ISO/IEC 14496-3:2001/Amd 1:2003/Cor 1:2004

Information technology — Coding of audio-visual objects — Part 3: Audio — Amendment 1: Bandwidth extension — Technical Corrigendum 1

Technologies de l'information — Codage des objets audiovisuels — Partie 3: Codage audio — Amendement 1: Extension de largeur de bande — Rectificatif technique 1

General Information

• Status: Withdrawn
• Publication Date: 13-Jun-2004
• Withdrawal Date: 13-Jun-2004
• ICS: 35.040 - Information coding
• ICS: 35.040.40 - Coding of audio, video, multimedia and hypermedia information
• Technical Committee: ISO/IEC JTC 1/SC 29 - Coding of audio, picture, multimedia and hypermedia information
• Drafting Committee: ISO/IEC JTC 1/SC 29 - Coding of audio, picture, multimedia and hypermedia information
• Current Stage: 9599 - Withdrawal of International Standard
• Completion Date: 21-Apr-2006

Relations

• Corrects: ISO/IEC 14496-3:2001/Amd 1:2003 - Information technology — Coding of audio-visual objects — Part 3: Audio — Amendment 1: Bandwidth extension
• Effective Date: 06-Jun-2022
• Child: ISO/IEC 14496-3:2001/Amd 1:2003 - Information technology — Coding of audio-visual objects — Part 3: Audio — Amendment 1: Bandwidth extension
• Effective Date: 15-Apr-2008
• Consolidated By: ISO 19009:2015 - Small craft — Electric navigation lights — Performance of LED lights
• Effective Date: 06-Jun-2022

Standards Content (Sample)

INTERNATIONAL STANDARD
ISO/IEC 14496-3:2001/Amd.1:2003/Cor.1:2004(E)

TECHNICAL CORRIGENDUM 1
Published 2004-06-01
INTERNATIONAL ORGANIZATION FOR STANDARDIZATION • МЕЖДУНАРОДНАЯ ОРГАНИЗАЦИЯ ПО СТАНДАРТИЗАЦИИ • ORGANISATION INTERNATIONALE DE NORMALISATION
INTERNATIONAL ELECTROTECHNICAL COMMISSION • МЕЖДУНАРОДНАЯ ЭЛЕКТРОТЕХНИЧЕСКАЯ КОМИССИЯ • COMMISSION ÉLECTROTECHNIQUE INTERNATIONALE


Information technology — Coding of audio-visual objects —
Part 3:
Audio
AMENDMENT 1: Bandwidth extension
TECHNICAL CORRIGENDUM 1
Technologies de l'information — Codage des objets audiovisuels —
Partie 3: Codage audio
AMENDEMENT 1: Extension de largeur de bande
RECTIFICATIF TECHNIQUE 1
Technical Corrigendum 1 to ISO/IEC 14496-3:2001/Amd.1:2003 was prepared by Joint Technical Committee
ISO/IEC JTC 1, Information technology, Subcommittee SC 29, Coding of audio, picture, multimedia and
hypermedia information.

NOTE This document specifies the first corrigendum to ISO/IEC 14496-3:2001/Amd.1:2003. With the exception of the
correction for Table 4.54 which is only relevant for multi-channel decoding of ER streams, all the corrections strive to
correct errors in the standard text and thus aligning the standard text with the reference software.

ICS 35.040 Ref. No. ISO/IEC 14496-3:2001/Amd.1:2003/Cor.1:2004(E)
©  ISO/IEC 2004 – All rights reserved
Published in Switzerland

---------------------- Page: 1 ----------------------
ISO/IEC 14496-3:2001/Amd.1:2003/Cor.1:2004(E)

In ISO/IEC 14496-3:2001/Amd.1, 4.4.2.8 Payloads for the audio object type SBR, replace Table 4.54 by the
table below (modified parts are marked in gray):
Table 4.54A – Syntax of sbr_extension_data()
Syntax No. of bits Mnemonic
sbr_extension_data(id_aac, crc_flag)
{
num_sbr_bits = 0;

if (crc_flag) {
 bs_sbr_crc_bits; 10 uimsbf
 num_sbr_bits += 10;
}

if (sbr_layer != SBR_STEREO_ENHANCE) { Note 1
 num_sbr_bits += 1;
 if (bs_header_flag) 1
  num_sbr_bits += sbr_header(); Note 2
}

num_sbr_bits += sbr_data(id_aac, bs_amp_res); Note 2

num_align_bits = (8*cnt - 4 - num_sbr_bits) % 8;
bs_fill_bits; num_align uimsbf
_bits

return ((num_sbr_bits + num_align_bits + 4) / 8)
}
Note 1: When the SBR tool is used with a non-scalable AAC core coder, the value of the helper variable
sbr_layer is SBR_NOT_SCALABLE. When the SBR tool is used with a scalable AAC core coder, the value of
the helper variable sbr_layer depends on the current layer and the scalability configuration of the AAC core
coder as defined in Table 4.86 in subclause 4.5.2.8.2.4.
Note 2: sbr_header() and sbr_data() return the number of bits read (cnt is a parameter in
extension_payload()).


In ISO/IEC 14496-3:2001/Amd.1:2003, subclause 4.4.2.8 Payloads for the Audio Object Type SBR, Table
4.55A, replace:
Note 3: If this bit is not set the default values for the underlying bitstream elements should be used.
by:
2 © ISO/IEC 2004 — All rights reserved

---------------------- Page: 2 ----------------------
ISO/IEC 14496-3:2001/Amd.1:2003/Cor.1:2004(E)
Note 3: If this bit is not set the default values for the underlying bitstream elements shall be used disregarded
any previous value.

In ISO/IEC 14496-3:2001/Amd.1, 4.4.2.8 Payloads for the audio object type SBR, Table 4.58, Table 4.59, and
Table 4.60, add the part marked in gray:
if (bs_extended_data) { 1
 cnt = bs_extension_size; 4 uimsbf
 if (cnt == 15)
  cnt += bs_esc_count; 8 uimsbf

 num_bits_left = 8 * cnt;
 while (num_bits_left > 7) {
  bs_extension_id; 2 uimsbf
  num_bits_left -= 2;
  sbr_extension(bs_extension_id, num_bits_left); Note 1
 }
 bs_fill_bits num_bits_left
}

In ISO/IEC 14496-3:2001/Amd.1, subclause 4.5.2.8.2.3 SBR Extension Payload for the Audio Object Types
ER AAC LC and ER AAC LTP replace:
The extension payload shall not include both DRC extension elements and SBR extension elements
simultaneously. If SBR extension elements are used, DRC extension elements are prohibited.
by:
The SBR extension elements shall be placed after any other extension elements.

In ISO/IEC 14496-3:2001/Amd.1, subclause 4.5.2.8.2.4 SBR Extension Payload for the Audio Object Types
AAC Scalable and ER AAC Scalable, replace:
The scalable SBR data is embedded into the MPEG-4 stream in the same way as for non-scalable SBR data
elements, by means of using the extension_payload().
by:
The scalable SBR data is embedded into the MPEG-4 stream in the same way as for non-scalable SBR data
elements, by means of using the extension_payload(). The SBR extension elements shall be placed after any
other extension elements.

In ISO/IEC 14496-3:2001/Amd.1, subclause 4.6.18.2.2 SBR specific notation, replace:
QQml−=sb,,l ikl,f (i)≤m ( ) ( ())
Mapped Orig TableNoise TableNoise Q E
where kl is defined by ttlk≥+l,1tl≤tkl+1
() () ( ()) ( ) ( () )
QQ
should be interpreted as follows. Q ml−sb,l equals Q ik, l
( ) ( ())
Mapped Orig
© ISO/IEC 2004 — All rights reserved 3

---------------------- Page: 3 ----------------------
ISO/IEC 14496-3:2001/Amd.1:2003/Cor.1:2004(E)
by:
QQmk−=,,l ikl,f (i)≤m ( ) ()
( )
Mapped x Orig TableNoise TableNoise Q E
where kl() is defined by tt(lk)≥+(l),1t(l )≤tk(l)+1
( ) ( )
QQ
should be interpreted as follows. Q mk− ,l equals Q ik, l
( ) ()
( )
Mapped x Orig

In ISO/IEC 14496-3:2001/Amd.1, subclause 4.6.18.2.5 Constants, replace:
ε A constant to avoid division by zero, e.g. 96 dB below maximum signal
input.
by:
ε=1 A constant to avoid division by zero, e.g. 96 dB below maximum signal
input.

In ISO/IEC 14496-3:2001/Amd.1, subclause 4.6.18.2.6, add:
W is the subband matrix where the QMF filtered subband samples are stored.

In ISO/IEC 14496-3:2001/Amd.1, subclause 4.6.18.4 and 4.6.18.8.2, replace all instances (in text flowcharts
and pictures) of
X
Low
by:
W

4 © ISO/IEC 2004 — All rights reserved

---------------------- Page: 4 ----------------------
ISO/IEC 14496-3:2001/Amd.1:2003/Cor.1:2004(E)
In ISO/IEC 14496-3:2001/Amd.1, subclause 4.6.18.4.3 Down sampled synthesis filterbank, in Figure 4.44,
where it says "127", it should say "127.5", hence replace it by the following figure:
Start
( for QMF subsample l )
for( n = 639; n >= 64; n--) {
v[n] = v[n -64]
}
for( n = 0; n <= 63; n++) {
v[n] = Real( X[0][l] / 64 * exp( i * π / 64 * ( 0.5 ) * ( 2 * n - 127.5 ) ))
for( k = 1; k <= 31; k++) {
v[n] = v[n] + Real( X[k][l] /64 * exp( i * π / 64 * ( k + 0.5 )*(2 *n -127.5) ))
}
}
for( n = 0; n <= 4; n++) {
for( k = 0; k <= 31; k++) {
g[64 * n + k]      = v[128 * n + k]
g[64 * n + 32 + k]  =  v[128 * n + 96 + k]
}
}
for( n = 0; n <= 319; n++) {
w[n] = g[n] * c[2*n]
}
for( k = 0; k <= 31; k++) {
temp = w[k]
for( n = 1; n <= 9; n++) {
temp = temp + w[32*n + k]
}
nextOutputAudioSample = temp
}
Done

© ISO/IEC 2004 — All rights reserved 5

---------------------- Page: 5 ----------------------
ISO/IEC 14496-3:2001/Amd.1:2003/Cor.1:2004(E)
In ISO/IEC 14496-3:2001/Amd.1, subclause 4.6.18.5 SBR tool overview, replace:
start-up l , lsb' and bsco' are set to zero. And where
Temp
bsco=⋅max INT maxAACLine 32 / frameLength− lsb,0 , and where
( ( ) )
by:
start-up l , k ' and bsco' are set to zero. Where bsco = 0 unless a scalable core coder is used, for which
Temp x
bsco=⋅max INT maxAACLine 32 / frameLength− k ,0 , and where
( ( ) )
x

In ISO/IEC 14496-3:2001/Amd.1, subclause 4.6.18.5 SBR Tool overview, replace:
The output from the filtering is stored in the matrix:
X k, l +≤ t , 0 k <32, 0 ≤ l < numTimeSlots⋅ RATE
( )
Low HFGen
by:
The subband filtered low band is defined by X according to:
Low
W(kl,−≤ t ) , 0 k < k, t ≤ l < l+ t

HFGen x HFGen f HFGen

0 , kk ≤< 32, t ≤l  
x HFGen f HFGen
X()kl, =

Low
W'(kl,+−l t ) , 0 ≤ k < k', 0 ≤ l < t
f HFGen x HFGen


0 , kk' ≤< 32, 0 ≤l   xHFGen
where W' is the W matrix from the previous frame, and k' is the k value from the previous frame, and
x x
where l=⋅numTimeSlots RATE . If scalable SBR is used the following apply instead of the equation above:
f
W(kl,−≤ t ) , 0 k <32, t ≤ l < l+ t

 HFGen HFGen f HFGen
X kl, =
()

Low
W'(kl,+−l t ) , 0 ≤ k < 32, 0 ≤ l < t
 f HFGen HFGen


In ISO/IEC 14496-3:2001/Amd.1, subclause 4.6.18.5 SBR Tool overview, replace:

X kl,+≤t , 0 k < k' + bsco', 0 ≤ l < l
( )
Low HFAdj x Temp


X k,,l+≤t 0 k < k+ bsco, l ≤ l < numTimeSlots⋅ RATE
()
Low HFAdj x Temp


X kl,= "
()
Y kl,++t , k' bsco' ≤ k < k + M, 0 ≤ l < l
()
HFAdj x x Temp

Y kl,,+≤t +k bsco k < k +≤ M, l l < numTimeSlots⋅ RATE
()
HFAdj x x Temp

0 , max k++bsco,k M+  M≤ k <64, 0 ≤ l < numTimeSlots⋅ RATE
()

xx

by:
6 © ISO/IEC 2004 — All rights reserved

---------------------- Page: 6 ----------------------
ISO/IEC 14496-3:2001/Amd.1:2003/Cor.1:2004(E)

X kl,,+≤t 0 k < k' + bsco', 0 ≤ l < l
( )
Low HFAdj x Temp


Y' kl,++t l , k'+bsco' ≤ k < k' + M', 0 ≤ l < l
()
HFAdj f x x Temp

0 , maxkb'++sco',k' M'≤ k <64, 0 ≤l < l
()
 xx Temp
X kl,=
()

X kl,,+≤t 0 k < k + bsco, l ≤< ll
()
 Low HFAdj x Temp f

Y kl,+≤t , k+ bsco k < k + M, l ≤ l < l
()

HFAdj x x Temp f

0, maxkb++sco,k M≤ k <64, l ≤l < l
()
 xx Temp f

where
l=⋅numTimeSlots RATE
f
and


In ISO/IEC 14496-3:2001/Amd.1, subclause 4.6.18.6.2 Inverse filtering , replace:
where bs_invf_mode´ are the bs_invf_mode values from the previous SBR frame.
by:
where bs_invf_mode´ are the bs_invf_mode values from the previous SBR frame, and are assumed to
be zero for the first frame.

In ISO/IEC 14496-3:2001/Amd.1, subclause 4.6.18.6.3 HF generator, replace Figure 4.46 by the following
figure:
© ISO/IEC 2004 — All rights reserved 7

---------------------- Page: 7 ----------------------
ISO/IEC 14496-3:2001/Amd.1:2003/Cor.1:2004(E)
Start
 msb = k0
 usb = kx
 numPatches = 0
 goalSb = NINT( 2.048E6 / Fs )
 if (goalSb < kx + M)
   for (i = 0, k = 0 ; f [i] < goalSb; i++){
Master
      k = i + 1
   }
 else
   k = N
Master
j = k
 sb = f [ j ]
Master
 j = j - 1
 odd = (sb - 2 + k0)%(2)
sb <= ( k0 - 1 + msb - odd ) False
False
True
 patchNumSubbands[numPatches] = max( sb - usb, 0)
 patchStartSubband[numPatches] = k0 - odd - patchNumSubbands[numPatches]
patchNumSubbands[numPatches] > 0 False
True
 usb = sb
 msb = kx
 msb = sb
 numPatches = numPatches + 1
True f [ k ] - sb < 3
Master
 k = N
Master
False
sb == ( kx + M ) True
(patchNumSubbands[numPatches-1] < 3)
True
&& (numPatches > 1)
numPatches = numPatches -1
False
Done

8 © ISO/IEC 2004 — All rights reserved

---------------------- Page: 8 ----------------------
ISO/IEC 14496-3:2001/Amd.1:2003/Cor.1:2004(E)
In ISO/IEC 14496-3:2001/Amd.1, subclause 4.6.18.7.2 Mapping, replace:

+ff(ii1)+ ()
TableHigh TableHigh
0 if m≠ INT
 

2
 

S mk−=,l
()

IndexMapped x
+ii1+ 
ff() ()
 TableHigh TableHigh
S ii⋅=δ ,lifmINT
() ( )

Index Step


2


for ffim≤< i+1 , 0,≤ () ( )
TableHigh TableHigh High E
where

11if l≥=l OR S′ i
() ( () )

AIndex
δ il, = ,
()

Step
0 otherwise



by:

+ff(ii1)+ ()
TableHigh TableHigh
0 if m≠ INT
 

2
 

S mk−=,l
()

IndexMapped x
+ffii1+ 
() ()
 TableHigh TableHigh
S im⋅−δ k ,lifm=INT
() ( )

Index Step x


2


for ffim≤< i+1 , 0,≤ () ( )
TableHigh TableHigh High E
where

1,if l≥−l OR S′ m L'1=1
() ( )
 ( )
A IndexMapped E
δ ml, = ,
()

Step
0 otherwise



and replace:
′
and S (i) is S (i) of the previous SBR frame.
Index Index
by:
′
and S is S of the previous SBR frame for the same frequency range. If the frequency range is
IndexMapped IndexMapped
larger for the current frame, the entries for the QMF subbands not covered by the previous S are
IndexMapped
assumed to be zero.

In ISO/IEC 14496-3:2001/Amd.1, subclause 4.6.18.7.2 Mapping, replace:

uk=+Fri 1,l ,l
(()())
 i
S mk−=,,l δ il,l≤m ()()

Mapped x S i i
lk=Fri,,l l
()()()
 i

© ISO/IEC 2004 — All rights reserved 9

---------------------- Page: 9 ----------------------
ISO/IEC 14496-3:2001/Amd.1:2003/Cor.1:2004(E)
for 0,≤ ( )
High E

FFi,,HI≥+k i l, LO,F i1, HI≤F k i,l+1, LO,r l= LO
( ) ( ( ) ) ( ) ( ( ) ) ()



ki,,l==i r l HI
() ()


and where

1,1∈≤SFk,l: ki,llk,r () (()()) (()())
 { }
IndexMapped
.
δ()il, =

S
0, otherwise


In order to handle the varying frequency resolution of the envelope scalefactors, ki,l is introduced. For a
( )
given high frequency resolution band, ki,l gives the proper indices to the corresponding low frequency
( )
resolution band of which the former is a subset, if the current SBR envelope is of low frequency resolution.
Finally, the δ il, function returns one if any entry in the S matrix is one within the given boundaries,
( )
IndexMapped
S
i.e. if an additional sinusoid is present within the present frequency band.

by:

ui=+Fr1,l
( ())
 i
S mk−=,,l δ il,l≤m ()()

Mapped x S i i
li=Fr,()l
()
i


for 0,≤ ()
( )
E
where

1,1∈−SFjk,l: i,rl≤j { () (()) ( ())}
 IndexMapped x
δ il, = .
()

S
0, otherwise



The δ il, function returns one if any entry in the S matrix is one within the given boundaries, i.e. if
( )
IndexMapped
S
an additional sinusoid is present within the present frequency band. The S matrix is hence one for all
Mapped
QMF subbands in the scalefactor bands where an additional sinusoid shall be added.


In ISO/IEC 14496-3:2001/Amd.1, subclause 4.6.18.7.6 Calculation of Additional HF Signal Components,
replace the equation:
S ml,
( )
Mapped
SE()ml,,=⋅(ml) ,0≤m M OrigMapped E
1,+Q ml
()
Mapped
by:
10 © ISO/IEC 2004 — All rights reserved

---------------------- Page: 10 ----------------------
ISO/IEC 14496-3:2001/Amd.1:2003/Cor.1:2004(E)
S ml,
( )
IndexMapped
SEml,,=⋅ml ,0≤m () ( )
M OrigMapped E
1,+Q ()ml
Mapped


In ISO/IEC 14496-3:2001/Amd.1, subclause 4.6.18.7.5 Calculation of gain, replace the equation:

E ml,
()
OrigMapped

if S m,0l =
()
M
εδ+⋅EQml,1+(l)⋅ (m,l)
()()()
Curr Mapped

G ml,,=≤0 m ()

E
EQml,,ml
 () ()
OrigMapped Mapped
⋅≠if S m,0l
()
 M
ε+E ml,1,+Q ml
()() ()
()
Curr Mapped


by:

E ()ml,
OrigMapped
 if S m,0l =
()
Mapped
εδ+⋅EQml,1+(l)⋅ (m,l)
()()()
Curr Mapped

G ml,,=≤0 m ()

E
EQml,,ml
 () ()
OrigMapped Mapped
⋅≠if S m,0l
()
 Mapped
ε+E ml,1,+Q ml
()() ()
()
Curr Mapped



In ISO/IEC 14496-3:2001/Amd.1, subclause 4.6.18.7.5 Calculation of gain, replace the equation:
f ()k+−11
TableLim
ε + E il,
()
0 ∑ OrigMapped
ik=f ()
TableLim
G k,,l=⋅limGain bs_limiter_gains0≤ k< N,0≤ l< L
() ()
Max LE
Temp f k+−11
()
TableLim
ε + E il,
()
0 ∑ Curr
ik=f
()
TableLim
5
GGml,m=≤in k m,l,10,0 m () (() )
( )
Max Max E
Temp
where km is defined by ffkm≤ ( ) ( ( )) ( ( ) )
TableLim TableLim
by:
f kk+−11−
()
TableLim x
ε + E il,
()
∑
0 OrigMapped
ik=−f k
()
TableLim x
G k,,l=⋅limGain bs_limiter_gains0≤ k< N,0≤ l< L
() ()
Max LE
Temp f ()kk+−11−
TableLim x
ε + E il,
()
0 ∑ Curr
ik=−f ()k
TableLim x
5
GG()ml,m=≤in k(m),l,10,0 m ( )
( )
Max Max E
Temp
where km is defined by ffkm( )≤+m k< km( )+1 ,
( ) ( ) ( )
TableLim x TableLim
© ISO/IEC 2004 — All rights reserved 11

---------------------- Page: 11 ----------------------
ISO/IEC 14496-3:2001/Amd.1:2003/Cor.1:2004(E)
In ISO/IEC 14496-3:2001/Amd.1, subclause 4.6.18.7.5 Calculation of Gain, replace:
0,S il, ≠0ORl=l
( )
MA
for 0,≤ ()()

L E M
1, otherwise

by:
≠0,S i,l 0 OR l= l OR l= l
( )
MAAPrev
for 0,≤ ()()

L E M
1, otherwise


In ISO/IEC 14496-3:2001/Amd.1, subclause 4.6.18.7.5 Calculation of gain, replace the equation
f ()k+−11
TableLim
ε + E il,
()
0 ∑ OrigMapped
ik=f
()
TableLim
G kl, =
()
Boost
Temp f ()k+−11
TableLim
22 2
εδ+⋅EGil,,il+S il,+S il,,l⋅Q i,l
() () ()()() ()
()
0 ∑ Curr Lim M M M
ik=f ()
TableLim
ffkm≤ ( ( )) ( ( ) )
TableLim TableLim
by:
f kk+−11−
()
TableLim x
ε + E ()il,
0 ∑ OrigMapped
ik=−f ()k
TableLim x
G kl, =
()
Boost
f kk+−11−
Temp ()
TableLim x
22 2
εδ+⋅EGil,,i l+S il,+S il,,l⋅Q i,l
() ( ) ()()() ()
∑()
0 Curr Lim M M M
Lim
ik=−f k
()
TableLim x
ffkm≤+m k< km+1
( ) ( )
( ) ( )
TableLim x TableLim

In ISO/IEC 14496-3:2001/Amd.1, subclause 4.6.18.7.6 Assembling HF signals, replace the equation:
h
SL

Ghmi,0−+j h ⋅ j if l≠l ANDh≠
() ()
∑ Temp SL Smooth A SL
G mi, =
()
 j=0
Filt

G mi, +h otherwise
()
Temp SL

rate⋅≤ttl i< rate⋅ l+1,0≤ m< M,0≤ l< L
() ( )
E EE
by:
h
SL

Gh()m,0i−+j h ⋅( j) if l≠ l AND h≠ ANDl≠ l
∑ Temp SL Smooth A SL APrev

G mi, =
() j=0

Filt

G mi, +h otherwise
()
Temp SL

RATEl⋅≤tti () ( )
E EE
12 © ISO/IEC 2004 — All rights reserved

---------------------- Page: 12 ----------------------
ISO/IEC 14496-3:2001/Amd.1:2003/Cor.1:2004(E)
In ISO/IEC 14496-3:2001/Amd.1, subclause 4.6.18.7.6 Assembling HF signals, replace the equation:
WGmi,,=⋅miX m+lsb,i+t, RATE⋅t0≤i ( ) ( ) ( ) ( )
( )
1 Filt High HfAdj E E E
by:
WGm,,i=⋅m iX m+ k,i+ t, RATE⋅t0≤ i< RATE⋅t L,0≤ m< M
( ) ( ) ( ) ( ) ( )
1 Filt High x HfAdj E E E

In ISO/IEC 14496-3:2001/Amd.1, subclause 4.6.18.7.6 Assembling HF signals, replace the equation:
h
SL

Qhmi,,− j+⋅h j if l≠l AND l≠l ANDS m l=0
() () ()
∑ Temp SL Smooth A APrev MBoost
Q mi, =
()
 j=0
Filt

0 otherwise

by:
≠QS(mi,,) if l l AND l≠l AND (ml)=0ANDh=0
Temp A APrev MBoost SL

h
SL

QQmi,,=−mi j+h⋅h j if l≠l AND l≠l ANDS m,l=0ANDh≠0
() () () ()

Filt ∑ Temp SL Smooth A APrev MBoost SL
j=0


0 otherwise


In ISO/IEC 14496-3:2001/Amd.1, subclause 4.6.18.7.6 Assembling HF signals, replace the equations:
ψ ml,,i=⋅SSm,l m,l⋅φ f i
( ) ( ) ( ) ()
( )
Re IndexMapped MBoost Re,sin IndexSine
ml+sb
ψ()ml,,i=⋅SS(m,l) (m,l)⋅(−1)⋅φ f (i)
( )
Im IndexMapped MBoost Im,sin IndexSine
by:
ψ ml,,i=⋅S m,l φ f i
( ) ( ) ( ())
Re MBoost Re,sin IndexSine
mk+
x
ψ ml,,i=⋅S m,l−1⋅φ f i
() ()() ( ())
Im MBoost Im,sin IndexSine

In ISO/IEC 14496-3:2001/Amd.1, subclause 4.6.18.7.6 Assembling HF signals, replace:
and inde
...