.rs
.\" Troff code generated by TPS Convert from ITU Original Files
.\"                 Not Copyright ( c) 1991 
.\"
.\" Assumes tbl, eqn, MS macros, and lots of luck.
.TA 1c 2c 3c 4c 5c 6c 7c 8c
.ds CH
.ds CF
.EQ
delim @@
.EN
.nr LL 40.5P
.nr ll 40.5P
.nr HM 3P
.nr FM 6P
.nr PO 4P
.nr PD 9p
.po 4P

.rs
\v | 5i'
.sp 2P
.LP
\fBRecommendation\ G.722\fR 
.RT
.sp 2P
.sp 1P
.ce 1000
\fB7\ kHz\ AUDIO\(hyCODING\ WITHIN\ 64\ KBIT/S\fR 
.EF '%	Fascicle\ III.4\ \(em\ Rec.\ G.722''
.OF '''Fascicle\ III.4\ \(em\ Rec.\ G.722	%'
.ce 0
.sp 1P
.ce 1000
\fI(Melbourne, 1988)\fR 
.sp 9p
.RT
.ce 0
.sp 1P
.LP
\fB1\fR 	\fBGeneral\fR 
.sp 1P
.RT
.sp 1P
.LP
1.1
	\fIScope and outline description\fR 
.sp 9p
.RT
.PP
This Recommendation describes the characteristics of an audio (50 to 7 
000\ Hz) coding system which may be used for a variety of higher quality 
speech applications. The coding system uses 
sub\(hyband adaptive differential pulse code modulation
(SB\(hyADPCM) within a bit rate of 64 kbitB/Fs. The
system is henceforth referred to as 64 kbit/s (7\ kHz) audio coding. In the
SB\(hyADPCM technique used, the frequency band is split into two sub\(hybands 
(higher and lower) and the signals in each sub\(hyband are encoded using 
ADPCM. 
.PP
The system has three basic modes of operation corresponding to the bit rates
used for 7\ kHz 
audio coding
:\ 64, 56 and 48\ kbit/s. The latter two modes allow an 
auxiliary data channel
of\ 8 and 16\ kbit/s respectively to be provided within the 64\ kbit/s 
by making use of bits from the lower sub\(hyband. 
.PP
Figure 1/G.722 identifies the main functional parts of the 64 kbit/s (7\ 
kHz) 
audio codec
as follows:
.RT
.LP
	i)
	64 kbit/s (7 kHz) audio encoder comprising:
.LP
	\(em
	a transmit audio part which converts an audio signal to
\fR 
a uniform digital signal which is coded using 14\ bits with
\fR 
16\ kHz sampling;
.LP
	\(em
	a SB\(hyADPCM encoder
which reduces the bit rate
to 64\ kbit/s.
.LP
	ii)
	64 kbitB/Fs (7 kHz) audio decoder comprising:
.LP
	\(em
	a 
SB\(hyADPCM decoder
which performs the reverse
operation to the encoder, noting that the effective audio
coding bit rate at the input of the decoder can be\ 64, 56
or 48\ kbit/s depending on the mode of operation;
.LP
\fR 
	\(em
	a receive audio part which reconstructs the audio
signal from the uniform digital signal which is encoded
using 14\ bits with 16\ kHz sampling.
.PP
The following two parts, identified in Figure 1/G.722 for
clarification, will be needed for applications requiring an auxiliary data
channel within the 64\ kbit/s:
.LP
	\(em
	a data insertion device at the transmit end which makes
use of, when needed, 1\ or 2\ audio bits per octet depending on
the mode of operation and substitutes data bits to provide an
auxiliary data channel of 8\ or 16\ kbit/s respectively;
.LP
\fR 
	\(em
	a data extraction device at the receive end which
determines the mode of operation according to a mode control
strategy and extracts the data bits as appropriate.
.PP
Paragraph 1.2 contains a functional description of the transmit and receive 
audio parts, \(sc 1.3 describes the modes of operation and the implication 
of inserting data bits on the algorithms, whilst \(sc\(sc\ 1.4 and\ 1.5 
provide the 
functional descriptions of the SB\(hyADPCM encoding and decoding algorithms
respectively. Paragraph\ 1.6 deals with the timing requirements. Paragraph\ 2
specifies the
transmission characteristics of the 64\ kbit/s (7\ kHz) audio codec and of the
transmit and receive audio parts, \(sc\(sc\ 3 and\ 4 give the principles of the
SB\(hyADPCM encoder respectively whilst \(sc\(sc\ 5 and\ 6 specify the 
computational 
details of the 
Quadrature Mirror Filters
(QMF) and of the ADPCM
encoders and decoders respectively.
.PP
Networking aspects and test sequences are addressed in Appendices\ I
and\ II respectively to this Recommendation.
.PP
Recommendation\ G.725 contains specifications for in\(hychannel
handshaking procedures for terminal identification and for mode control
strategy, including interworking with existing 64\ kbit/s PCM
terminals.
.bp
.RT
.LP
.rs
.sp 31P
.ad r
\fBFigure 1/G.722, p.\fR 
.sp 1P
.RT
.ad b
.RT
.sp 1P
.LP
1.2
	\fIFunctional description of the audio parts\fR 
.sp 9p
.RT
.PP
Figure 2/G.722 shows a possible arrangement of audio parts in a
64\ kbit/s (7\ kHz) 
audio coding terminal
. The microphone, pre\(hyamplifier, power amplifier and loudspeaker are 
shown simply to identify the audio parts 
and are not considered further in this Recommendation.
.PP
In order to facilitate the measurement of the transmission
characteristics as specified in \(sc\ 2, test points\ A and\ B need to 
be provided as shown. These test points may either be for test purposes 
only or, where the 
audio parts are located in different units from the microphone, loudspeaker,
etc., correspond to physical interfaces.
.PP
The transmit and receive audio parts comprise either the following
functional units or any equivalent items satisfying the specifications
of \(sc\ 2:
.RT
.LP
	i)
	transmit:
.LP
	\(em
	an input level adjustment device,
.LP
	\(em
	an input 
anti\(hyaliasing filter
,
.LP
\fR 
	\(em
	a sampling device operating at 16 kHz,
.LP
	\(em
	an 
analogue\(hyto\(hyuniform digital converter
with
14\ bits and with 16\ kHz sampling;
.LP
	ii)
	receive:
.LP
	\(em
	a 
uniform digital\(hyto\(hyanalogue converter
with
14\ bits and with 16\ kHz sampling,
.LP
	\(em
	a 
reconstructing filter
which includes x/sin
x\ correction,
.LP
	\(em
	an output level adjustment device.
.bp
.LP
.rs
.sp 20P
.ad r
\fBFigure 2/G.722, p.\fR 
.sp 1P
.RT
.ad b
.RT
.sp 1P
.LP
1.3
	\fIPossible modes of operation and implications of inserting data\fR 
.sp 9p
.RT
.PP
\fR The three basic possible modes of operation which correspond to the 
bit rates available for audio coding at the input of the decoder are defined 
in Table\ 1/G.722. 
.RT
.ce
\fBH.T. [T1.722]\fR 
.ce
TABLE\ 1/G.722
.ce
\fBBasic possible modes of operation\fR 
.ps 9
.vs 11
.nr VS 11
.nr PS 9
.TS
center box;
cw(36p) | cw(72p) | cw(72p) .
Mode	7 kHz audio coding bit rate	 {
Auxiliary data channel bit rate
 }
_
.T&
cw(36p) | cw(72p) | cw(72p) .
1	64 kbit/s	\ 0 kbit/s
.T&
cw(36p) | cw(72p) | cw(72p) .
2	56 kbit/s	\ 8 kbit/s
.T&
cw(36p) | cw(72p) | cw(72p) .
3	48 kbit/s	16 kbit/s
_
.TE
.nr PS 9
.RT
.ad r
\fBTable 1/G.722 [T1.722], p.\fR 
.sp 1P
.RT
.ad b
.RT
.PP
.sp 3
See Appendix I for examples of applications using one or several of these 
modes and for their corresponding subjective quality. 
.bp
.PP
The 64 kbit/s (7 kHz) audio encoder uses 64 kbit/s for audio coding
at all times irrespective of the mode of operation. The audio coding algorithm 
has been chosen such that, without sending any indication to the encoder, 
the least significant bit or two least significant bits of the lower sub\(hyband 
may 
.PP
be used downstream from the 64\ kbit/s (7\ kHz) audio encoder in order to
substitute the auxiliary data channel bits. However, to maximize the audio
performance for a given mode of operation, the 64\ kbit/s (7\ kHz) audio 
decoder must be optimized to the bit rate available for audio coding. Thus, 
this 
Recommendation describes three variants of the SB\(hyADPCM decoder and, for
applications requiring an auxiliary data channel, an indication must be
forwarded to select in the decoder the variant appropriate to the mode of
operation. Figure\ 1/G.722 illustrates the arrangement. It should be noted 
that the bit rate at the input of the 64\ kbit/s (7\ kHz) audio decoder 
is always 
64\ kbit/s but comprising\ 64, 56 or 48\ kbit/s for audio coding depending 
on the mode of operation. From an algorithm viewpoint, the variant used 
in the 
SB\(hyADPCM decoder can be changed in any octet during the transmission. 
When no indication about the mode of operation is forwarded to the decoder, 
the variant corresponding to Mode\ 1 should be used. 
.PP
A mode mismatch situation, where the variant used in the 64\ kbit/s
(7\ kHz) audio decoder for a given octet does not correspond to the mode of
operation, will not cause misoperation of the decoder. However, to maximize 
the audio performance, it is recommended that the mode control strategy 
adopted in 
.PP
the data extraction device should be such as to minimize the duration of the
mode mismatch. Appendix\ I gives further information on the effects of a mode
mismatch. To ensure compatibility between various types of 64\ kbit/s (7\ kHz)
audio coding terminals, it is recommended that, as a minimum, the variant
corresponding to Mode\ 1 operation is always implemented in the decoder.
.PP
\fR 
The mode control strategy could be derived from the auxiliary data
channel protocol (see Recommendation\ G.725).
.RT
.sp 1P
.LP
1.4
	\fIFunctional description of the\fR 
\fISB\(hyADPCM encoder\fR 
.sp 9p
.RT
.PP
\fR 
Figure 3/G.722 is a block diagram of the SB\(hyADPCM encoder. A
functional description of each block is given below in \(sc\(sc\ 1.4.1 
to\ 1.4.4. 
.RT
.LP
.rs
.sp 13P
.ad r
\fBFigure 3/G.722, p.\fR 
.sp 1P
.RT
.ad b
.RT
.sp 1P
.LP
1.4.1
	\fITransmit quadrature mirror filters\fR (QMFs)
.sp 9p
.RT
.PP
The transmit QMFs comprise two linear\(hyphase non\(hyrecursive digital 
filters which split the frequency band 0\ to 8000\ Hz into two sub\(hybands: 
the 
lower sub\(hyband (0\ to 4000\ Hz) and the higher sub\(hyband (4000\ to 
8000\ Hz). The 
.PP
input to the transmit QMFs,\ \fIx\fR\d\fIi\fR\\d\fIn\fR\u, is the output 
from the transmit audio part and is sampled at 16\ kHz. The outputs,\ \fIx\fR\d\fIL\fR\uand\ 
\fIx\fR\d\fIH\fR\u, 
for the lower and higher sub\(hybands respectively, are sampled at 8\ kHz.
.RT
.sp 1P
.LP
1.4.2
	\fILower sub\(hyband ADPCM encoder\fR 
.sp 9p
.RT
.PP
Figure 4/G.722 is a block diagram of the lower sub\(hyband ADPCM
encoder. The lower sub\(hyband input signal,\ \fIx\fR\d\fIL\fR\uafter subtraction 
of an 
estimate,\ \fIs\fR\d\fIL\fR\u, of the input signal produces the difference 
signal, 
\fIe\fR\d\fIL\fR\u. An adaptive 60\(hylevel 
non linear quantizer
is used to
assign six binary digits to the value of the difference signal to produce a
48\ kbit/s signal,\ \fII\fR\d\fIL\fR\u.
.bp
.PP
In the feedback loop, the two least significant bits of \fII\fR\d\fIL\fR\uare
deleted to produce a 4\(hybit signal\ \fII\fR\d\fIL\fR\\d\fIt\fR\u, which 
is used for the 
quantizer adaptation and applied to a 15\(hylevel inverse adaptive quantizer to
produce a quantized difference signal,\ \fId\fR\d\fIL\fR\\d\fIt\fR\u. The 
signal estimate, \fIs\fR\d\fIL\fR\uis added to this quantized difference 
signal to produce a reconstructed version,\ \fIr\fR\d\fIL\fR\\d\fIt\fR\u, 
of the lower sub\(hyband input signal. Both the 
reconstructed signal and the quantized difference signal are operated upon 
by an adaptive predictor which produce the estimate,\ \fIs\fR\d\fIL\fR\u, 
of the input signal, thereby completing the feedback loop. 
.PP
4\(hybit operation, instead of 6\(hybit operation, in the feedback loops 
of both the lower sub\(hyband ADPCM encoder, and the lower sub\(hyband 
ADPCM decoder 
allows the possible insertion of data in the two least significant bits as
described in \(sc\ 1.3 without causing misoperation in the decoder. Use of a
60\(hylevel quantizer (instead of 64\(hylevel) ensures that the pulse density
requirements as described in Recommendation\ G.802 are met under all conditions 
and in all modes of operation. 
.RT
.LP
.rs
.sp 29P
.ad r
\fBFigure 4/G.722, p.\fR 
.sp 1P
.RT
.ad b
.RT
.sp 1P
.LP
1.4.3
	\fIHigher sub\(hyband ADPCM encoder\fR 
.sp 9p
.RT
.PP
Figure 5/G.722 is a block diagram of the higher sub\(hyband ADPCM
encoder. The higher sub\(hyband input signal,\ \fIx\fR\d\fIH\fR\uafter 
subtraction of an 
estimate,\ \fIs\fR\d\fIH\fR\u, of the input signal, produces the difference 
signal, \fIe\fR\d\fIH\fR\u. An adaptive 4\(hylevel non linear quantizer 
is used to assign two binary digits to the value of the difference signal 
to produce a 16\ kbit/s 
signal,\ \fII\fR\d\fIH\fR\u.
.PP
\fR An inverse adaptive quantizer produces a quantized difference
signal,\ \fId\fR\d\fIH\fR\u, from these same two binary digits. The signal 
estimate, \fIs\fR\d\fIH\fR\u, is added to this quantized difference signal 
to produce a 
reconstructed version,\ \fIr\fR\d\fIH\fR\u, of the higher sub\(hyband input 
signal. Both the reconstructed signal and the quantized difference signal 
are operated upon by an adaptive predictor which produces the estimate,\ 
\fIs\fR\d\fIH\fR\u, of the 
input signal, thereby completing the feedback loop.
.bp
.RT
.LP
.rs
.sp 23P
.ad r
\fBFigure 5/G.722, p.\fR 
.sp 1P
.RT
.ad b
.RT
.sp 1P
.LP
1.4.4
	\fIMultiplexer\fR 
.sp 9p
.RT
.PP
The multiplexer (MUX) shown in Figure 3/G.722 is used to combine the signals,\ 
\fII\fR\d\fIL\fR\uand\ \fII\fR\d\fIH\fR\u, from the lower and higher sub\(hyband 
ADPCM encoders respectively into a composite 64\ kbit/s signal,\ I, with 
an octet format for transmission. 
.PP
The output octet format, after multiplexing, is as
follows:
\v'6p'
.RT
.sp 1P
.ce 1000
\fII\fR\d\fIH\fR\\d1\u\fII\fR\d\fIH\fR\\d2\u\fII\fR\d\fIL\fR\\d1\u\fII\fR\d\fIL\fR\\d2\u\fI\fR 
\fII\fR\d\fIL\fR\\d3\u\fII\fR\d\fIL\fR\\d4\u\fII\fR\d\fIL\fR\\d5\u\fII\fR\d\fIL\fR\\d6\u 
.ce 0
.sp 1P
.LP
.sp 1
where \fII\fR\d\fIH\fR\\d1\uis the first bit transmitted, and
where\ \fII\fR\d\fIH\fR\\d1\uand\ \fII\fR\d\fIL\fR\\d1\uare the most significant 
bits of \fII\fR\d\fIH\fR\uand\ \fII\fR\d\fIL\fR\urespectively, whilst\ 
\fII\fR\d\fIH\fR\\d2\uand\ \fII\fR\d\fIL\fR\\d6\uare the least significant 
bits of\ \fII\fR\d\fIH\fR\uand\ \fII\fR\d\fIL\fR\u 
respectively.
.sp 1P
.LP
1.5
	\fIFunctional description of the\fR 
\fISB\(hyADPCM decoder\fR 
.sp 9p
.RT
.PP
\fR 
Figure 6/G.722 is a block diagram of the SB\(hyADPCM decoder. A
functional description of each block is given below in \(sc\(sc\ 1.5.1 
to\ 1.5.4. 
.RT
.LP
.rs
.sp 13P
.ad r
\fBFigure 6/G.722, p.\fR 
.sp 1P
.RT
.ad b
.RT
.LP
.bp
.sp 1P
.LP
1.5.1
	\fIDemultiplexer\fR 
.sp 9p
.RT
.PP
The demultiplexer (DMUX) decomposes the received 64 kbit/s
octet\(hyformatted signal,\ \fII\fR\d\fIr\fR\u, into two signals,\ \fII\fR\d\fIL\fR\\d\fIr\fR\uand 
\fII\fR\d\fIH\fR\u, which form the codeword inputs to the lower and higher 
sub\(hyband ADPCM decoders respectively. 
.RT
.sp 1P
.LP
1.5.2
	\fILower sub\(hyband ADPCM decoder\fR 
.sp 9p
.RT
.PP
Figure 7/G.722 is a block diagram of the lower sub\(hyband ADPCM
decoder. This decoder can operate in any of three possible variants depending 
on the received indication of the mode of operation. 
.RT
.LP
.rs
.sp 35P
.ad r
\fBFigure 7/G.722, p.\fR 
.sp 1P
.RT
.ad b
.RT
.PP
The path which produces the estimate, \fIs\fR\d\fIL\fR\u, of the input 
signal including the 
quantizer adaptation
, is identical to the feedback portion of the lower sub\(hy band ADPCM 
encoder described in \(sc\ 1.4.2. The 
reconstructed signal,\ \fIr\fR\d\fIL\fR\u, is produced by adding to the signal
estimate one of three possible quantized difference
signals,\ \fId\fR\d\fIL\fR\\d,\\d6\u,\ \fId\fR\d\fIL\fR\\d,\\d5\uor\ \fId\fR\d\fIL\fR\\d,\\d4\u(=\ 
\fId\fR\d\fIL\fR\\d\fIt\fR\u\(hy see note), selected according to the received 
indication of 
the mode of operation. For each indication, Table\ 2/G.722 shows the quantized 
difference signal selected, the inverse adaptive quantizer used and the 
number of least significant bits deleted from the input codeword. 
.bp
.ce
\fBH.T. [T2.722]\fR 
.ce
TABLE\ 2/G.722
.ce
\fBLower sub\(hyband ADPCM decoder variants\fR 
.ps 9
.vs 11
.nr VS 11
.nr PS 9
.TS
center box;
cw(60p) | cw(54p) | cw(60p) | cw(54p) .
 {
Received indication of mode of operation
 }	 {
Quantized difference signal selected
 }	 {
Inverse adaptive quantizer used
 }	 {
Number of least significant bits deleted from input codeword,
I
L
r
 }
_
.T&
cw(60p) | cw(54p) | cw(60p) | cw(54p) .
Mode 1	d L , 6	60\(hylevel	0
.T&
cw(60p) | cw(54p) | cw(60p) | cw(54p) .
Mode 2	d L , 5	30\(hylevel	1
.T&
cw(60p) | cw(54p) | cw(60p) | cw(54p) .
Mode 3	d L , 4	15\(hylevel	2
.TE
.LP
\fINote\fR
\ \(em\ For clarification purposes, all three inverse quantizers have been
indicated in the upper portion of Figure\ 7/G.722. In an optimized
implementation, the signal\ d
L
t, produced in the predictor loop, could be substituted for\ d
L , 4.
.nr PS 9
.RT
.ad r
\fBTable 2/G.722 [T2.722], p.\fR 
.sp 1P
.RT
.ad b
.RT
.sp 1P
.LP
.sp 2
1.5.3
	\fIHigher sub\(hyband ADPCM decoder\fR 
.sp 9p
.RT
.PP
Figure 8/G.722 is a block diagram of the higher sub\(hyband ADPCM
decoder. This decoder is identical to the feedback portion of the higher
sub\(hyband ADPCM encoder described in \(sc\ 1.4.3, the output being the 
reconstructed signal,\ \fIr\fR\d\fIH\fR\u. 
.RT
.LP
.rs
.sp 17P
.ad r
\fBFigure 8/G.722, p.\fR 
.sp 1P
.RT
.ad b
.RT
.sp 1P
.LP
1.5.4
	\fIReceive QMFs\fR 
.sp 9p
.RT
.PP
The receive QMFs shown in Figure 6/G.722 are two 
linear\(hyphase   non\(hyrecursive digital filters
which interpolate the outputs,\ \fIr\fR\d\fIL\fR\uand
\fIr\fR\d\fIH\fR\u, of the lower and higher sub\(hyband ADPCM decoders 
from 8\ kHz to 
16\ kHz and which then produce an output,\ \fIx\fR\d\fIo\fR\\d\fIu\fR\\d\fIt\fR\u, 
sampled at 
16\ kHz which forms the input to the receive audio parts.
.PP
\fR Excluding the ADPCM coding processes, the combination of the
transmit and the receive QMFs has an impulse response which closely
approximates a simple delay whilst, at the same time, the 
aliasing
effects associated with the 8\ kHz 
sub\(hysampling
are cancelled.
.bp
.RT
.sp 1P
.LP
1.6
	\fITiming requirements\fR 
.sp 9p
.RT
.PP
64 kHz bit timing and 8 kHz octet timing should be provided by the network 
to the audio decoder. 
.PP
For a correct operation of the audio coding system, the precision of the 
16\ kHz sampling frequencies of the A/D and D/A converters must be better 
than \(+- | 0 | (mu | 0\uD\dlF261\u6\d.
.RT
.sp 2P
.LP
\fB2\fR 	\fBTransmission characteristics\fR 
.sp 1P
.RT
.sp 1P
.LP
2.1
	\fICharacteristics of the audio ports and the\fR 
\fItest points\fR 
.sp 9p
.RT
.PP
\fR 
Figure 2/G.722 indicates the audio input and output ports and the test 
points (A and\ B). It is for the designer to determine the characteristics 
of the audio ports and the test points (i.e.\ relative levels, impedances, 
whether balanced or unbalanced). The microphone, pre\(hyamplifier, power 
amplifier and loudspeaker should be chosen with reference to the specifications 
of the 
audio parts: in particular their nominal bandwidth, idle noise and
distortion.
.PP
It is suggested that input and ouput impedances
should be high and low, respectively, for an unbalanced termination
whilst for a balanced termination these impedances should be 600 ohms.
However, the audio parts should meet all audio parts specifications for 
their respective input and output impedance conditions. 
.RT
.sp 1P
.LP
2.2
	\fIOverload point\fR 
.sp 9p
.RT
.PP
The overload point for the analogue\(hyto\(hydigital and
digital\(hyto\(hyanalogue converters should be + 9\ dBm0 \(+- | .3\ dB. 
This assumes 
the same nominal speech level (see Recommendation\ G.232) as for 64\ kbit/s 
PCM, but with a wider margin for the maximum signal level which is likely 
to be 
necessary with conference arrangements. The measurement method of the overload 
point is under study. 
.RT
.sp 1P
.LP
2.3
	\fINominal reference frequency\fR 
.sp 9p
.RT
.PP
Where a nominal reference frequency of 1000\ Hz is indicated below, the 
actual frequency should be chosen equal to 1020\ Hz. The frequency 
tolerance should be +2 to \(em7\ Hz.
.RT
.sp 1P
.LP
2.4
	\fITransmission characteristics of the 64 kbit/s (7 kHz) audio\fR 
\fIcodec\fR 
.sp 9p
.RT
.PP
The values and limits specified below should be met with a
64\ kbit/s (7\ kHz) audio encoder and decoder connected back\(hyto\(hyback. For
practical reasons, the measurements may be performed in a looped configuration 
as shown in Figure\ 9a)/G.722. However, such a looped configuration is 
only 
intended to simulate an actual situation where the encoder and decoder are
located at the two ends of a connection.
.PP
These limits apply to operation in Mode 1.
.RT
.sp 1P
.LP
2.4.1
	\fINominal bandwidth\fR 
.sp 9p
.RT
.PP
The nominal 3 dB bandwidth is 50 to 7000 Hz.
.RT
.sp 1P
.LP
2.4.2
	\fIAttenuation/frequency distortion\fR 
.sp 9p
.RT
.PP
The variation with frequency of the attenuation should satisfy the limits 
shown in the mask of Figure\ 10/G.722. The nominal reference frequency 
is 1000\ Hz and the test level is \(em10\ dBm0.
.RT
.sp 1P
.LP
2.4.3
	\fIAbsolute group delay\fR 
.sp 9p
.RT
.PP
The absolute group delay, defined as the minimum group delay for a sine 
wave signal between\ 50 and 7000\ Hz, should not exceed 4\ ms. The test 
level is \(em10\ dBm0.
.RT
.sp 1P
.LP
2.4.4
	\fIIdle noise\fR 
.sp 9p
.RT
.PP
The unweighted noise power measured in the frequency range 50 to
7000\ Hz with no signal at the input port (test point\ A) should not exceed
\(em66\ dBm0. When measured in the frequency range 50\ Hz to 20\ kHz the 
unweighted noise power should not exceed \(em60\ dBm0. 
.bp
.RT
.LP
.rs
.sp 47P
.ad r
\fBFigure 9/G.722, p.\fR 
.sp 1P
.RT
.ad b
.RT
.LP
.bp
.LP
.rs
.sp 19P
.ad r
\fBFigure 10/G.722, p.\fR 
.sp 1P
.RT
.ad b
.RT
.sp 1P
.LP
2.4.5
	\fISingle frequency noise\fR 
.sp 9p
.RT
.PP
The level of any single frequency (in particular 8000 Hz, the
sampling frequency and its multiples), measured selectively with no signal 
at the input port (test point\ A) should not exceed \(em70\ dBm0. 
.RT
.sp 1P
.LP
2.4.6
	\fISignal\(hyto\(hytotal distortion ratio\fR 
.sp 9p
.RT
.PP
Under study.
.RT
.sp 1P
.LP
2.5
	\fITransmission characteristics of the audio parts\fR 
.sp 9p
.RT
.PP
When the measurements indicated below for the audio parts are from audio\(hyto\(hyaudio, 
a looped configuration as shown in Figure\ 9b)/G.722 should be used. The 
audio parts should also meet the specifications of \(sc\ 2.4 with the 
measurement configuration of Figure\ 9b)/G.722.
.RT
.sp 1P
.LP
2.5.1
	\fIAttenuationB/Ffrequency response of the input\fR 
\fIanti\(hyaliasing\fR \fIfilter\fR 
.sp 9p
.RT
.PP
The in\(hyband and out\(hyof\(hyband attenuation/frequency response of 
the input anti\(hyaliasing filter should satisfy the limits of the mask 
shown in 
Figure\ 11/G.722. The nominal reference frequency is 1000\ Hz and the test 
level for the in\(hyband characteristic is \(em10\ dBm0. Appropriate measurements 
should be made to check the out\(hyof\(hyband characteristic taking into 
account the aliasing due to the 16\ kHz sampling. 
.RT
.sp 1P
.LP
2.5.2
	\fIAttenuationB/Ffrequency response of the output\fR 
\fIreconstructing\fR \fIfilter\fR 
.sp 9p
.RT
.PP
The in\(hyband and out\(hyof\(hyband attenuation/frequency response of 
the output reconstructing filter should satisfy the limits of the mask 
shown in 
Figure\ 12/G.722. The nominal reference frequency is 1000\ Hz and the test 
level for the in\(hyband characteristic is \(em10\ dBm0. Appropriate measurements 
should be made to check the out\(hyof\(hyband characteristic taking into 
account the aliasing due to the 16\ kHz sampling. The mask of Figure\ 12/G.722 
is valid for the whole of the receive audio part including any pulse amplitude 
modulation distortion and x/sin x\ correction. 
.bp
.RT
.LP
.rs
.sp 20P
.ad r
\fBFigure 11/G.722, p.\fR 
.sp 1P
.RT
.ad b
.RT
.LP
.rs
.sp 21P
.ad r
\fBFigure 12/G.722, p.\fR 
.sp 1P
.RT
.ad b
.RT
.LP
.bp
.sp 1P
.LP
2.5.3
	\fIGroup\(hydelay distortion with frequency\fR 
.sp 9p
.RT
.PP
The group\(hydelay distortion, taking the minimum value of group delay 
as a reference, should satisfy the limits of the mask shown in 
Figure\ 13/G.722.
.RT
.LP
.rs
.sp 21P
.ad r
\fBFigure 13/G.722, p.\fR 
.sp 1P
.RT
.ad b
.RT
.sp 1P
.LP
2.5.4
	\fIIdle noise for the receive audio part\fR 
.sp 9p
.RT
.PP
The unweighted noise power of the receive audio part measured in
the frequency range 50\ to 7000\ Hz with a 14\(hybit all\(hyzero signal 
at its input 
should not exceed \(em75\ dBm0.
.RT
.sp 1P
.LP
2.5.5
	\fISignal\(hyto\(hytotal distortion ratio as a function of input\fR 
\fIlevel\fR 
.sp 9p
.RT
.PP
With a sine wave signal at a frequency excluding simple harmonic
relationships with the 16\ kHz sampling frequency, applied to test point\ 
A, the ratio of signal\(hyto\(hytotal distortion power as a function of 
input level measured unweighted in the frequency range 50\ to 7000\ Hz 
at test point\ B, should 
satisfy the limits of the mask shown in Figure\ 14/G.722. Two measurements
should be performed, one at a frequency of about 1\ kHz and the other at a
frequency of about 6\ kHz.
.RT
.LP
.rs
.sp 17P
.ad r
\fBFigure 14/G.722, p.\fR 
.sp 1P
.RT
.ad b
.RT
.LP
.bp
.sp 1P
.LP
2.5.6
	\fISignal\(hyto\(hytotal distortion ratio as a function of\fR 
\fIfrequency\fR 
.sp 9p
.RT
.PP
With a sine wave signal at a level of \(em10 dBm0 applied to test
point\ A, the ratio of signal\(hyto\(hytotal distortion power as a function of
frequency measured unweighted in the frequency range 50\ to 7000\ Hz at test
point\ B should satisfy the limits of the mask shown in Figure\ 15/G.722.
.RT
.LP
.rs
.sp 18P
.ad r
\fBFigure 15/G.722, p.\fR 
.sp 1P
.RT
.ad b
.RT
.sp 1P
.LP
2.5.7
	\fIVariation of gain with input level\fR 
.sp 9p
.RT
.PP
With a sine wave signal at the nominal reference frequency of
1000\ Hz, but excluding the sub\(hymultiple of the 16\ kHz sampling frequency,
applied to test point\ A, the gain variation as a function of input level
relative to the gain at an input level of \(em10\ dBm0 measured selectively 
at test point\ B, should satisfy the limits of the mask shown in Figure\ 
16/G.722. 
.RT
.LP
.rs
.sp 24P
.ad r
\fBFigure 16/G.722, p.\fR 
.sp 1P
.RT
.ad b
.RT
.LP
.bp
.sp 1P
.LP
2.5.8
	\fIIntermodulation\fR 
.sp 9p
.RT
.PP
Under study.
.RT
.sp 1P
.LP
2.5.9
	\fIGo/return crosstalk\fR 
.sp 9p
.RT
.PP
The crosstalk from the transmit direction to the receive direction should 
be such that, with a sine wave signal at any frequency in the range 
50\ to 7000\ Hz and at a level of +6\ dBm0 applied to test point\ A, the 
crosstalk level measured selectively at test point\ B should not exceed 
\(em64\ dBm0. The 
measurement should be made with a 14\(hybit all\(hyzero digital signal 
at the input to the receive audio part. 
.PP
The crosstalk from the receive direction to the transmit direction
should be such that, with a digitally simulated sine wave signal at any
frequency in the range of 50\ to 7000\ Hz and a level of +6\ dBm0 applied 
to the input of the receive audio part, the crosstalk level measured selectively 
and with the measurement made digitally at the output of the transmit audio 
part 
should not exceed \(em64\ dBm0. The measurement should be made with no 
signal at 
test point\ A, but with the test point correctly terminated.
.RT
.sp 1P
.LP
2.6
	\fITranscoding to and from 64 kbit/s PCM\fR 
.sp 9p
.RT
.PP
For compatibility reasons with 64 kbit/s PCM, transcoding between 64\ kbit/s 
(7\ kHz) audio coding and 64\ kbit/s PCM should take account of the 
relevant specifications of Recommendations\ G.712, G.713 and\ G.714. When the
audio signal is to be heard through a loudspeaker, more stringent
specifications may be necessary. Further information may be found in
Appendix\ I.
.RT
.sp 2P
.LP
\fB3\fR 	\fBSB\(hyADPCM encoder principles\fR 
.sp 1P
.RT
.PP
A block diagram of the SB\(hyADPCM encoder is given in Figure 3/G.722. 
Block diagrams of the lower and higher sub\(hyband ADPCM encoders are given 
respectively in Figures\ 4/G.722 and\ 5/G.722.
.PP
Main variables used for the descriptions in \(sc\(sc 3 and 4 are
summarized in Table\ 3/G.722. In these descriptions, index (\fIj\fR ) indicates 
a value corresponding to the current 16\ kHz sampling interval, index (\fIj\fR 
\(eml) 
indicates a value corresponding to the previous 16\ kHz sampling interval,
index (\fIn\fR ) indicates a value corresponding
to the current 8\ kHz sampling interval, and index\ (\fIn\fR \(em1)
indicates a value corresponding to the previous 8\ kHz sampling interval.
Indices are not used for internal variables, i.e.\ those employed only within
individual computational blocks.
.RT
.sp 1P
.LP
3.1
	\fITransmit QMF\fR 
.sp 9p
.RT
.PP
A 24\(hycoefficient QMF is used to compute the lower and higher
sub\(hyband signal components. The QMF coefficient values, \fIh\fR\d\fIi\fR\u, 
are 
given in Table\ 4/G.722.
.PP
The output variables, \fIx\fR\d\fIL\fR\u(\fIn\fR ) and
\fIx\fR\d\fIH\fR\u(\fIn\fR ), are computed in the following way:
\v'6p'
.RT
.ad r
.ad b
.RT
.ad r
.ad b
.RT
.ad r
.ad b
.RT
.ad r
.ad b
.RT
.sp 1P
.LP
3.2
	\fIDifference signal computation\fR 
.sp 9p
.RT
.PP
The difference signals, \fIe\fR\d\fIL\fR\u(\fIn\fR ) and
\fIe\fR\d\fIH\fR\u(\fIn\fR ), are computed by subtracting predicted values,
\fIs\fR\d\fIL\fR\u(\fIn\fR ) and \fIs\fR\d\fIH\fR\u(\fIn\fR ), from the 
lower and 
higher sub\(hyband input values, \fIx\fR\d\fIL\fR\u(\fIn\fR ) and
\fIx\fR\d\fIH\fR\u(\fIn\fR ):
\v'6p'
.RT
.ad r
.ad b
.RT
.ad r
.ad b
.RT
.LP
.bp
.ce
\fBH.T. [T3.722]\fR 
.ce
TABLE\ 3/G.722
.ce
\fBVariables used in the SB\(hyADPCM encoder and decoder descriptions\fR 
.ps 9
.vs 11
.nr VS 11
.nr PS 9
.TS
center box;
cw(48p) | cw(180p) .
Variable	Description
_
.T&
lw(48p) | lw(180p) .
x i n	 {
Input value
(uniform representation)
 }
_
.T&
lw(48p) | lw(180p) .
x L, x H	QMF output signals
_
.T&
lw(48p) | lw(180p) .
S L p, S H p	 {
Pole\(hypredictor output signals
 }
_
.T&
lw(48p) | lw(180p) .
a L , i, a H , i	 {
Pole\(hypredictor coefficients
 }
_
.T&
lw(48p) | lw(180p) .
r L, r L t, r H	 {
Reconstructed signals (non truncated and truncated)
 }
_
.T&
lw(48p) | lw(180p) .
b L , i, b H , i	 {
Zero\(hypredictor coefficients
 }
_
.T&
lw(48p) | lw(180p) .
d L, d L t, d H	 {
Quantized difference signals (non truncated and truncated)
 }
_
.T&
lw(48p) | lw(180p) .
S L z, S H z	 {
Zero\(hypredictor output signals
 }
_
.T&
lw(48p) | lw(180p) .
S L, S H	Predictor output signals
_
.T&
lw(48p) | lw(180p) .
e L, e H	 {
Difference signals to be quantized
 }
_
.T&
lw(48p) | lw(180p) .
\(gr L, \(gr H	 {
Logarithmic quantizer scale factors
 }
_
.T&
lw(48p) | lw(180p) .
?63 L, ?63 H	 {
Quantizer scale factor (linear)
 }
_
.T&
lw(48p) | lw(180p) .
I L, I L t, I H	 {
Codewords (non truncated and truncated)
 }
_
.T&
lw(48p) | lw(180p) .
P L t, P H	 {
Partially reconstructed signals
 }
_
.T&
lw(48p) | lw(180p) .
I L r	 {
Received lower sub\(hyband codeword
 }
_
.T&
lw(48p) | lw(180p) .
X o u t	Output value (uniform)
.TE
.LP
\fINote\fR
\ \(em\ Variables used exclusively within one section are not listed.
Subscripts L and H refer to lower sub\(hyband and higher sub\(hyband values.
Subscript Lt denotes values generated from the truncated\ 4\(hybit codeword as
opposed to the nontruncated 6\(hybit (encoder) or\ 6\(hy, 5\(hy or 4\(hybit (decoder)
codewords.
.nr PS 9
.RT
.ad r
\fBTableau 3/G.722 [T3.722], p. 19\fR 
.sp 1P
.RT
.ad b
.RT
.LP
.rs
.sp 8P
.ad r
Blanc
.ad b
.RT
.LP
.bp
.ce
\fBH.T. [T4.722]\fR 
.ce
TABLE\ 4/G.722
.ce
\fBTransmit and receive OMF coefficient values\fR 
.ps 9
.vs 11
.nr VS 11
.nr PS 9
.TS
center box;
cw(90p) | cw(90p) .
 {
h
0\fB\fR\(da\fB1\fR
, h
2
3
 }	 {
\fB\(em\fR
0.366211E\(em03
 }
.T&
cw(90p) | cw(90p) .
 {
h
1\fB\fR\(da\fB1\fR
, h
2
2
 }	\(em0.134277E\(em02
.T&
cw(90p) | cw(90p) .
 {
h
2\fB\fR\(da\fB1\fR
, h
2
1
 }	\(em0.134277E\(em02
.T&
cw(90p) | cw(90p) .
 {
h
3\fB\fR\(da\fB1\fR
, h
2
0
 }	 {
\fB\(em\fR
0.646973E\(em02
 }
.T&
cw(90p) | cw(90p) .
 {
h
4\fB\fR\(da\fB1\fR
, h
1
9
 }	 {
\fB\(em\fR
0.146484E\(em02
 }
.T&
cw(90p) | cw(90p) .
 {
h
5\fB\fR\(da\fB1\fR
, h
1
8
 }	\(em0.190430E\(em01
.T&
cw(90p) | cw(90p) .
 {
h
6\fB\fR\(da\fB1\fR
, h
1
7
 }	 {
\fB\(em\fR
0.390625E\(em02
 }
.T&
cw(90p) | cw(90p) .
 {
h
7\fB\fR\(da\fB1\fR
, h
1
6
 }	 {
\fB\(em\fR
0.441895E\(em01
 }
.T&
cw(90p) | cw(90p) .
 {
h
8\fB\fR\(da\fB1\fR
, h
1
5
 }	\(em0.256348E\(em01
.T&
cw(90p) | cw(90p) .
 {
h
9\fB\fR\(da\fB1\fR
, h
1
4
 }	\(em0.982666E\(em01
.T&
cw(90p) | cw(90p) .
h 1 0, h 1 3	\fB\(em\fR 0.116089E+00
.T&
cw(90p) | cw(90p) .
h 1 1, h\fR 1 2	\fB\(em\fR 0.473145E+00
_
.TE
.nr PS 9
.RT
.ad r
\fBTableau 4/G.722 [T4.722], p. 20\fR 
.sp 1P
.RT
.ad b
.RT
.sp 1P
.LP
3.3
	\fIAdaptive quantizer\fR 
.sp 9p
.RT
.PP
The difference signals, \fIe\fR\d\fIL\fR\u(\fIn\fR ) and
\fIe\fR\d\fIH\fR\u(\fIn\fR ), are quantized to 6\ and 2\ bits for the lower and
higher sub\(hybands respectively. Tables\ 5/G.722 and\ 6/G.722 give the 
decision levels
and the output codes for the 6\(hy and 2\(hybit quantizers respectively. 
In these tables, only the positive decision levels are indicated, the negative 
levels can be determined by symmetry.\ \fIm\fR\d\fIL\fR\uand\ \fIm\fR\d\fIH\fR\uare 
indices for the quantizer intervals. The interval boundaries,\ \fILL\fR 
6, \fILU\fR 6, \fIHL\fR and\ \fIHU\fR , are scaled by computed scale factors,\ 
?63 
\fI\fI\d\fIL\fR\u(\fIn\fR )
and\ ?63
\fI\fI\d\fIH\fR\u(\fIn\fR ) (see \(sc\ 3.5). Indices,\ \fIm\fR\d\fIL\fR\u
and\ \fIm\fR\d\fIH\fR\u, are then determined to satisfy the following:
\v'6p'
.RT
.ad r
.ad b
.RT
.ad r
.ad b
.RT
.LP
for the lower and higher sub\(hybands respectively.
.PP
The output codes, \fIILN\fR and \fIIHN\fR , represent negative intervals,
whilst the output codes,\ \fIILP\fR and\ \fIIHP\fR , represent positive 
intervals. The 
output codes,\ \fII\fR\d\fIL\fR\u(\fIn\fR ) and\ \fII\fR\d\fIH\fR\u(\fIn\fR 
), are then given by: 
\v'6p'
.ad r
.ad b
.RT
.ad r
.ad b
.RT
.LP
for the lower and higher sub\(hybands respectively.
.bp
.ce
\fBH.T. [T5.722]\fR 
.ce
TABLE\ 5/G.722
.ce
\fBDecision levels and output codes for the 6\(hybit lower sub\(hyband\fR 
.ce
\fBquantizer\fR 
.ps 9
.vs 11
.nr VS 11
.nr PS 9
.TS
center box;
cw(36p) | cw(48p) | cw(48p) | cw(48p) | cw(48p) .
m L	LL6	LU6	ILN	ILP
_
.T&
cw(36p) | cw(48p) | cw(48p) | cw(48p) | cw(48p) .
\ 1 \ 2	0.00000 0.06817	0.06817 0.14103	111111 111110	111101 111100
_
.T&
cw(36p) | cw(48p) | cw(48p) | cw(48p) | cw(48p) .
\ 3 \ 4 \ 5 \ 6	 {
0.14103
0.21389
0.29212
0.37035
 }	 {
0.21389
0.29212
0.37035
0.45482
 }	011111 011110 011101 011100	111011 111010 111001 111000
_
.T&
cw(36p) | cw(48p) | cw(48p) | cw(48p) | cw(48p) .
\ 7 \ 8 \ 9 10	 {
0.45482
0.53929
0.63107
0.72286
 }	 {
0.53929
0.63107
0.72286
0.82335
 }	011011 011010 011001 011000	110111 110110 110101 110100
_
.T&
cw(36p) | cw(48p) | cw(48p) | cw(48p) | cw(48p) .
11 12 13 14	 {
0.82335
0.92383
1.03485
1.14587
 }	 {
0.92383
1.03485
1.14587
1.26989
 }	010111 010110 010101 010100	110011 110010 110001 110000
_
.T&
cw(36p) | cw(48p) | cw(48p) | cw(48p) | cw(48p) .
15 16 17 18	 {
1.26989
1.39391
1.53439
1.67486
 }	 {
1.39391
1.53439
1.67486
1.83683
 }	010011 010010 010001 010000	101111 101110 101101 101100
_
.T&
cw(36p) | cw(48p) | cw(48p) | cw(48p) | cw(48p) .
19 20 21 22	 {
1.83683
1.99880
2.19006
2.38131
 }	 {
1.99880
2.19006
2.38131
2.61482
 }	001111 001110 001101 001100	101011 101010 101001 101000
_
.T&
cw(36p) | cw(48p) | cw(48p) | cw(48p) | cw(48p) .
23 24 25 26	 {
2.61482
2.84833
3.14822
3.44811
 }	 {
2.84833
3.14822
3.44811
3.86796
 }	001011 001010 001001 001000	100111 100110 100101 100100
_
.T&
cw(36p) | cw(48p) | cw(48p) | cw(48p) | cw(48p) .
27 28 29 30	 {
3.86796
4.28782
4.99498
5.70214
 }	4.28782 4.99498 5.70214 \(if	000111 000110 000101 000100	 {
100011
100010
100001
100000
 }
.TE
.LP
\fINote\fR
\ \(em\ If a transmitted codeword for the lower sub\(hyband signal has been
transformed, due to transmission errors to one of the four suppressed
codewords\ \*Q0000XX\*U, the received code word is set at\ \*Q111111\*U.
.nr PS 9
.RT
.ad r
\fBTable 5/G.722 [T5.722], p.\fR 
.sp 1P
.RT
.ad b
.RT
.ce
\fBH.T. [T6.722]\fR 
.ce
TABLE\ 6/G.722
.ce
\fBDecision levels and output codes for the 2\(hybit higher sub\(hyband\fR 
.ce
\fBquantizer\fR 
.ps 9
.vs 11
.nr VS 11
.nr PS 9
.TS
center box;
cw(36p) | cw(48p) | cw(48p) | cw(48p) | cw(48p) .
m H	HL	HH	IHN	IHP
_
.T&
cw(36p) | cw(48p) | cw(48p) | cw(48p) | cw(48p) .
1 2	0 1.10156	1.10156 \(if	01 00	11 10
_
.TE
.nr PS 9
.RT
.ad r
\fBTable 6/G.722 [T6.722], p.\fR 
.sp 1P
.RT
.ad b
.RT
.LP
.bp
.sp 2P
.LP
3.4
	\fIInverse adaptive quantizers\fR 
.sp 1P
.RT
.sp 1P
.LP
3.4.1
	\fIInverse adaptive quantizer in the lower sub\(hyband ADPCM encoder\fR 
.sp 9p
.RT
.PP
The 
lower sub\(hyband
output code, \fII\fR\d\fIL\fR\u(\fIn\fR ), is truncated by two bits to 
produce\ \fII\fR\d\fIL\fR\\d\fIt\fR\u(\fIn\fR ). The 4\(hybit 
codeword,\ \fII\fR\d\fIL\fR\\d\fIt\fR\u(\fIn\fR ), is converted to the 
truncated
quantized difference signal
, \fId\fR\d\fIL\fR\\d\fIt\fR\u(\fIn\fR ), using the
\fIQL\fR 4\uD\dlF261\u1\d output values of Table\ 7/G.722, and scaled by the 
scale
factor
,\ ?63
\fI\fI\d\fIL\fR\u(\fIn\fR ):
\v'6p'
.RT
.ad r
.ad b
.RT
.LP
where sgn [\fII\fR\d\fIL\fR\\d\fIt\fR\u(\fIn\fR )) is derived from the sign of
\fIe\fR\d\fIL\fR\u(\fIn\fR ) defined in Equation\ 3\(hy9.
.PP
There is a unique mapping, shown in Table 7/G.722, between four
adjacent 6\(hybit quantizer intervals and the\ \fIQL\fR 4\uD\dlF261\u1\d 
output values. 
\fIQL\fR 4\uD\dlF261\u1\d[\fII\fR\d\fIL\fR\\d\fIt\fR\u(\fIn\fR )] is determined 
in two steps: first determination of the 
quantizer interval index
,\ \fIm\fR\d\fIL\fR\u,
corresponding to\ \fII\fR\d\fIL\fR\u(\fIn\fR ) from Table\ 5/G.722, and then
determination of\ \fIQ\fR\d\fIL\fR\u | \uD\dlF261\u1\d(\fIm\fR\d\fIL\fR\u) 
by reference to 
Table\ 7/G.722.
.ce
\fBH.T. [T7.722]\fR 
.ce
TABLE\ 7/G.722
.ce
\fBOutput values and multipliers for 6, 5 and 4\(hybit lower sub\(hyband\fR 
.ce
\fBinverse quantizers\fR 
.ps 9
.vs 11
.nr VS 11
.nr PS 9
.TS
center box;
cw(36p) | cw(48p) | cw(48p) | cw(48p) | cw(48p) .
m L	QL6\uD\dlF261\u1\d	QL5\uD\dlF261\u1\d	QL4\uD\dlF261\u1\d	W L
_
.T&
cw(36p) | cw(48p) | cw(48p) | cw(48p) | cw(48p) .
\fB.\fR \ 1 \ 2	\fB.\fR 0.03409 0.10460	 {
\fB.\fR
0.06817
\fB.\fR
 }	 {
0.0000
\fB.\fR
\fB.\fR
 }	 {
\(em0.02930
\fB.\fR
\fB.\fR
 }
_
.T&
cw(36p) | cw(48p) | cw(48p) | cw(48p) | cw(48p) .
\ 3 \ 4 \ 5 \ 6	 {
0.17746
0.25300
0.33124
0.41259
 }	 {
0.21389
\fB.\fR
0.37035
\fB.\fR
 }	 {
\fB.\fR
0.29212
\fB.\fR
\fB.\fR
 }	 {
\fB.\fR
\(em0.01465
\fB.\fR
\fB.\fR
 }
_
.T&
cw(36p) | cw(48p) | cw(48p) | cw(48p) | cw(48p) .
\ 7 \ 8 \ 9 10	 {
0.49706
0.58518
0.67697
0.77310
 }	 {
0.53929
\fB.\fR
0.72286
\fB.\fR
 }	 {
\fB.\fR
0.63107
\fB.\fR
\fB.\fR
 }	 {
\fB.\fR
\ 0.02832
\fB.\fR
\fB.\fR
 }
_
.T&
cw(36p) | cw(48p) | cw(48p) | cw(48p) | cw(48p) .
11 12 13 14	 {
0.87359
0.97934
1.09036
1.20788
 }	 {
0.92383
\fB.\fR
1.14587
\fB.\fR
 }	 {
\fB.\fR
1.03485
\fB.\fR
\fB.\fR
 }	 {
\fB.\fR
\ 0.08398
\fB.\fR
\fB.\fR
 }
_
.T&
cw(36p) | cw(48p) | cw(48p) | cw(48p) | cw(48p) .
15 16 17 18	 {
1.33191
1.46415
1.60462
1.75585
 }	 {
1.39391
\fB.\fR
1.67486
\fB.\fR
 }	 {
\fB.\fR
1.53439
\fB.\fR
\fB.\fR
 }	 {
\fB.\fR
\ 0.16309
\fB.\fR
\fB.\fR
 }
_
.T&
cw(36p) | cw(48p) | cw(48p) | cw(48p) | cw(48p) .
19 20 21 22	 {
1.91782
2.09443
2.28568
2.49806
 }	 {
1.99880
\fB.\fR
2.38131
\fB.\fR
 }	 {
\fB.\fR
2.19006
\fB.\fR
\fB.\fR
 }	 {
\fB.\fR
\ 0.26270
\fB.\fR
\fB.\fR
 }
_
.T&
cw(36p) | cw(48p) | cw(48p) | cw(48p) | cw(48p) .
23 24 25 26	 {
2.73157
2.99827
3.29816
3.65804
 }	 {
2.84833
\fB.\fR
3.44811
\fB.\fR
 }	 {
\fB.\fR
3.14822
\fB.\fR
\fB.\fR
 }	 {
\fB.\fR
\ 0.58496
\fB.\fR
\fB.\fR
 }
_
.T&
cw(36p) | cw(48p) | cw(48p) | cw(48p) | cw(48p) .
27 28 29 30	 {
4.07789
4.64140
5.34856
6.05572
 }	 {
4.28782
\fB.\fR
5.70214
\fB.\fR
 }	 {
\fB.\fR
4.99498
\fB.\fR
\fB.\fR
 }	 {
\fB.\fR
\ 1.48535
\fB.\fR
\fB.\fR
 }
_
.TE
.nr PS 9
.RT
.ad r
\fBTable 7/G.722 [T7.722], p.\fR 
.sp 1P
.RT
.ad b
.RT
.LP
.bp
.sp 1P
.LP
3.4.2
	\fIInverse adaptive quantizer in the higher sub\(hyband ADPCM\fR 
\fIencoder\fR 
.sp 9p
.RT
.PP
The 
higher sub\(hyband
output code, \fII\fR\d\fIH\fR\u(\fIn\fR ) is converted to the quantized 
difference signal,\ \fId\fR\d\fIH\fR\u(\fIn\fR ), 
using the\ \fIQ\fR 2\uD\dlF261\u1\d output values of Table\ 8/G.722 and 
scaled by the scale factor,\ ?63\fI\fI\d\fIH\fR\u(\fIn\fR ): 
\v'6p'
.RT
.ad r
.ad b
.RT
.LP
where sgn[\fII\fR\d\fIH\fR\u(\fIn\fR )] is derived from the sign of
\fIe\fR\d\fIH\fR\u(\fIn\fR ) defined in Equation\ (3\(hy10), and where
\fIQ\fR\d2\u\uD\dlF261\u1\d[\fII\fR\d\fIH\fR\u(\fIn\fR )] is determined 
in two steps: first 
determine the quantizer interval index,\ \fIm\fR\d\fIH\fR\u, corresponding
to\ \fII\fR\d\fIH\fR\u(\fIn\fR ) from Table\ 6/G.722 and then determine
\fIQ\fR 2\uD\dlF261\u1\d(\fIm\fR\d\fIH\fR\u) by reference to Table\ 8/G.722.
.ce
\fBH.T. [T8.722]\fR 
.ce
TABLE\ 8/G.722
.ce
\fBOutput values and multipliers for the 2\(hybit higher sub\(hyband\fR 
.ce
\fBquantizer\fR 
.ps 9
.vs 11
.nr VS 11
.nr PS 9
.TS
center box;
cw(36p) | cw(48p) | cw(48p) .
m H	Q2\uD\dlF261\u1\d	W H
_
.T&
cw(36p) | cw(48p) | cw(48p) .
1	0.39453	\(em0.10449
.T&
cw(36p) | cw(48p) | cw(48p) .
2	1.80859	\ 0.38965
_
.TE
.nr PS 9
.RT
.ad r
\fBTableau 8/G.722 [T8.722],\fR 
.sp 1P
.RT
.ad b
.RT
.LP
.sp 1
.sp 1P
.LP
3.5
	\fIQuantizer adaptation\fR 
.sp 9p
.RT
.PP
This block defines ?63
\fI\fI\d\fIL\fR\u(\fIn\fR ) and
?63
\fI\fI\d\fIH\fR\u(\fIn\fR ), the scaling factors for the lower and higher
sub\(hyband quantizers. The scaling factors are updated in the log domain and
subsequently converted to a linear representation. For the lower sub\(hyband, 
the input is\ \fII\fR\d\fIL\fR\\d\fIt\fR\u(\fIn\fR ), the codeword truncated 
to preserve the four most significant bits. For the higher sub\(hyband, 
the 2\(hybit quantizer 
output, \fII\fR\d\fIH\fR\u(\fIn\fR ), is used directly.
.PP
Firstly the log scaling factors, ?63
\fI\fI\d\fIL\fR\u(\fIn\fR ) and
?63
\fI\fI\d\fIH\fR\u(\fIn\fR ), are updated as follows:
\v'6p'
.RT
.ad r
\fI\fI\fI\fR 
.ad b
.RT
.ad r
.ad b
.RT
.LP
where \fIW\fR\d\fIL\fR\uand \fIW\fR\d\fIH\fR\uare logarithmic scaling factors 
multipliers given in Tables\ 7/G.722 and\ 8/G.722, and B is a 
leakage constant
equal
to\ 127/128.
.PP
Then the 
log scaling factors
are limited, according
to:
\v'6p'
.ad r
.ad b
.RT
.ad r
.ad b
.RT
.PP
Finally, the 
linear scaling factors
are computed from the
log scaling factors, using an approximation of the inverse log\d2\ufunction:
\v'6p'
.ad r
.ad b
.RT
.ad r
.ad b
.RT
.LP
where ?63
\fI\fI\d\fIm\fR\\d\fIi\fR\\d\fIn\fR\uis equal to half the quantizer step 
size of the 
14\ bit analogue\(hyto\(hydigital converter.
.bp
.sp 2P
.LP
3.6
	\fIAdaptive prediction\fR 
.sp 1P
.RT
.sp 1P
.LP
3.6.1
	\fIPredicted value computations\fR 
.sp 9p
.RT
.PP
The adaptive predictors compute predicted signal values,
\fIs\fR\d\fIL\fR\u(\fIn\fR ) and\ \fIs\fR\d\fIH\fR\u(\fIn\fR ), for the 
lower and higher sub\(hybands respectively. 
.PP
Each 
adaptive predictor
comprises two sections: a second\(hyorder section that models poles, and 
a sixth\(hyorder section that models zeroes in the input signal. 
.PP
The second order 
pole sections
(coefficients \fIa\fR\d\fIL\fR\\d,\u\fI\d\fIi\fR\uand\ \fIa\fR\d\fIH\fR\\d,\u\fIi\fR 
) use the quantized reconstructed signals, 
\fIr\fRL\fI\d\fIt\fR\u(\fIn\fR ) and\ \fIr\fR\d\fIH\fR\u(\fIn\fR ), for 
prediction. The 
sixth order 
zero sections
(coefficients\ \fIb\fR\d\fIL\fR\\d,\u\fI\fI\d\fIi\fR\u) and
\fIb\fR\d\fIH\fR\\d,\u\fIi\fR ) use the quantized difference
signals,\ \fId\fR\d\fIL\fR\\d\fIt\fR\u(\fIn\fR ) and\ \fId\fR\d\fIH\fR\u(\fIn\fR 
). The 
zero\(hybased predicted signals,\ \fIs\fR\d\fIL\fR\\d\fIz\fR\u(\fIn\fR )
and\ \fIs\fR\d\fIH\fR\\d\fIz\fR\u(\fIn\fR ), are also employed to compute 
partially 
reconstructed signals as described in \(sc\ 3.6.2.
.PP
Firstly, the outputs of the pole sections are computed as
follows:
\v'6p'
.RT
.ad r
.ad b
.RT
.ad r
.ad b
.RT
.PP
Similarly, the outputs of the zero sections are computed as
follows:
\v'6p'
.ad r
.ad b
.RT
.ad r
.ad b
.RT
.PP
Then, the intermediate predicted values are summed to produce the predicted 
signal values: 
\v'6p'
.ad r
.ad b
.RT
.ad r
.ad b
.RT
.sp 1P
.LP
3.6.2
	\fIReconstructed signal computation\fR 
.sp 9p
.RT
.PP
The quantized reconstructed signals, \fIr\fR\d\fIL\fR\\d\fIt\fR\u(\fIn\fR 
) and\ \fIr\fR\d\fIH\fR\u(\fIn\fR ), are computed as follows: 
\v'6p'
.RT
.ad r
.ad b
.RT
.ad r
.ad b
.RT
.PP
The partially reconstructed signals, \fIp\fR\d\fIL\fR\\d\fIt\fR\u(\fIn\fR 
) and \fIp\dH\u\fR (\fIn\fR ), used for the pole section adaptation, are 
then 
computed:
\v'6p'
.ad r
.ad b
.RT
.ad r
.ad b
.RT
.LP
.bp
.sp 1P
.LP
3.6.3
	\fIPole section adaptation\fR 
.sp 9p
.RT
.PP
The second order pole section is adapted by updating the
coefficients,\ \fIa\fR\d\fIL\fR\\d,\\d1\u, \fIa\fR\d\fIH\fR\\d,\\d1\u, 
\fIa\fR\d\fIH\fR\\d,\\d2\u, using a simplified gradient algorithm: 
\v'6p'
.RT
.ad r
.ad b
.RT
.ad r
.ad b
.RT
.LP
where
\v'6p'
.ad r
.ad b
.RT
.ad r
.ad b
.RT
.LP
with
\v'6p'
.ad r
.ad b
.RT
.LP
and
\v'6p'
.ad r
.ad b
.RT
.PP
Then the following stability constraints are imposed:
\v'6p'
.ad r
.ad b
.RT
.ad r
.ad b
.RT
.PP
\fIa\fR\d\fIH\fR\\d,\\d1\u(\fIn\fR ) and \fIa\fR\d\fIH\fR\\d,\\d2\u(\fIn\fR 
) are 
similarly computed, replacing\ \fIa\fR\d\fIL\fR\\d,\\d1\u(\fIn\fR ),
\fIa\fR\d\fIL\fR\\d,\\d2\u(\fIn\fR ) and \fIP\fR\d\fIL\fR\\d\fIt\fR\u | \fIn\fR 
) 
by\ \fIa\fR\d\fIH\fR\\d,\\d1\u(\fIn\fR ), \fIa\fR\d\fIH\fR\\d,\\d2\u(\fIn\fR )
and\ \fIP\fR\d\fIH\fR\u(\fIn\fR ), respectively.
.sp 1P
.LP
3.6.4
	\fIZero section adaptation\fR 
.sp 9p
.RT
.PP
The sixth order zero predictor is adapted by updating the
coefficients\ \fIb\fR\d\fIL\fR\\d,\u\fI\fI\d\fIi\fR\uand\ \fIb\fR\d\fIH\fR\\d,\u\fI\fI\d\fIi\fR\uusing 
a simplified 
gradient algorithm:
\v'6p'
.RT
.ad r
.ad b
.RT
.LP
for \fIi\fR = 1, 2\ . |  | \ 6
.LP
and with
.ad r
.ad b
.RT
.LP
where \fIb\fR\d\fIL\fR\\d,\u\fI\fI\d\fIi\fR\u | \fIn\fR ) is implictly 
limited to \(+- | . 
.PP
\fIb\fR\d\fIH\fR\\d,\u\fI\fI\d\fIi\fR\u(\fIn\fR ) are similarly updated,
replacing\ \fIb\fR\d\fIL\fR\\d,\u\fI\fI\d\fIi\fR\u(\fIn\fR ) and\ \fId\fR\d\fIL\fR\\d\fIt\fR\u(\fIn\fR 
) by\ \fIb\fR\d\fIH\fR\\d,\u\fI\fI\d\fIi\fR\u(\fIn\fR ) and\ \fId\fR\d\fIH\fR\u(\fIn\fR 
) 
respectively.
.sp 2P
.LP
\fB4\fR 	\fBSB\(hyADPCM decoder principles\fR 
.sp 1P
.RT
.PP
A block diagram of the SB\(hyADPCM decoder is given in Figure 6/G.722 and 
block diagrams of the lower and higher sub\(hyband ADPCM decoders are given 
respectively in Figures\ 7/G.722 and\ 8/G.722.
.PP
The input to the lower sub\(hyband ADPCM decoder, \fII\fR\d\fIL\fR\\d\fIr\fR\u, 
may differ from\ \fII\fR\d\fIL\fR\ueven in the absence of transmission 
errors, in that one 
or two least significant bits may have been replaced by data.
.bp
.RT
.sp 2P
.LP
4.1
	\fIInverse adaptive quantizer\fR 
.sp 1P
.RT
.sp 1P
.LP
4.1.1
	\fIInverse adaptive quantizer selection for the lower sub\(hyband\fR 
\fIADPCM decoder\fR 
.sp 9p
.RT
.PP
According to the received indication of the mode of operation the number 
of least significant bits which should be truncated from the input 
codeword\ \fII\fR\d\fIL\fR\\d\fIr\fR\u, and the choice of the inverse adaptive 
quantizer 
are determined, as shown in Table\ 2/G.722.
.PP
For operation in mode 1, the 6\(hybit codeword,
\fII\fR\d\fIL\fR\\d\fIr\fR\u(\fIn\fR ), is converted to the quantized difference, 
\fId\fR\d\fIL\fR\u(\fIn\fR ), according to\ \fIQL\fR 6\uD\dlF261\u1\d output 
values of 
Table\ 7/G.722, and scaled by the scale factor,
\ ?63
\fI\fI\d\fIL\fR\u(\fIn\fR ):
\v'6p'
.RT
.ad r
.ad b
.RT
.LP
where sgn[\fII\fR\d\fIL\fR\\d\fIr\fR\u(\fIn\fR )] is derived from the sign of
\fII\fR\d\fIL\fR\u(\fIn\fR ) defined in equation (3\(hy9).
.PP
Similarly, for operations in mode 2 or mode 3, the 
truncated   codeword
(by one or two bits) is converted to the quantized difference
signal, \fId\fR\d\fIL\fR\u(\fIn\fR ), according to\ \fIQL\fR 5\uD\dlF261\u1\d\fR 
or\ \fIQL\fR 4\uD\dlF261\u1\d\fR output values of Table\ 7/G.722 respectively.
.PP
\fR There are unique mappings, shown in Table 7/G.722, between two or
four adjacent 6\(hybit quantizer intervals and the\ \fIQL\fR 5\uD\dlF261\u1\d 
or\ \fIQL\fR 4\uD\dlF261\u1\d output values respectively. 
.PP
In the computations above, the output values are determined in two
steps: first determination of the quantizer interval index,\ \fIm\fR\d\fIL\fR\u, 
corresponding to\ \fII\fR\d\fIL\fR\\d\fIr\fR\u(\fIn\fR ) from Table\ 5/G.722, 
and then 
determination of the output values corresponding to\ \fIm\fR\d\fIL\fR\uby 
reference to 
Table\ 7/G.722.
.PP
The inverse adaptive quantizer, used for the computation of the
predicted value and for adaptation of the quantizer and predictor, is described 
in \(sc\ 3.4.1, but with\ \fII\fR\d\fIL\fR\u(\fIn\fR ) replaced 
by\ \fII\fR\d\fIL\fR\\d\fIr\fR\u(\fIn\fR ).
.RT
.sp 1P
.LP
4.1.2
	\fIInverse adaptive quantizer for the higher sub\(hyband ADPCM\fR 
\fIdecoder\fR 
.sp 9p
.RT
.PP
See \(sc 3.4.2.
.RT
.sp 1P
.LP
4.2
	\fIQuantizer adaptation\fR 
.sp 9p
.RT
.PP
See \(sc 3.5.
.RT
.sp 2P
.LP
4.3
	\fIAdaptive prediction\fR 
.sp 1P
.RT
.sp 1P
.LP
4.3.1
	\fIPredicted value computation\fR 
.sp 9p
.RT
.PP
See \(sc 3.6.1.
.RT
.sp 1P
.LP
4.3.2
	\fIReconstructed signal computation\fR 
.sp 9p
.RT
.PP
See \(sc 3.6.2.
.PP
The output reconstructed signal for the lower sub\(hyband ADPCM
decoder,\ \fIr\fR\d\fIL\fR\u(\fIn\fR ), is computed from the quantized 
difference 
signal,\ \fId\fR\d\fIL\fR\u(\fIn\fR ), as follows:
\v'6p'
.RT
.ad r
.ad b
.RT
.sp 1P
.LP
4.3.3
	\fIPole section adaptation\fR 
.sp 9p
.RT
.PP
See \(sc 3.6.3.
.RT
.sp 1P
.LP
4.3.4
	\fIZero section adaptation\fR 
.sp 9p
.RT
.PP
See \(sc 3.6.4.
.RT
.sp 1P
.LP
4.4
	\fIReceive QMF\fR 
.sp 9p
.RT
.PP
A 24\(hycoefficient QMF is used to reconstruct the output signal,
\fIx\fR\do\\du\\dt\u(
\fIj\fR ), from the reconstructed lower and
higher sub\(hyband
signals,\ \fIr\fR\d\fIL\fR\u(\fIn\fR ) and\ \fIr\fR\d\fIH\fR\u(\fIn\fR 
). The QMF 
coefficient values,\ \fIh\fR\d\fIi\fR\u, are the same as those used in 
the transmit QMF and are given in Table\ 4/G.722. 
.bp
.PP
The output signals, \fIx\fR\do\\du\\dt\u(
\fIj\fR ) and
\fIx\fR\do\\du\\dt\u(
\fIj\fR \ +\ 1), are computed in the following
way:
\v'6p'
.RT
.ad r
.ad b
.RT
.ad r
.ad b
.RT
.LP
where
\v'6p'
.ad r
.ad b
.RT
.ad r
.ad b
.RT
.sp 2P
.LP
\fB5\fR 	\fBComputational details for QMF\fR 
.sp 1P
.RT
.sp 1P
.LP
5.1
	\fIInput and output signals\fR 
.sp 9p
.RT
.PP
Table 9/G.722 defines the input and output signals for the transmit and 
receive\ QMF. All input and output signals have 16\(hybit\ word lengths, 
which are limited to a range of \(em16384 to\ 16383 in\ 2's complement 
notation. Note that the most significant magnitude bit of the\ A/D output 
and the\ D/A input appears at the third bit location in\ XIN and\ XOUT, 
respectively. 
.RT
.ce
\fBH.T. [T9.722]\fR 
.ce
TABLE\ 9/G.722
.ce
\fBRepresentation of input and output signals\fR 
.ps 9
.vs 11
.nr VS 11
.nr PS 9
.TS
center box;
cw(228p) .
Transmit QMF
.TE
.TS
center box;
lw(36p) | cw(24p) | cw(84p) | cw(84p) .
	Name	Binary representation	Description
_
.T&
lw(36p) | lw(24p) | cw(84p) | lw(84p) .
Input	XIN	 {
S, S, \(em2, \(em3, . |  | , \(em14, \(em15
 }	 {
Input value
(uniformly quantized)
 }
.T&
lw(36p) | lw(24p) | cw(84p) | lw(84p) .
Output	XL	 {
S, S, \(em2, \(em3, . |  | , \(em14, \(em15
 }	 {
Output signal for lower sub\(hyband encoder
 }
.T&
lw(36p) | lw(24p) | cw(84p) | lw(84p) .
Output	XH	 {
S, S, \(em2, \(em3, . |  | , \(em14, \(em15
 }	 {
Output signal for higher sub\(hyband encoder
 }
_
.T&
cw(228p) .
Receive QMF
_
.T&
lw(36p) | cw(24p) | cw(84p) | cw(84p) .
	Name	Binary representation	Description
_
.T&
lw(36p) | lw(24p) | cw(84p) | lw(84p) .
Input	RL	 {
S, S, \(em2, \(em3, . |  | , \(em14, \(em15
 }	 {
Lower sub\(hyband reconstructed signal
 }
.T&
lw(36p) | lw(24p) | cw(84p) | lw(84p) .
Input	RH	 {
S, S, \(em2, \(em3, . |  | , \(em14, \(em15
 }	 {
Higher sub\(hyband reconstructed signal
 }
.T&
lw(36p) | lw(24p) | cw(84p) | lw(84p) .
Output	XOUT	 {
S, S, \(em2, \(em3, . |  | , \(em14, \(em15
 }	 {
Output value
(uniformly quantized)
 }
.TE
.LP
\fINote\fR
\ \(em\ XIN and XOUT are represented in a sign\(hyextended 15\(hybit format,
where the LSB is set to \*Q0\*U for 14\(hybit converters.
.nr PS 9
.RT
.ad r
\fBTable 9/G.722 [T9.722], p.\fR 
.sp 1P
.RT
.ad b
.RT
.LP
.bp
.sp 1P
.LP
5.2
	\fIDescription of variables and detailed specification of sub\(hyblocks\fR 
.sp 9p
.RT
.PP
This section contains a detailed expansion of the transmit and
receive\ QMF. The expansions are illustrated in Figures\ 17/G.722 and\ 18/G.722
with the internal variables given in Table\ 10/G.722, and the\ QMF coefficients 
given in Table\ 11/G.722. The word lengths of internal variables, XA, XB 
and\ WD must be equal to or greater than 24\ bits (see Note). The other 
internal 
variables have a minimum of 16 bit word lengths. A brief functional description 
and the full specification is given for each sub\(hyblock. 
.PP
\fR The notations used in the block descriptions are as
follows:
.RT
.LP
	>
> | 
	denotes an \fIn\fR \(hybit arithmetic shift right operation
(sign extension),
.LP
	+
	denotes arithmetic addition with saturation control which
forces the result to the minimum or maximum representable value
in case of underflow or overflow, respectively,
.LP
	\(em
	denotes arithmetic subtraction with saturation control
which forces the result to the minimum or maximum representable
value in case of underflow or overflow, respectively.
.LP
	*
	denotes arithmetic multiplication which can be performed
with either truncation or rounding,
.LP
	<
	denotes the \*Qless than\*U condition as \fIx\fR < | fIy\fR ; \fIx\fR is less
than \fIy\fR ,
.LP
	>
	denotes the \*Qgreater than\*U condition, as \fIx\fR > | fIy\fR ; \fIx\fR is
greater than \fIy\fR ,
.LP
	=
	denotes the substitution of the right\(hyhand variable for the
left\(hyhand variable.
.PP
\fINote\ 1\fR \ \(em\ Some freedom is offered for the implementation of
the accumulation process in the QMF: the word lengths of the internal variables 
can be equal to or greater than 24\ bits, and the arithmetic multiplications 
can be performed with either truncation or rounding. It allows a simplified 
implementation on various types of processors. The counterpart is that it
excludes the use of digital test sequence for the test of the QMF.
.ce
\fBH.T. [T10.722]\fR 
.ce
TABLE\ 10/G.722
.ce
\fBRepresentation of internal processing variables and QMF\fR 
.ce
\fBcoefficients\fR 
.ps 9
.vs 11
.nr VS 11
.nr PS 9
.TS
center box;
cw(228p) .
Transmit QMF
_
.T&
cw(54p) | cw(84p) | cw(90p) .
Name	Binary representation	Description
_
.T&
lw(54p) | lw(84p) | lw(90p) .
XA	 {
S, \(em1, \(em2, \(em3, . |  | , \(emy+1, \(emy
 }	 {
Output signal of sub\(hyblock, ACCUMA
 }
.T&
lw(54p) | lw(84p) | lw(90p) .
XB	 {
S, \(em1, \(em2, \(em3, . |  | , \(emy+1, \(emy
 }	 {
Output signal of sub\(hyblock, ACCUMB
 }
.T&
lw(54p) | lw(84p) | lw(90p) .
XIN1, XIN2, . |  | , XIN23	 {
S, \ S, \(em2, \(em3, . |  | , \(em14, \(em15
 }	 {
Input signal with delays 1 to 23
 }
_
.T&
cw(228p) .
Receive QMF
_
.T&
cw(54p) | cw(84p) | cw(90p) .
Name	Binary representation	Description
_
.T&
lw(54p) | lw(84p) | lw(90p) .
XD, XD1, . |  | , XD11	 {
S, \(em1, \(em2, \(em3, . |  | , \(em14, \(em15
 }	 {
Input signal for sub\(hyblock, ACCUMC, with delays 0 to 11
 }
.T&
lw(54p) | lw(84p) | lw(90p) .
XOUT1	 {
S, \ S, \(em2, \(em3, . |  | , \(em14, \(em15
 }	8 kHz sampled output value
.T&
lw(54p) | lw(84p) | lw(90p) .
XOUT2	 {
S, \ S, \(em2, \(em3, . |  | , \(em14, \(em15
 }	8 kHz sampled output value
.T&
lw(54p) | lw(84p) | lw(90p) .
XS, XS1, . |  | , XS11	 {
S, \(em1, \(em2, \(em3, . |  | , \(em14, \(em15
 }	 {
Input signal for sub\(hyblock, ACCUMD, with delays 0 to 11
 }
.T&
lw(54p) | lw(84p) | lw(90p) .
WD	 {
S, \(em1, \(em2, \(em3, . |  | , \(emy+1, \(emy
 }	Partial sum
_
.T&
cw(228p) .
QMF coefficients
_
.T&
cw(54p) | cw(84p) | cw(90p) .
Name	Binary representation	Description
_
.T&
lw(54p) | lw(84p) | lw(90p) .
H0, H1, . |  | , H23	 {
S, \(em2, \(em3, \(em4, . |  | , \(em12, \(em13
 }	 {
Filter coefficient values
 }
.TE
.LP
\fINote\fR
\ \(em\ y is equal to or greater than 23.
.nr PS 9
.RT
.ad r
\fBTable 10/G.722 [T10.722], p.\fR 
.sp 1P
.RT
.ad b
.RT
.LP
.bp
.ce
\fBH.T. [T11.722]\fR 
.ce
TABLE\ 11/G.722
.ce
\fBQMF coefficient\fR 
.ps 9
.vs 11
.nr VS 11
.nr PS 9
.TS
center box;
cw(84p) | cw(60p) .
Coefficient	Scaled values  (see Note)
_
.T&
cw(84p) | cw(60p) .
H0\  , H23	\ \ \ 3
.T&
cw(84p) | cw(60p) .
H1\  , H22	\ \(em11
.T&
cw(84p) | cw(60p) .
H2\  , H21	\ \(em11
.T&
cw(84p) | cw(60p) .
H3\  , H20	\ \ 53
.T&
cw(84p) | cw(60p) .
H4\  , H19	\ \ 12
.T&
cw(84p) | cw(60p) .
H5\  , H18	\(em156
.T&
cw(84p) | cw(60p) .
H6\  , H17	\ \ 32
.T&
cw(84p) | cw(60p) .
H7\  , H16	\ 362
.T&
cw(84p) | cw(60p) .
H8\  , H15	\(em210
.T&
cw(84p) | cw(60p) .
H9\  , H14	\(em805
.T&
cw(84p) | cw(60p) .
H10 , H13	\ 951
.T&
cw(84p) | cw(60p) .
H11 , H12	\ 3876
.TE
.LP
\fINote\fR
\ \(em\ QMF coefficients are scaled by 2\u1\d\u3\d with respect to the
representation specified in Table\ 10/G.722.
.nr PS 9
.RT
.ad r
\fBTable 11/G.722 [T11.722], p.\fR 
.sp 1P
.RT
.ad b
.RT
.sp 1P
.LP
5.2.1
	\fIDescription of the transmit QMF\fR 
.sp 9p
.RT
.LP
.rs
.sp 20P
.ad r
\fBFigure 17/G.722, p.\fR 
.sp 1P
.RT
.ad b
.RT
.ad r
.ad b
.RT
.LP
.bp
.sp 1P
.ce 1000
DELAYX
.sp 9p
.RT
.ce 0
.sp 1P
.LP
Input:
	x
.LP
Output:
	y
.LP
\fINote\fR \ \(em\ Index (
\fIj\fR ) indicates the current 16\(hykHz sample period,
while index (
\fIj\fR \ \(em\ 1) indicates the previous one.
.LP
Function:
	Memory block. For any input x, the output is given by:
.LP
\fIy\fR (
\fIj\fR ) = \fIx\fR (
\fIj\fR \ \(em\ 1)
.ad r
.ad b
.RT
.sp 1P
.ce 1000
ACCUMA
.sp 9p
.RT
.ce 0
.sp 1P
.LP
Inputs:
	XIN, XIN2, XIN4, . |  | , XIN22
.LP
Output:
	XA
.LP
\fINote\ 1\fR \ \(em\ H0, H2, . |  | , H22 are obtained from Table 11/G.722.
.LP
\fINote\ 2\fR \ \(em\ The values XIN, XIN2, . |  | , XIN22 and H0, H2, 
. |  | , H22 may be 
shifted before multiplication, if so desired. The result\ XA must be rescaled
accordingly, In performing these scaling operations the following rules 
must be obeyed: 
.LP
	1)
	 the precision of XIN, XIN2, . |  | , XIN22 and H0, H2, . |  | , H22 as 
given in Table\ 9/G.722 and Table\ 10/G.722 must be retained, 
.LP
	2)
	the partial products and the ouptut signal XA must be
retained to a significance of at least\ 2\uD\dlF261\u2\d\u3\d,
.LP
	3)
	no saturation should occur in the calculation of the
function\ XA.
.LP
\fINote\ 3\fR \ \(em\ No order of summation is specified in accumulating 
the partial 
products.
.LP
Function:
	Multiply the even order QMF coefficients by the appropriately
delayed input signals, and accumulate these products.
.LP
XA = (XIN 
*
H0) + (XIN2 
*
H2) + (XIN4 
*
H4) + . |  |   + (XIN22 
*
H22)
.ad r
.ad b
.RT
.sp 1P
.ce 1000
ACCUMB
.sp 9p
.RT
.ce 0
.sp 1P
.LP
Inputs:
	XIN1, XIN3, XIN5, . |  | , XIN23
.LP
Output:
	XB
.LP
\fINote\ 1\fR \ \(em\ H1, H3, . |  | , H23 are obtained from Table 11/G.722.
.LP
\fINote\ 2\fR \ \(em\ The values XIN1, XIN3, . |  | , XIN23 and H1, H3, 
. |  | , H23 may be 
shifted before multiplication, if so desired. The result\ XB must be rescaled
accordingly. In performing these scaling operations the following rules 
must be obeyed: 
.LP
	1)
	 the precision of XIN1, XIN3, . |  | , XIN23 and H1, H3, . |  | , H23 
as given in Table\ 9/G.722 and Table\ 10/G.722 must be retained, 
.LP
	2)
	the partial products and the output signal X3 must be
retained to a significance of at least\ 2\uD\dlF261\u2\d\u3\d,
.LP
	3)
	no saturation should occur in the calculation of the
function\ XB.
.LP
\fINote\ 3\fR \ \(em\ No order of summation is specified in accumulating 
the partial 
products.
.LP
Function:
	Multiply the odd order QMF coefficients by the appropriately
delayed input signals, and accumulate these products.
.LP
XB = (XIN1 
*
H1) + (XIN3 
*
H3) + (XIN5 
*
H5) + . |  |  + (XIN23 
*
H23)
.ad r
.ad b
.RT
.LP
.bp
.sp 1P
.ce 1000
LOWT
.sp 9p
.RT
.ce 0
.sp 1P
.LP
Inputs:
	XA, XB
.LP
Output:
	XL
.LP
Function:
	Compute the lower sub\(hyband signal component.
.LP
XL = (XA + XB) >
> (y \(em 15)
[Formula Deleted]
.LP