 02376816 2001-12-10
for selecting and weighting two signals from a plurality
of output signals from the delay adjustment circuit in
accordance with the control signal, and a switching
circuit for selecting one signal from a plurality of
output signals from the delay adjustment circuit and an
output signal from the addition circuit in accordance
with the control signal.
In this case, the switching circuit may switch
a signal before switching of output signals from the
delay adjustment circuit to an output signal from the
addition circuit at a timing set in consideration of a
delay time in the delay adjustment circuit from a
switching timing designated by the control signal,
outputs the output signal from the addition circuit for
a predetermined interval, and then output the signal
after switching.
According to the present invention, there
provided a speech switching apparatus for receiving a
plurality of input signal sampled with a plurality of
different sampling frequencies and a control signal for
designating a signal of the plurality of input signals
which is to be reconstructed, and selecting and
outputting one of the plurality of input signals in
accordance with the control signal, characterized by
comprising a plurality of sampling frequency conversion
circuits for converting sampling frequencies of the
plurality of input signals to a predetermined frequency,
CA 02376816 2001-12-10
a delay adjustment circuit for adjusting phases between
output signals from the plurality of sampling frequency
conversion circuits and outputting the signals, an
addition circuit for selecting and weighting two signals
from a plurality of output signals from the delay
adjustment circuit in accordance with the control signal,
and a switching circuit for selecting one signal from a
plurality of output signals from the delay adjustment
circuit and an output signal from the addition circuit
in accordance with the control signal.
In this case, the switching circuit may switch
a signal before switching of output signals from the
delay adjustment circuit to an output signal from the
addition circuit at a timing set in consideration of a
delay time in the delay adjustment circuit from a
switching timing designated by the control signal,
output the output signal from the addition circuit for a
predetermined interval, and then output the signal after
switching.
In addition, the above speech switching
apparatus may further comprise a speech decoding circuit
for decoding a plurality of signals sampled from one bit
stream with different sampling frequencies, and
outputting the signals as the plurality of input signals
to the sampling frequency conversion circuit or the
delay adjustment circuit; and one signal is selected
from a plurality of output decoded signals from the
_ g _
CA 02376816 2001-12-10
speech decoding circuit in accordance with a bit rate at
the time of reception and the control signal and output.
The above speech switching apparatus may
further comprise a bit stream switching circuit for
receiving bit streams obtained by multiplexing a
plurality of bit streams in which a plurality of types
of signals having different sampling frequencies, and
switching/outputting the bit streams to a plurality of
output terminals in accordance with types of bit streams,
and a plurality of speech decoding circuits for decoding
the respective bit streams output from the bit stream
switching circuit, and outputting the bit streams as the
plurality of input signals to the sampling frequency
conversion circuit or the delay adjustment circuit, and
one signal may be selected from output decoded signals
from the plurality of speech decoding circuits in
accordance with the control signal and output.
Brief Description of Drawings
Fig. 1 is a view showing the arrangement of
the first embodiment of the present invention;
Fig. 2 is a view showing the arrangement of
the second embodiment of the present invention;
Fig. 3 is a view showing the arrangement of
the third embodiment of the present invention;
Fig. 4 is a view showing the arrangement of
the fourth embodiment of the present invention;
Fig. 5 is a view showing the arrangement of
_ g _
CA 02376816 2001-12-10
the fifth embodiment of the present invention;
Fig. 6 is a view showing the arrangement of
the sixth embodiment of the present invention; and
Fig. 7 is a view showing an example of the
arrangement of a conventional speech switching apparatus.
Best Mode of Carrying Out the Invention
The mode of carrying out the present invention
will be described below. According to the present
invention, when digital speech signals having different
sampling frequencies are to be reconstructed/switched,
in order to eliminate interruptions in reconstructed
speech, a plurality of digital speech signals having
different sampling frequencies are converted into
signals having the same sampling frequency, and the
resultant phases are adjusted, thereby reconstructing
the signals.
More specifically, the present invention
includes a sampling frequency conversion circuit (1 in
Fig. 1) for converting the sampling frequencies of
digital speech signals and a delay adjustment circuit (2
in Fig. 1) for adjusting a phase shift caused by
sampling frequency conversion between a plurality of
digital speech signals.
To eliminate discontinuity caused between
samples when digital speech signals having the same
sampling frequency and different signal bandwidths are
continuously reconstructed, the digital speech signals
- 10 -
CA 02376816 2001-12-10
before and after switching are weighted/added for a
predetermined interval first, and then are
switched/reconstructed. More specifically, the present
invention includes a sampling frequency conversion
circuit (1 in Fig. 3), a delay adjustment circuit (2 in
Fig. 3), an addition circuit (6 in Fig. 3) for
weighting/adding output signals from the delay
adjustment circuit for a predetermined interval, and a
switching circuit (7 in Fig. 3) for switching output
signals from the addition circuit in accordance with a
control signal after outputting an output signal from
the addition circuit for the interval.
According to the present invention, the
sampling frequency conversion circuit and delay
adjustment circuit make digital signals before and after
switching have the same sampling frequency and phase.
This reduces the tendency to cause interruptions in
reconstructed speech without requiring sampling
frequency setting in a D/A circuit.
In addition, according to the present
invention, the addition circuit weights/adds digital
signals before and after switching to reduce
discontinuity between the final sample of a speech
signal before switching and the first sample in the
interval as compared with a case where no
weighting/adding operation is performed. The switching
circuit performs switching after an output signal from
- 11 -
CA 02376816 2001-12-10
the addition circuit is output for a predetermined
interval. This reduces discontinuity between samples at
the start and end of the interval, and hence reduces the
tendency to produce strange sounds in reconstructed
speech.
To describe the above mode in more detail, the
embodiments of the present invention will be described
with reference to the accompanying drawings.
Fig. 1 is a block diagram showing the
arrangement of the first embodiment of the present
invention. Referring to Fig. 1, in the first embodiment
of the present invention, two types of speech signals
(first and second speech signals) having different
sampling frequencies (e.g., 8 kHz and 16 kHz) and a
control signal for instructing to reconstruct one of the
two types of speech signals are input, and the speech
signals are switched and reconstructed in accordance
with the control signal.
A sampling frequency conversion circuit 1
performs sampling frequency conversion to match the
sampling frequency of the first speech signal to the
sampling frequency of the second speech signal (e. g.,
converts the sampling frequency from 8 kHz to 16 kHz),
and outputs the resultant signal to a delay adjustment
circuit 2. In this case, the sampling frequency
conversion circuit 1 performs frequency conversion by
using a frequency-multiplying or frequency-dividing
- 12 -
CA 02376816 2001-12-10
circuit or performing interpolation or decimation
processing. This frequency conversion is performed by
using a known circuit. For its operation, refer to, for
example, P.P. Vaidyanathan, "Multirate Systems and
Filter Banks", Section 4. 1. 1 (Figure 4.1-8).
Owing to the processing performed by the
sampling frequency conversion circuit 1, the output
signal undergoes a phase delay with respect to the input
signal. Let D be the delay time in this case.
The delay adjustment circuit 2 outputs the
signal obtained by delaying the input second speech
signal by the delay time D using a delay circuit (not
shaven) and the output signal from the sampling frequency
conversion circuit 1 to a switching circuit 3. As the
delay circuit, an arbitrary circuit such as an inverter
array or delay line is used.
The switching circuit 3 receives the first
speech signal having undergone sampling frequency
conversion and the second delay signal having undergone
delay adjustment from the delay adjustment circuit 2,
switches the two types of speech signals, in
consideration of the delay time D, in accordance with
the control signal, and outputs the resultant signal to
a D/A conversion circuit 4.
The D/A conversion circuit 4 converts the
input digital speech signal into an analog signal and
outputs it. The analog signal is provided for a user
- 13 -
CA 02376816 2001-12-10
through a speaker, headphone, or the like.
Fig. 2 is a block diagram showing the
arrangement of the second embodiment of the present
invention. The second embodiment of the present
invention additionally has a sampling frequency circuit
5 for performing sampling frequency conversion of a
second speech signal, as compared with the first
embodiment. A sampling frequency conversion circuit 1
converts the sampling frequency of a first speech signal
into a predetermined sampling frequency, and outputs the
resultant signal to a delay adjustment circuit 2.
Likewise, the sampling frequency conversion circuit 5
converts the sampling frequency of the second speech
signal into a predetermined sampling frequency, and
outputs the resultant signal to the delay adjustment
circuit 2. Let D1 be the delay time produced in the
sampling frequency conversion circuit 1, and D2 be the
delay time produced in the sampling frequency conversion
circuit 5.
The delay adjustment circuit 2 performs delay
adjustment to set the first and second speech signals
having undergone sampling frequency conversion in phase,
and outputs the resultant signals to a switching circuit
3.
For delay adjustment, letting D be one of the
delay times D1 and D2 which is longer than the other,
the two signals are delayed by the same time, i.e., the
- 14 -
CA 02376816 2001-12-10
delay time D, using a delay circuit (not shown).
The switching circuit 3 receives the first and
second speech signals having undergone sampling
frequency conversion and delay adjustment from the delay
adjustment circuit 2, switches the two types of speech
signals, in consideration of the delay time D, in
accordance with the control signal, and outputs the
resultant signal to a D/A conversion circuit 4.
The D/A conversion circuit 4 converts the
input digital speech signal into an analog signal and
outputs it. The analog signal is provided for a user
through a speaker, headphone, or the like.
In this embodiment, for example, when the
sampling frequencies of the first and second speech
signals are 8 kHz and 12 kHz, respectively, the sampling
frequencies of the first and second speech signals are
converted into 24 kHz. This makes it possible to
further reduce the processing amount of sampling
frequency conversion as compared with the first
embodiment in which only the sampling frequency of the
first speech signal is converted into 12 kHz.
Fig. 3 is a block diagram showing the
arrangement of the third embodiment of the present
invention. Referring to Fig. 3, the third embodiment of
the present invention further includes an addition
circuit 6 as compared with the first embodiment. In
addition, the operation of a switching circuit 7 differs
- 15 -
CA 02376816 2001-12-10
from that in the first embodiment.
A sampling frequency conversion circuit 1
performs sampling frequency conversion to match the
sampling frequency of a first speech signal to the
sampling frequency of a second speech signal, and
outputs the resultant signal to a delay adjustment
circuit 2. Let the delay time produced in the sampling
frequency conversion circuit 1 be D. The delay
adjustment circuit 2 outputs to the addition circuit 6
the signal obtained by delaying the input second speech
signal by the delay time D and the output signal from
the sampling frequency conversion circuit 1.
The addition circuit 6 weights/adds the first
speech signal having undergone sampling frequency
conversion and the second speech signal having undergone
delay adjustment, and outputs the resultant signal to
the switching circuit 7.
For example, in weighting/adding operation, if
signals before and after switching are given by
S1 (n) , S2 (n) , n = 0, 1, . . . , T-1
then, an output signal S3(n) from the addition circuit 5
is given by
S3 (n) - (n/ (T-1) ) S2 (n) + ( (T-1 - n) / (T-1) ) S1 (n) ,
n = 0, 1,..., T-1, ...(1)
where T is a sample count which represents intervals at
which output signals from the addition circuit are used
and is determined for each sampling frequency of an
- 16 -
CA 02376816 2001-12-10
input speech signal.
In addition, as signals before and after
switching, either the first speech signal having
undergone sampling frequency conversion or the second
speech signal having undergone delay adjustment is
assigned.
The switching circuit 7 receives the first
speech signal having undergone sampling frequency
conversion, the second speech signal having undergone
delay adjustment, the output signal from the addition
circuit 6, and a control signal, and switches the signal
to be output from a signal S1(n) before switching to the
output signal S3(n) from the addition circuit 6 at a
timing set, in consideration of the delay time D, on the
basis of the switching timing designated by the control
signal. The switching circuit 7 then outputs the signal
Sl(n) after switching to a D/A conversion circuit after
outputting the signal S3(n) for a predetermined interval.
A D/A conversion circuit 4 converts the input
digital speech signal into an analog signal and outputs
it. The analog signal is provided for a user through a
speaker, headphone, or the like.
Fig. 4 is a block diagram showing the
arrangement of the fourth embodiment of the present
invention. Referring to Fig. 4, the fourth embodiment
of the present invention further includes an addition
circuit 6 as compared with the second embodiment. In
- 17 -
CA 02376816 2001-12-10
addition, the operation of a switching circuit 7 differs
from that in the second embodiment.
In the fourth embodiment of the present
invention, the operations of the addition circuit 6 and
switching circuit 7 are the same as those described in
the third embodiment.
A sampling frequency conversion circuit 1
converts the sampling frequency of a first speech signal
into a predetermined sampling frequency (e. g., 24 kHz),
and outputs the resultant signal to a delay adjustment
circuit 2. Likewise, a sampling frequency conversion
circuit 5 converts the sampling frequency of a second
speech signal into a predetermined sampling frequency,
and outputs the resultant signal to the delay adjustment
circuit 2. Let D1 be the delay time produced in the
sampling frequency conversion circuit l, and D2 be the
delay time produced in the sampling frequency conversion
circuit 5. The delay adjustment circuit 2 performs
delay adjustment to set the first and second speech
signals having undergone sampling frequency conversion
in phase, and outputs the resultant signals to the
addition circuit 6 and switching circuit 7. For example,
in delay adjustment, letting D be one of the delay times
D1 and D2 which is longer than the other, the two
signals are delayed by the delay time D.
The addition circuit 6 weights/adds the first
and second speech signals having undergone sampling
- 18 -
CA 02376816 2001-12-10
frequency conversion and delay adjustment, and outputs
the resultant signal to the switching circuit 7.
For example, in weighting/adding operation,
equation (1) is used. In this case, as signals S1(n)
and S2(n) before and after switching, one of the first
and second speech signals having undergone sampling
frequency conversion and delay adjustment is assigned.
The switching circuit 7 receives the first and
second speech signals having undergone sampling
frequency conversion and delay adjustment, the output
signal from the addition circuit 6, and a control signal,
and switches the signal to be output from the signal
S1(n) before switching to an output signal S3(n) from
the addition circuit 5 at a timing set, in consideration
of the delay time D, on the basis of the switching
timing designated by the control signal. The switching
circuit 7 then outputs the signal S1(n) after switching
to a D/A conversion circuit after the signal S3(n) is
output for a predetermined interval.
A D/A conversion circuit 4 converts the input
digital speech signal into an analog signal and outputs
it. The analog signal is provided for a user through a
speaker, headphone, or the like.
Fig. 5 is a block diagram showing the
arrangement of a speech switching apparatus according to
the fifth embodiment of the present invention, which is
a combination of a speech decoding circuit 8 and the
- 19 -
CA 02376816 2001-12-10
arrangement of the third embodiment. Referring to
Fig. 5, in the fifth embodiment of the present invention,
the bandwidth-hierarchized speech decoding circuit 8
outputs as first and second digital speech signals
digital speech signals obtained by decoding an input bit
stream to a sampling frequency conversion circuit 1 and
delay circuit 2.
The bandwidth-hierarchized speech decoding
circuit 8 outputs, to an addition circuit 6 and
switching circuit 7, a control signal for instructing
which one of two types of speech signals is to be
reconstructed.
In this case, the bit stream is constituted by
a fundamental portion indispensable to decoding of
compressed speech signal information and an expansion
portion for improving the quality of the speech signal
by expanding the bandwidth.
When, therefore, the bandwidth-hierarchized
speech decoding circuit 8 receives only a fundamental
portion, it decodes the portion into a speech signal
with a narrow bandwidth (e.g., a digital signal having a
sampling frequency of 8 kHz), and outputs it to the
sampling frequency conversion circuit 1.
When this circuit also receives an expansion
portion, it decodes the signal into a speech signal with
a winder bandwidth (e.g., a digital signal having a
sampling frequency of 16 kHz), and outputs it to the
- 20 -
CA 02376816 2001-12-10
delay adjustment circuit 2.
For the decoding operation of the
bandwidth-hierarchized speech decoding circuit 8, refer
to, for example, Japanese Patent Laid-Open No. 11-30997.
When the bandwidth-hierarchized speech
decoding circuit 8 receives the expansion portion of a
bit stream as well as the fundamental portion, the
circuit can simultaneously obtain a plurality of decoded
signals, i.e., a decoded signal obtained by using only
the fundamental portion and a signal obtained by using
both the fundamental portion and the expansion portion.
Assume that in this embodiment, a decoded
signal using only the fundamental portion of a bit
stream is always obtained and output to the delay
adjustment circuit 2.
The operations of the sampling frequency
conversion circuit 1, delay adjustment circuit 2,
addition circuit 6, switching circuit 7, and D/A
conversion circuit 4 are the same as those in the second
embodiment, and hence a description thereof will be
omitted.
Fig. 6 is a block diagram showing the
arrangement of a speech switching apparatus according to
the sixth embodiment of the present invention, which is
a combination of a plurality of speech decoding circuits
and the first embodiment described above. Referring to
Fig. 6, in the sixth embodiment of the present invention,
- 21 -
CA 02376816 2001-12-10
a bit stream switching circuit 11 receives a bit steam
obtained by multiplexing a plurality of bit streams as
compressed signals having different sampling frequencies,
and outputs the input bit stream to a first speech
decoding circuit 9 or second speech decoding circuit 10
depending on the type of the received bit stream.
In this case, as a method of multiplexing bit
streams, a method of simultaneously multiplexing a
plurality of bit streams or a method of switching and
multiplexing them may be used. In the former method,
two types of speech signals are simultaneously decoded
from two types of bit streams. In the latter method, a
speech signal is decoded from only one of the bit
streams. Assume that in this embodiment, a bit stream
obtained by switching/multiplexing a plurality of bit
streams is input.
The bit stream switching circuit 11 outputs,
to a switching circuit 3, a control signal for
instructing which one of the two types of speech signals
is to be reconstructed.
The first speech decoding circuit 9 outputs a
speech signal (e. g., a digital signal having a sampling
frequency of 8 kHz) obtained by decoding a bit stream
having a lower bit rate (e.g., 8 kbit/s) than in the
second speech decoding circuit 10 as a first digital
speech signal to a sampling frequency conversion circuit
1.
- 22 -
" CA 02376816 2001-12-10
The second speech decoding circuit 10 outputs
a speech signal (e.g., a digital signal having a
sampling frequency of 16 kHz) obtained by decoding a bit
stream having a higher bit rate (e. g., 16 kbit/s) than
in the first speech decoding circuit 9 as a second
digital speech signal to a delay adjustment circuit 2.
_...--.
In this case, for the first speech decoding
circuit 9 and second speech decoding circuit 10, refer
to, for example, Japanese Patent Laid-Open No. 10-207496.
The operations of the sampling frequency
conversion circuit 1, delay adjustment circuit 2,
switching circuit 3, and D/A conversion circuit 4 are
the same as those in the first embodiment, and a
description thereof will be omitted.
Fig. 5 shows a combination of the
bandwidth-hierarchized speech decoding circuit and the
arrangement of the third embodiment. Fig. 6 shows a
combination of the plurality of speech decoding circuits
and the arrangement of the first embodiment. Obviously,
however, the above embodiments may be arbitrarily
combined with each other.
In the third and fourth embodiments, since the
addition circuit simultaneously requires a plurality of
signals, when the first and second speech signals are
switched, the two signals must overlap each other.
In the third and fourth embodiments each
combined with the speech decoding circuit, therefore,
- 23 -
CA 02376816 2001-12-10
the apparatus must be combined with a
bandwidth-hierarchized speech decoding circuit.
Alternatively, if a plurality of speech decoding
circuits are used, an input bit stream must be the one
obtained by simultaneously multiplexing a plurality of
bit streams.
Each embodiment exemplifies the case where two
type of input speech signals are processed. An
apparatus designed to process three or more types of
input speech signals can be realized by adding necessary
numbers of sampling frequency conversion circuits and
input/output lines to be connected thereto.
As has been described above, according to the
present invention, strange sounds that are produced when
a plurality of different speech signals are
reconstructed and switched can be reduced.
This is because, in the present invention,
matching the sampling frequencies and phases of signals
before and after switching of a plurality of speech
signals obviates the necessity to change the sampling
frequency settings.
In addition, by weighting/adding speech
signals before and after switching for a predetermined
interval, discontinuity between samples at the start and
end of the interval can be reduced.
- 24 -