ISO/IEC 14496-3:2001/Amd 2:2004/Cor 1:2005
(Corrigendum)Information technology — Coding of audio-visual objects — Part 3: Audio — Amendment 2: Parametric coding for high-quality audio — Technical Corrigendum 1
Information technology — Coding of audio-visual objects — Part 3: Audio — Amendment 2: Parametric coding for high-quality audio — Technical Corrigendum 1
Technologies de l'information — Codage des objets audiovisuels — Partie 3: Codage audio — Amendement 2: Codage paramétrique pour le codage audio de haute qualité — Rectificatif technique 1
General Information
Relations
Standards Content (Sample)
INTERNATIONAL STANDARD ISO/IEC 14496-3:2001/Amd.2:2004
TECHNICAL CORRIGENDUM 1
Published 2005-08-15
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 2: Parametric coding for high-quality audio
TECHNICAL CORRIGENDUM 1
Technologies de l'information — Codage des objets audiovisuels —
Partie 3: Codage audio
AMENDEMENT 2: Codage paramétrique pour le codage audio de haute qualité
RECTIFICATIF TECHNIQUE 1
Technical Corrigendum 1 to ISO/IEC 14496-3:2001/Amd.2:2004 was prepared by Joint Technical Committee
ISO/IEC JTC 1, Information technology, Subcommittee SC 29, Coding of audio, picture, multimedia and
hypermedia information.
ICS 35.040 Ref. No. ISO/IEC 14496-3:2001/Amd.2:2004/Cor.1:2005(E)
© ISO/IEC 2005 – All rights reserved
Published in Switzerland
---------------------- Page: 1 ----------------------
ISO/IEC 14496-3:2001/Amd.2:2004/Cor.1:2005(E)
In ISO/IEC 14496-3:2001/Amd.2:2004, subpart 1, update Table 1.1. Replacing SSC by SSC(Transient,
Sinusoid, Noise) and add a PS tool, as illustrated in the table below.
AOT/Tools SSC(Transient, Sinusoid, Noise) Parametric
stereo
SSC X X
In ISO/IEC 14496-3:2001/Amd.2:2004, subclause 8.5.1, replace Table 8.10 with (modified parts are marked in
gray):
Table 8.10 - Decoder level
decoder_ Level of max_nrof_sinusoids max_nrof_den #bits for s_nrof_ #bits for
level complexity continuations n_nrof_lsf
00 Reserved Na Na Na Na
01 Medium 60 24 6 4
10 Reserved Na Na Na Na
11 Reserved Na Na Na Na
In ISO/IEC 14496-3:2001/Amd.2:2004, subclause 8.5.2, replace (modified parts are marked in gray):
refresh_sinusoids_next_frame – One bit providing an additional frame look ahead for the ADPCM decoding
of sinusoidal parameters. If this bit is set to %1, the next frame is a refresh frame. In that case the bit
refresh_sinusoids shall be set to %1 in the next frame.
with:
refresh_sinusoids_next_frame – One bit providing an additional frame look ahead for the ADPCM decoding
of sinusoidal parameters. If this bit is set to %1, the next frame is a refresh frame. In that case the bit
refresh_sinusoids shall be set to %1 in the next frame. If this bit is set to %0, the next frame is not a refresh
frame.
In ISO/IEC 14496-3:2001/Amd.2:2004, subclause 8.5.2, replace (modified parts are marked in gray):
t_phi[sf][ch][i] – For a transient of the Meixner type in sub-frame sf of channel ch, these bits represent the
phase of the i-th sinusoid under the transient envelope. The decoded value is converted into a phase value in
radians in the range [-π, π> and is specified for the start of the transient.
tp [i]= 2⋅tp ⋅t_phi[sf][ch][i] ,
q e
π
where tp represents the absolute phase error (tp = ) and tp represents the dequantized absolute phase
e q
e
32
(in radians). The allowed range for t_phi is [-16, 15]; the representation level +16 is represented by –16
(because +π==-π).
with:
t_phi[sf][ch][i] – For a transient of the Meixner type in sub-frame sf of channel ch, these bits represent the
phase of the i-th sinusoid under the transient envelope. The decoded value is converted into a phase value in
radians in the range [-π, π> and is specified for the start of the transient.
tp [i]= 2⋅tp ⋅t_phi[sf][ch][i] ,
q e
π
where tp represents the absolute phase error (tp = ) and tp represents the dequantized absolute phase
e q
e
32
(in radians). The allowed range for t_phi is [-16, 15].
2 © ISO/IEC 2005 – All rights reserved
---------------------- Page: 2 ----------------------
ISO/IEC 14496-3:2001/Amd.2:2004/Cor.1:2005(E)
In ISO/IEC 14496-3:2001/Amd.2:2004, subclause 8.5.2, replace (modified parts are marked in gray):
If no ICC data is sent in the bit-stream, all ICC parameters are reset to 1 (i.e. index=0). The default
quantization grid for ICC is provided in Table 8.B.19.
with:
If no ICC data is sent in the bit-stream, all ICC parameters are reset to 1 (i.e. index=0). The default
quantization grid for ICC is provided in Table 8.23
In ISO/IEC 14496-3:2001/Amd.2:2004, subclause 8.5.2, replace (modified parts are marked in gray):
iid_dt[e] – This flag describes for envelope index n, whether the IID parameters are coded differentially over
time (iid_dt==%1) or over frequency (iid_dt==%0). In the case iid_mode is different from the previous envelope
(e-1), iid_dt[e] shall have the value 0% forcing frequency differential coding.
with:
iid_dt[e] – This flag describes for envelope index e, whether the IID parameters are coded differentially over
time (iid_dt==%1) or over frequency (iid_dt==%0). In the case iid_mode is different from the previous envelope
(e-1), iid_dt[e] shall have the value 0% forcing frequency differential coding.
In ISO/IEC 14496-3:2001/Amd.2:2004, subclause 8.6.2.4, replace (modified parts are marked in gray):
8.6.2.4 Synthesis of sinusoids for segments without a transient
with:
8.6.2.4 Synthesis of sinusoids for segments with a transient
In ISO/IEC 14496-3:2001/Amd.2:2004, subclause 8.6.4.3, replace (modified parts are marked in gray):
In order to compensate for the overall delay of the hybrid analysis filterbank, the first 10 sets (6 from delay and
4 from QMF filter) of hybrid subbands are flushed and therefore not taken into account for processing.
with:
In order to compensate for the overall delay of the hybrid analysis filterbank, the first 10 sets (6 from delay and
4 from QMF filter) of hybrid subbands are flushed and therefore not taken into account for processing. Note
that in Figure 8.24 this delay has already been accounted for.
In ISO/IEC 14496-3:2001/Amd.2:2004, subclause 8.6.4.4, Figure 8.24, replace caption (modified parts are
marked in gray):
SSC frame
384
Monaural sub
frames
66 6 6 6 66 6
region region region region region
0 1 2 4
3
Stereo Stereo
frame i frame i+1
Figure 8.24 - One SSC frame comprises two stereo frames of data. The solid line illustrates the
interpolation between stereo parameters for slots that have not been assigned stereo parameters to
© ISO/IEC 2005 – All rights reserved 3
Stereo
bands
---------------------- Page: 3 ----------------------
ISO/IEC 14496-3:2001/Amd.2:2004/Cor.1:2005(E)
with:
SSC frame
384
Monaural sub
frames
66 6 6 6 66 6
region region region region region
0 1 2 4
3
Stereo Stereo
frame i frame i+1
Figure 8.24 - One SSC frame comprises two stereo frames of data. The solid line illustrates the
interpolation between stereo parameters for slots that have not been assigned stereo parameters to.
Note that the delay introduced by the hybrid QMF analysis filterbank has been compensated for in this
figure.
In ISO/IEC 14496-3:2001/Amd.2:2004, subclause 8.6.4.5.1, remove:
F is the output sampling rate.
s
Replace equation:
53 ,F <32kHz,10or 20stereobands
⎧
s
⎪
73 ,F <32kHz,34stereobands
⎪
s
NR__ALLPASS BANDS=
⎨
30 ,F ≥32kHz,10or 20stereobands
s
⎪
⎪
50 ,F ≥32kHz,34stereobands
⎩ s
with:
30 ,10or 20stereobands
⎧
NR__ALLPASS BANDS=
⎨
50 ,34stereobands
⎩
Replace equation:
71 ,F <32kHz,10or 20stereobands
⎧
s
⎪
91 ,F <32kHz,34stereobands
⎪
s
SHORT__DELAY BAND=
⎨
,F ≥32kHz,10or 20stereobands
42
s
⎪
⎪
62 ,F ≥32kHz,34stereobands
⎩ s
with:
42 ,10or 20stereobands
⎧
SHORT__DELAY BAND=
⎨
62 ,34stereobands
⎩
Replace equation:
0.6 ,F <32kHz
⎧
s
a =
⎨
Smooth
0.25 ,F ≥32kHz
⎩ s
with:
a = 0.25
Smooth
4 © ISO/IEC 2005 – All rights reserved
Stereo
bands
---------------------- Page: 4 ----------------------
ISO/IEC 14496-3:2001/Amd.2:2004/Cor.1:2005(E)
In ISO/IEC 14496-3:2001/Amd.2:2004, subclause 8.6.4.5.3, replace equation:
⎧
P(in, )
,,γ⋅>PPin i,n
() ( )
⎪
SmoothPeakDecayDiffNrg
G in, = γ⋅P in,
() ()
⎨
TransientRatio SmoothPeakDecayDiffNrg
⎪
1,otherwise
⎩
with:
⎧ P in,
( )
SmoothNrg
,,γ⋅>PPin in,
() ()
⎪
SmoothPeakDecayDiffNrg SmoothNrg
GPin,,=⋅γ in
() ()
⎨
TransientRatio SmoothPeakDecayDiffNrg
⎪
1,otherwise
⎩
In ISO/IEC 14496-3:2001/Amd.2:2004, subclause 8.6.4.5.4, replace Table 8.34 with:
Table 8.34 - Filter coefficient vector and delay length vector d m .
()
m
a m d m
() ( )
0 0.65143905753106 3
1 0.56471812200776 4
2 0.48954165955695 5
Remove equation:
d ,3F <2kHz
⎧
24kHz s
Delay length vector, d= .
⎨
d ,3F ≥2kHz
48kHz s
⎩
In ISO/IEC 14496-3:2001/Amd.2:2004, subclause 8.6.4.5.4, replace Table 8.35 with (modified parts are
marked in gray):
Table 8.35 - Delay length vector f .
center _20
k k
f k f k
() ()
center_20 center_20
0 -3/8 5 7/8
1 -1/8 6 5/4
2 1/8 7 7/4
3 3/8 8 9/4
4 5/8 9 11/4
Replace Table 8.38 with:
Table 8.36 - Peak Decay Factor α .
α
0.76592833836465
Remove equation:
α ,3F <2kHz
⎧
Decay24kHz s
Peak decay factor, α= .
⎨
α ,3F ≥2kHz
Decay48kHz s
⎩
© ISO/IEC 2005 – All rights reserved 5
---------------------- Page: 5 ----------------------
ISO/IEC 14496-3:2001/Amd.2:2004/Cor.1:2005(E)
In ISO/IEC 14496-3:2001/Amd.2:2004, subclause 8.6.4.6.1, replace (modified parts are marked in gray):
The number of stereo bands that is actually used for the processing of the cues depends on the number of
available parameters for IID and ICC according to the relation given in Table 8.39.
with:
The number of stereo bands that is actually used for the processing of the cues depends on the number of
available parameters for IID and ICC according to the relation given in Table 8.39. In case no IID or no ICC
parameters have been transmitted in the current frame (enable_iid==%0 or enable_icc==%0), the number of
IID or ICC parameters, respectively, is assumed to be 20 for the purpose of Table 8.39. In case no IID and no
ICC parameters have been transmitted in the current frame (enable_iid==%0 and enable_icc==%0), the
number of stereo bands in the previous frame is kept unchanged and used also for the processing of the
current frame.
In ISO/IEC 14496-3:2001/Amd.2:2004, subclause 8.6.4.6.1, replace (modified parts are marked in gray):
If the number of stereo bands changes from 10,20 in the previous frame to 34 in the current frame, the stereo
parameters from the previous frame are mapped to 34 stereo bands
...
Questions, Comments and Discussion
Ask us and Technical Secretary will try to provide an answer. You can facilitate discussion about the standard in here.