readme_usage

MPEG2/MPEG4编解码参考程序(实现了MPEG4的部分功能)

The following is usages of the MPEG-4 CELP reference software.

Encoder Usage:
--------------------------------------------------------
LPC-based encoder core V5.0 13-nov-97
Usage:
     mp4auenc -m lpc -r <rate> -c "<options>" <speech file>
      where <rate> is the desired bit rate
            <speech file> is a 8kHz/16kHz speech file
            <options> are:
                e <0/1> : Multi-Pulse Excitation (0) or Regular-Pulse Excitation (1)
                n <NumEnhStage> : the number of enhancement layers 
                f <0/1> : Fine Rate Control OFF (0) or ON (1)
                b <NBbitrate> : the bitrate for NarrowBand speech in the BandWidthScalable coder
                p <0/1> : Pre-processing OFF (0) or ON (1)
                d <0/1> : Debug Level OFF (0) or ON (1)

--------------------------------------------------------

Decoder Usage:
--------------------------------------------------------
CELP-based decoder core V5.0 13-nov-97
Usage: mp4audec -c "<options>" <name bitstream file>
       where <options> are:
             n <DecEnhStage> : Decoded number of enhancement layers (0, 1, 2, 3)
             b <0/1>         : Decoding NarrowBand speech (0) or WideBand speech (1)
             p <0/1>         : Post filter OFF (0) or ON (1). Default: OFF
             d <0/1>         : Debug Level OFF (0) or ON (1). Default: OFF
--------------------------------------------------------

The following is examples of command line options for the
MPEG-4 CELP (MPE mode) encoder/decoder.

1. Narrowband/Wideband without scalability mode 

Encode:
mp4auenc -m lpc -r <rate> -c "e=0,n=0,f=0,b=0,p=1" input.wav -o bitstream.mp4

input.wav:         input speech file [8/16 kHz] (wav format)
bitstream.mp4:     encoded bitstream file (mp4 format)
e=0:               multi-pulse excitation mode
n=0:               number of enhancement layers for bitrate scalable mode.
                   "0" indicates encoding without bitrate scalability.
f=0:               FineRateControl is off.
b=0:               bitrate for narrowband core CELP in the bandwidth 
                   scalable mode.
                   "0" indicates encoding without bandwidth scalability.
p=1:               pre-processing is on.
rate:              the following bitrates are supported.
                   Narrowband mode:
                    3850,  4250,  4650,  4900,  5200,  5500,  6200,
                    5700,  6000,  6300,  6900,  7100,  7300,  7700,
                    8300,  8700,  9100,  9500,  9900, 10300, 10500,
                   10700, 11000, 11400, 11800, 12000, 12200
                   Wideband mode:
                   10900, 11500, 12100, 12700, 13300, 13900, 14300,
                   14700, 15900, 17100, 17900, 18700, 19500, 20300,
                   21100, 13600, 14200, 14800, 15400, 16000, 16600,
                   17000, 17400, 18600, 19800, 20600, 21400, 22200,
                   23000, 23800

Decode:
mp4audec -c "p=1" bitstream.mp4 -o decout.wav

bitstream.mp4:     encoded bitstream file (mp4 format)
decout.wav:        decoded speech file [8/16 kHz] (wav format)
p=1:               post-filter is on.

2. Narrowband/Wideband bitrate scalable mode 

Encode:
mp4auenc -m lpc -r <rate> -c "e=0,n=num,f=0,b=0,p=1" input.wav -o bitstream.mp4

input.wav:         input speech file [8/16 kHz] (wav format)
bitstream.mp4:     encoded bitstream file (mp4 format)
e=0:               multi-pulse excitation mode
n=num:             number of enhancement layers for bitrate scalable mode.
                   The available values are "1", "2" and "3".
f=0:               FineRateControl is off.
b=0:               bitrate for narrowband core CELP in the bandwidth 
                   scalable mode.
                   "0" indicates encoding without bandwidth scalability.
p=1:               pre-processing is on.
rate:              Total bitrate of core layer and enhancement layers.

                   Narrowband mode: rate = core_rate + num * 2000
                   Wideband mode:   rate = core_rate + num * 4000

                   where core_rate indicates bitrate of core layer
                   (see example 1).

Decode:
mp4audec -c "n=dnum,p=1" bitstream.mp4 -o decout.wav

bitstream.mp4:     encoded bitstream file (mp4 format)
decout.wav:        decoded speech file [8/16 kHz] (wav format)
p=1:               post-filter is on.
n=dnum:            number of enhancement layers to be decoded.
                   The available values are equal to or less than 
                   "num" for encoding.

3. Bandwidth scalable mode 

Encode:
mp4auenc -m lpc -r <rate> -c "e=0,n=0,f=0,b=nbrate,p=1" input.wav -o bitstream.mp4

input.wav:         input speech file [16 kHz] (wav format)
bitstream.mp4:     encoded bitstream file (mp4 format)
e=0:               multi-pulse excitation mode
n=0:               number of enhancement layers for bitrate scalable mode.
                   "0" indicates encoding without bitrate scalability.
f=0:               FineRateControl is off.
b=nbrate:          bitrate for narrowband core CELP in the bandwidth 
                   scalable mode.
p=1:               pre-processing is on.
rate:              Total bitrate of core layer and enhancement layer.

                   rate = nbrate + bwsrate

                   where nbrate indicates bitrate of core layer
                   (see example 1).
                   The available bitrates for enchancement layer can 
                   be selected from 4 descrete spteps for each core layer
                   bitrate as follows:
                    3850 <= nbrate <=  4650:  9200, 10400, 11600, 12400
                    4900 <= nbrate <=  5500:  9467, 10667, 11867, 12677
                    5700 <= nbrate <= 10700: 10000, 11200, 12400, 13200
                   11000 <= nbrate <= 12200: 11600, 12800, 14000, 14800

Decode for Naorrowband:
mp4audec -c "b=0,p=1" bitstream.mp4 -o decout.wav

bitstream.mp4:     encoded bitstream file (mp4 format)
decout.wav:        decoded speech file [8 kHz] (wav format)
p=1:               post-filter is on.
b=0:               "0" indicates decoding narrowband speech.

Decode for Wideband:
mp4audec -s 16000 -c "b=1,p=1" bitstream.mp4 -o decout.wav

bitstream.mp4:     encoded bitstream file (mp4 format)
decout.wav:        decoded speech file [16 kHz] (wav format)
p=1:               post-filter is on.
b=1:               "1" indicates decoding wideband speech.



**********************************************************************
**********************************************************************

3   TwinVQ
==========


gmake makedirs
gmake TF=1 VERSION=1 ERRORCONCEALMENT=0 STATISTICS=0 all

switch           description                              default

-core_br         bitrtae for core [bit/s]                  8000.0
	         priority is higher than -r option
-tvq_wshape      enable KB window                          sine window
-tvq_tns         enable tns                                tns off
-tvq_960         set frame length 960                      1024
-tvq_bandlimit   enable bandlimit                          bandlimit off
-tvq_ppc         enable ppc(periodic peak component)       ppc off
-tvq_postproc    enable postprocess                        postproc off
-tvq_msmask1     enable ms_mask mode 1                     ms_mask=0 or 2


#TV00
setenv MP4_RAWAUDIOFILE "integer16, 0 , 8000.0, native, 1, 1.0"
<platform>_v1/mp4auenc -d 1 -m tf \
 -c "-mp4ff -core_coder 10 -mode 0  -core_br 8000" -r 9400 ABBA.08
<platform>_v1/mp4audec -d 1 -c "-mp4ff "  ABBA -ed raw -o ABBA8.8raw

#TV01
setenv MP4_RAWAUDIOFILE "integer16, 0 , 16000.0, native, 1, 1.0"
<platform>_v1/mp4auenc -d 1 -m tf \
-c "-mp4ff -core_coder 10 -mode 0  -tvq_wshape -core_br 16000"\
-r 18400 ABBA.16
<platform>_v1/mp4audec -d 1 -c "-mp4ff "  ABBA -ed raw -o ABBA16.16raw

#TV02
setenv MP4_RAWAUDIOFILE "integer16, 0 , 16000.0, native, 1, 1.0"
<platform>_v1/mp4auenc -d 1 -m tf \
-c "-mp4ff -core_coder 10 -mode 0  -pred_ltp -core_br 16000" -r 18400 ABBA.16
<platform>_v1/mp4audec -d 1 -c "-mp4ff "  ABBA -ed raw -o ABBA16.16raw

#TV03
setenv MP4_RAWAUDIOFILE "integer16, 0 , 16000.0, native, 1, 1.0"
<platform>_v1/mp4auenc -d 1 -m tf \
-c "-mp4ff -core_coder 10 -mode 0  -tvq_tns -core_br 16000" -r 18400 ABBA.16
<platform>_v1/mp4audec -d 1 -c "-mp4ff  "  ABBA -ed raw -o ABBA16.16raw

#TV04
setenv MP4_RAWAUDIOFILE "integer16, 0 , 16000.0, native, 1, 1.0"
<platform>_v1/mp4auenc -d 1 -m tf \
-c "-mp4ff -core_coder 10 -mode 0  -tvq_960 -core_br 16000" -r 18400 ABBA.16
<platform>_v1/mp4audec -d 1 -c "-mp4ff  "  ABBA -ed raw -o ABBA16.16raw

#TV05
setenv MP4_RAWAUDIOFILE "integer16, 0 , 16000.0, native, 1, 1.0"
<platform>_v1/mp4auenc -d 1 -m tf \
-c "-mp4ff -core_coder 10 -mode 0  -tvq_bandlimit -core_br 16000" \
-r 18400 ABBA.16
<platform>_v1/mp4audec -d 1 -c "-mp4ff "  ABBA -ed raw -o ABBA16.16raw

#TV06
setenv MP4_RAWAUDIOFILE "integer16, 0 , 16000.0, native, 1, 1.0"
<platform>_v1/mp4auenc -d 1 -m tf \
-c "-mp4ff -core_coder 10 -mode 0  -tvq_ppc -core_br 16000" -r 18400 ABBA.16
<platform>_v1/mp4audec -d 1 -c "-mp4ff "  ABBA -ed raw -o ABBA16.16raw

#TV07
setenv MP4_RAWAUDIOFILE "integer16, 0 , 16000.0, native, 1, 1.0"
<platform>_v1/mp4auenc -d 1 -m tf \
-c "-mp4ff -core_coder 10 -mode 0  -tvq_postproc -core_br 16000" \
-r 18400 ABBA.16
<platform>_v1/mp4audec -d 1 -c "-mp4ff "  ABBA -ed raw -o ABBA16.16raw

#TV08
-------no bitstream

#TV09
-------no bitstream

#TV10
setenv MP4_RAWAUDIOFILE "integer16, 0 , 16000.0, native, 2, 1.0"
<platform>_v1/mp4auenc -d 1 -m tf \
-c "-mp4ff -core_coder 10 -mode 1  -tvq_msmask1 -core_br 16000 -nttScl 16000" \
-r 38400 ABBA.16
<platform>_v1/mp4audec -d 1 -c "-mp4ff -out 0"  ABBA -ed raw -o ABBA8.16raw

#TV11
setenv MP4_RAWAUDIOFILE "integer16, 0 , 16000.0, native, 2, 1.0"
<platform>_v1/mp4auenc -d 1 -m tf \
-c "-mp4ff -core_coder 10 -mode 1  -tvq_msmask1 -core_br 16000 -nttScl 16000"\
-r 38400 ABBA.16
<platform>_v1/mp4audec -d 1 -c "-mp4ff -out 1"  ABBA -ed raw -o ABBA88.16raw

#TV12
setenv MP4_RAWAUDIOFILE "integer16, 0 , 24000.0, native, 1, 1.0"
<platform>_v1/mp4auenc -d 1 -m tf \
-c "-mp4ff -core_coder 10 -mode 0  -core_br 8000 -nttScl 8000 8000"\
-r 38400 ABBA.16
<platform>_v1/mp4audec -d 1 -c "-mp4ff -out 0"  ABBA -ed raw -o ABBA8.24raw

#TV13
setenv MP4_RAWAUDIOFILE "integer16, 0 , 24000.0, native, 1, 1.0"
<platform>_v1/mp4auenc -d 1 -m tf \
-c "-mp4ff -core_coder 10 -mode 0  -core_br 8000 -nttScl 8000 8000"\
-r 38400 ABBA.16
<platform>_v1/mp4audec -d 1 -c "-mp4ff -out 1"  ABBA -ed raw -o ABBA88.24raw

#TV14
setenv MP4_RAWAUDIOFILE "integer16, 0 , 24000.0, native, 1, 1.0"
<platform>_v1/mp4auenc -d 1 -m tf \
-c "-mp4ff -core_coder 10 -mode 0  -core_br 8000 -nttScl 8000 8000"\
-r 38400 ABBA.24
<platform>_v1/mp4audec -d 1 -c "-mp4ff -out 2"  ABBA -ed raw -o ABBA888.24raw

#TV15
setenv MP4_RAWAUDIOFILE "integer16, 0 , 48000.0, native, 2, 1.0"
<platform>_v1/mp4auenc -d 1 -m tf \
-c "-mp4ff -core_coder 10 -mode 1  -core_br 32000 -nttScl 32000 32000" \
-r 120000 ABBA.48
<platform>_v1/mp4audec -d 1 -c "-mp4ff -out 2" ABBA -ed raw -o ABBA161616.48raw

#TV16
setenv MP4_RAWAUDIOFILE "integer16, 0 , 44100.0, native, 1, 1.0"
<platform>_v1/mp4auenc -d 1 -m tf \
-c "-mp4ff -core_coder 10 -mode 0  -core_br 12000 -nttScl 12000 12000 12000 12000 12000 12000 12000" \
-r 120000 ABBA.44
<platform>_v1/mp4audec -d 1 -c "-mp4ff -out 7"  ABBA -ed raw -o ABBA12_8.44raw



**********************************************************************
**********************************************************************

4   AAC scalable
================


Compilation:
------------

gmake makedirs
gmake TF=1 LPC=1 VERSION=1 all


Operation modes:
----------------

-aac_raw         plain MPEG-2 AAC encoding/decoding mode
-aac_sca         old AAC scalable mode (now obsolete)
-aac_sys         AAC scalable mode with systems bitstream syntax


Switches:
---------

-aac_tns                   enable Temporal Noise Shaping
-aac_pns <start band>      enable Perceptual Noise substitution
-pred_ltp                  enable Nokia Long Term Predictor


Options for the scalable mode:
------------------------------

-mode <n>                  selects the number of layers and the channel configuration
-core_coder <n>            selects the core coder type (TVQ or CELP)


Examples:
---------

Plain AAC:
----------
mp4auenc -m tf -c "-aac_raw -aac_tns" -r <bitrate>  <input file> -o <bitstream file>
mp4audec -c "-aac_raw" -aaceof  <bitstream file> -o <decoded file> -s <sampling freq> -n <num channels> 


Scalable AAC:
-------------

1 TVQ layer only (mode 0):
-----------------------------
mp4auenc -m tf -c "-aac_sys -mode 0 -core_br <bitrate>"  <input file> -o <bitstream file>
mp4audec -c "-mp4ff -out <x>" -aaceof <bitstream file> -o <decoded file>


1 AAC layer only (mode 2):
--------------------------
mp4auenc -m tf -c "-aac_sys -mode 2" -r <bitrate>  <input file> -o <bitstream file>
mp4audec -c "-mp4ff -out <x>" -aaceof <bitstream file> -o <decoded file>


1 TVQ + 1 AAC layer (mode 4):
-----------------------------
mp4auenc -m tf -c "-aac_sys -mode 4 -core_br <core bitrate>" -r <bitrate>  <input file> -o <bitstream file>
mp4audec -c "-mp4ff -out <x>" -aaceof <bitstream file> -o <decoded file>


1 TVQ + 2 AAC layer (mode 7):
-----------------------------
mp4auenc -m tf -c "-aac_sys -mode 7 -core_br <core bitrate>" -r <bitrate>  <input file> -o <bitstream file>
mp4audec -c "-mp4ff -out <x>" -aaceof <bitstream file> -o <decoded file>


2 AAC layers (mode 11):
-----------------------
mp4auenc -m tf -c "-aac_sys -mode 11" -r <bitrate>  <input file> -o <bitstream file>
mp4audec -c "-mp4ff -out <x>" -aaceof <bitstream file> -o <decoded file>


for further informations, please contact:

[e-mail]
[e-mail]



**********************************************************************
**********************************************************************

--
$Id: README_usage,v 1.9 1999/08/09 16:23:43 purnhage Exp $