how do i make svcd with 5.1 CD image?

[redbullssg]

is it at all possible. i would really like 5.1 surround svcd's that i could lay in my stand alone dvd player!! it supports 48khz!
and also, what is the best setting for the TooLame mode? Stereo, dual channel, or joint stereo?

[chainsaw135]

moved.

[ulfschack]

Nope, not possible. (or rather ...) While there is support for more than two channels in mpeg-2 audio layers standards there aren't any standalones so far capable of playing it.

AC3 is multichannel, but for the DVD standard.

The option of downmixing from ac3 to mp2 with a nice "sourround like" effect is described somewhere on this forum. Do a search.

Cheers

[spinspan]

What about ReMPEG2 v1.52 :)
I have heard that can convert an AC3 5.1 audio track to MPEG2 5.1
I tried once and i saw that can do this, but with problems.:(
The program crashes everytime and i think that can't convert the rear left channel.:mad:

Another way to convert the AC3 5.1 to MPEG2 5.1 (which is the standard multichannel format for SVCDs) is using some special utilities for LINUX :rolleyes:

But the last think i know is that ONLY the I-Author (SVCD authoring program) can multiplex the multichannel audio into the SVCD image. AND this program is very s..t!:angry:

Thanx
====================================================
Where you hear a lot of cherries, hold small basket!
(greek words)

[fistula]

It is possible to convert ac3 5.1 to mpeg 5.1, but goodluck finding a standalone player that will play it (at least in the USA)... i have never seen one myself

[spinspan]

Goodevening from the Olympic City of 2004 :)

Look :cool:
Here in europe the standard audio format for the DVD was (at the beginning) the MPEG2.
The most standalone DVD players have on their body both the Dolby Digital and the MPEG2 multichannel LOGOS !!! Neither me, i do not know if any DVD in our market (greece) support for sure the MPEG2. But all the theme magazines about DVD and home cinema do make tests for this compatibility, and they tell us that the most of the players support it. Even mine which is a china made (amoisonic).

Hey I thing that I have to do some BIIIIIIIIIIG TESTS :D
And I must involve in the game many of my friends :rolleyes:
The MPEG2 multichannel must be RISED:devil:
We all have to help for this...:sly:

[frank]

SCVD MPEG-2 multichannel audio is possible!

SVCD authoring:
1. Demux AC3 tracks with DVD2AVI.
2. BeeSweet 1.4: convert to AIF 5.1 multi-channel.
Plays 2ch compatible in Winamp.
3. Search the internet for the Philips tool pub_enc.exe from Philips website. Anywhere in the forum the link was released, I don't remember.
4. pub_enc.exe: Make a batch file and encode the aif to mpg audio 320 kbps. Rename to mp2.
5. Mux with bbMPEG (video max = 2400kbit/s)
6. Authoring and burning with TSCV 0.84 (includes VCDimager 0.12, cdrao)

Your player must be switched to raw output mode.
DVD player: Philips DVD, SEG Las Vegas (ger), VESEG (turky) do it.
Then digital link to your amplifier including the MPEG-2 mc audiodecoder.
There are Philips and Pioneer receivers with MPEG audio decoder.

I made about 10 SVCD with MPEG-2 mc, sounds very fine!
BTW, ReMPEG2 v1.52 can only convert 5.0 tracks, and sounds shitty!

And now, good luck!

[ulfschack]

Thanks, Frank

The few times I venture to be categorical like above there are always someone tempted to put me in my place. Works beautifully :)

cheers

[irulat]

Frank, How do you make the batch file for the Philip's program?

[frank]

My AIF2MP2.BAT
SET MPEGTABLES=D:\DVD\philips\TABLES
"D:\DVD\philips\pub_enc.exe" -ms -ns -L -r44.1 -e -p3 -b320 E:\movies\mcsound.aif
You must change the path and filename.

[Latexxx]

http://212.68.196.9/eclipse/mpegmc.htm <- a gui (includes besweet and philips encoder)

[htc10825]

all links to the files for downloading are all dead.
Great work, nice guy!

[hypercube]

I have made some test with Rempeg and it seems to be completly buggy:
the LFE channel is swaped with Rear Left channel.

pub_enc.exe works perfectly but allow only multichannel AIFF as input
file. So you have to do stuff like that:

AC3 => 6 PCM WAV FILES => 1 Multichannel AIFF => 1 MPEG Multichannel

if you don't want to lose your time you can use Hypercube Transcoder.
http://hypercube.is.dreaming.org
My freeware is like ReMpeg but it works ! :p

now about multiplexing:

there is no "special multiplexer" for MPEG Multichannel.
you can use any multiplexer you want.
I-Author Multiplexer works
TMPGEnc Multiplexing tools too.

note that many people think that I-Author multiplexer generate
MPEG stream that can be used only by I-Author. It's not true.
You can check "PS" in I-Author Multiplexer to generate regular
MPEG Program Stream. (which can be burn by nero for example)

[smuck]

Hello All,

Frank wrote;

>SCVD MPEG-2 multichannel audio is possible!
>SVCD authoring:
>1. Demux AC3 tracks with DVD2AVI.
>2. BeeSweet 1.4: convert to AIF 5.1 multi-channel.
>Plays 2ch compatible in Winamp.
>3. Search the internet for the Philips tool pub_enc.exe from Philips >website. Anywhere in the forum the link was released, I don't >remember.
>4. pub_enc.exe: Make a batch file and encode the aif to mpg audio 320 >kbps. Rename to mp2.

I tried this too, and it works. My DVD played the 5.1 sound in SVCD.
But now I'am backing up "The Insider". Quite a long movie (2:31).
I split the movie in three parts (CD's). In the third part the sound is suddenly gone.

I noticed that when I tried to listen to the file "Encoded_audio_1.aif", it could no played. The reported error was that it had a bad reported file length.
I used the next command to covert to the aif format;
d:\Program Files\DVD2SVCD\BeSweet\BeSweet.exe -core( -input P:\Ripped\The_Insider\Extracted_audio_1.ac3 -output P:\Ripped\THE_IN~2\Encoded_audio_1.aif -6chaiff -logfile P:\Ripped\THE_IN~2\Encoded_audio_1.log ) -azid( -L -3db -c normal -s surround -g max )

The actual file length is 4.194.292 bytes (4Gb limit ?)

When the .aif file is wrong, the obviously the .mp2 file is also wrong.

Is this a problem with BeSweet v1.5b1 ?

Can anyone tell me what went wrong?


Update 22/1/2003;
I tried to rip the same movie in stereo and now I do not have the problem above. But that is probably obvious, because the conversion to the .aif format is not done for stereo.

CU, Smuck

[htc10825]

Use HeadAC3he to cut the .ac3 file in 3 parts(according to the length of 3 CDs) at first. Then decode/encode them separately.

[DSPguru]

no need to cut the ac3. just use the partial encoding feature.

[hypercube]

I have noticed that my (any ?) DVD player does not claim if
the start of MPEG is trashed. So I am doing this terrible thing: :p

encode a regular mpeg for the entire movie (TMPGENC + Hypercube transcoder).
cut without respecting MPEG format.
burn with nero.

[smuck]

Originally posted by DSPguru
no need to cut the ac3. just use the partial encoding feature.

Hi DSPguru,

I noticed you are the author of BeSweet ?! Great!

What do you mean with "just use the partial encoding feature". I tried to find it in DVD2SVCD, but could not find it. Please tell me more.

CU, Smuck

[DSPguru]

this feature is availabe via BeSweetGUI or commandline.

[smuck]

Originally posted by DSPguru
this feature is availabe via BeSweetGUI or commandline.

Right, I'll give it a try. I was looking in the settings of DVD2SVCD for this feature.

I'll try to do it by hand now (or via BeSweetGUI).

CU, Smuck

[hypercube]

dear smuck, I thing you are on the wrong way.
you are following this way:

>SCVD MPEG-2 multichannel audio is possible!
>SVCD authoring:
>1. Demux AC3 tracks with DVD2AVI.
>2. BeeSweet 1.4: convert to AIF 5.1 multi-channel.
>Plays 2ch compatible in Winamp.
>3. Search the internet for the Philips tool pub_enc.exe from Philips >website. Anywhere in the forum the link was released, I don't >remember.
>4. pub_enc.exe: Make a batch file and encode the aif to mpg audio 320 >kbps. Rename to mp2.

this is a very long way ;)

like htc10825 say: Use HeadAC3he to cut the .ac3 file in 3 parts(according to the length of 3 CDs) at first.

now, DO NOT DECODE THEM.

use "hypercube transcoder" to convert directly (in batch mode) this
3 AC3 files in 3 multichannel MP2 files.

that's all. (and hypercube transcoder is faster than pub_enc.exe !)

[smuck]

Originally posted by hypercube
dear smuck, I thing you are on the wrong way.
you are following this way:

>SCVD MPEG-2 multichannel audio is possible!
>SVCD authoring:
>1. Demux AC3 tracks with DVD2AVI.
>2. BeeSweet 1.4: convert to AIF 5.1 multi-channel.
>Plays 2ch compatible in Winamp.
>3. Search the internet for the Philips tool pub_enc.exe from Philips >website. Anywhere in the forum the link was released, I don't >remember.
>4. pub_enc.exe: Make a batch file and encode the aif to mpg audio 320 >kbps. Rename to mp2.

this is a very long way ;)

like htc10825 say: Use HeadAC3he to cut the .ac3 file in 3 parts(according to the length of 3 CDs) at first.

now, DO NOT DECODE THEM.

use "hypercube transcoder" to convert directly (in batch mode) this
3 AC3 files in 3 multichannel MP2 files.

that's all. (and hypercube transcoder is faster than pub_enc.exe !)

Alright, I will try this. But now the ease of using DVD2SVCD is gone :( . Have to search for hypercube transcoder first).

Thanks for the reply.


Update 25 februari;
I know waht the actual problem was. I converted my FAT32 partition to NTFS (XP). This solved the 4Gb limit in my *.AIF file. So it had/has nothing to do with BeSweet (whatever version).

CU, Smuck

[DSPguru]

Originally posted by hypercube
hypercube transcoder is faster than pub_enc.exe !but what about its quality ?
on which version of musicin is it based ?

[hypercube]

hypercube transcoder is based on musicin 2.0
pub_enc is nothing else than musicin 2.5
rempeg is based on musicin 2.0

I don't know if there is another MPEG2 multichanel encoder around...

audio quality is good. There is some limitation due to
MPEG Multichannel maximum bitrate: 384kbps. So high frequency
are not very well reproduced in rear speakers. This problem
is the same for every encoders. (try to encode a splash
symbal in MP3 at 96 kbps, you will hear the same artifacts)

some users say hypercube transcoder is fast. So I tell you
that it is fast. :p

there is some advice about hypercube transcoder in vcdhelp forum.

smuck, you can download this freeware here:
http://hypercube.is.dreaming.org/html/download.html

snapshots are here:
http://hypercube.is.dreaming.org/html/hypercube_transcoder.html

Currently , I am working on 2.0, so it was a beta version.
there is 2 little bugs but all works very well.

[DSPguru]

Originally posted by hypercube
hypercube transcoder is based on musicin 2.0
pub_enc is nothing else than musicin 2.5i don't have musicin's changelog, and never performed quality tests, but since v2.0/v1.9 (dist10) & v2.5 has basicly the same functionalty, i would bet that there is quality improvement in v2.5 .

I don't know if there is another MPEG2 multichanel encoder around...since tooLame is also based on dits10, i believe there were tooLame (http://sourceforge.net/tracker/index.php?func=detail&aid=450874&group_id=19022&atid=369022) versions who could encode m2mc, but mike have stopped offering this feature after a while.

snapshots are here:very nice :).

[smuck]

Originally posted by hypercube
hypercube transcoder is based on musicin 2.0

smuck, you can download this freeware here:
http://hypercube.is.dreaming.org/html/download.html

snapshots are here:
http://hypercube.is.dreaming.org/html/hypercube_transcoder.html


Ok, I downloaded it. Will try this weekend. I am at work ;-0 at the moment.

[hypercube]

quality improvement in v2.5 are:

-p model psycho model 1 or 3
-R steps DRC gain in 0.2 dB steps (dflt 0)
-X Amatr 7.1-ch aug. matrix (dflt 0)
-K Atca set 7.1-ch aug. tca (dflt 0)
-Y AdynX 7.1-ch aug. DynX mode (dflt 0)

here musicin 2.5:

ISO MPEG Audio Subgroup Software Simulation Group
ISO 13818-3 MPEG-2 Audio Multichannel Encoder
Revision: 2.5 990316

pub_enc: unrecognised option h

usage: pub_enc queries for all arguments, or
pub_enc [-l lay][-m mode][-r sfrq][-b br][-B br][-d emp][-C config]
[-x matr][-k tca][-y dynX][-X aug_matr][-K aug_tca][-Y aug_dynX]
[-c][-o][-e][-p model][-w][-g][-v verb][-L][-a][-P][-D] inputPCM [outBS]
where
-l lay use layer <lay> coding (dflt 2)
-m mode channel mode : s/d/j/m (dflt s)
-n mode surround mode : s/d/j/m (dflt s)
-r sfrq input smpl rate in kHz (dflt 48.0)
-b br total bitrate in kbps (dflt 384)
-B br MPEG1 bitrate in kbps (dflt 384)
-d emp de-emphasis n/5/c (dflt n)
-c mark as copyright
-o mark as original
-e add error protection
-p model psycho model 1 or 3
-x matr matrix mode (dflt 0)
-k tca set channel-switching (dflt -2)
-y dynX set dyn_cross mode (dflt 0)
-X Amatr 7.1-ch aug. matrix (dflt 0)
-K Atca set 7.1-ch aug. tca (dflt 0)
-Y AdynX 7.1-ch aug. DynX mode (dflt 0)
-w set phantom coding (dflt off)
-L set LFE on (dflt off)
-C mode set configuration (dflt 320)
-g set MultiLingual file (dflt off) input ML file = inputPCM+".ml"
-P set mc prediction on (dflt off)
-D set dyn_cross_LR on (dflt off)
-n nr number of anc data Bytes (dflt 2)
-R steps DRC gain in 0.2 dB steps (dflt 0)
-a set ancillary data on (dflt off)
-v verb level of verbosity 0-3 (dflt 1)

inputPCM input PCM sound file (standard or AIFF)
outBS output bit stream of encoded audio (dflt inName+.mpg)

here musicin 2.0:

ISO MPEG Audio Subgroup Software Simulation Group (1996)
ISO 13818-3 MPEG-2 Audio Multichannel Encoder
$Revision: 2.0 $
Enter PCM input file name <required>: ^C
C:\>musicin -h
ISO MPEG Audio Subgroup Software Simulation Group (1996)
ISO 13818-3 MPEG-2 Audio Multichannel Encoder
$Revision: 2.0 $
musicin: unrec option h
usage: musicin queries for all arguments, or
musicin [-l lay][-m mode][-r sfrq][-b br][-B br][-d emp][-C config]
[-x matr][-k tca][-y dynX]
[-c][-o][-e][-w][-g][-v verb][-L][-a][-P][-D] inputPCM [outBS]
where
-l lay use layer <lay> coding (dflt 2)
-m mode channel mode : s/d/j/m (dflt r)
-n mode surround mode : s/d/j/m (dflt r)
-r sfrq input smpl rate in kHz (dflt 0.0)
-b br total bitrate in kbps (dflt 384)
-B br MPEG1 bitrate in kbps (dflt 384)
-d emp de-emphasis n/5/c (dflt n)
-c mark as copyright
-o mark as original
-e add error protection
-x matr matrix (dflt 0)
-k tca set channel-switching (dflt -2)
-y dynX set dyn_cross mode (dflt 0)
-w set phantom coding (dflt off)
-L set LFE on (dflt off)
-C mode set configuration (dflt 320)
-g set MultiLingual file (dflt off) input ML filename = inputPCM+".ml"

-P set mc prediction on (dflt off)
-D set dyn_cross_LR on (dflt off)
-n nr number of anc data Bytes (dflt 2)
-a set ancillary data on (dflt off)
-v verb level of verbosity 0-3 (dflt 1)
inputPCM input PCM sound file (standard or AIFF)
outBS output bit stream of encoded audio (dflt inName+.mpg)

>since tooLame is also based on dits10, i believe there were tooLame >versions who could encode m2mc, but mike have stopped offering this >feature after a while.
dist10 is very buggy. I worked very hard to debug this release.
(translate trashed C into clean C++, port from linux to win32,
correct some random crash, etc...). Perhaps mike didn't want to
work any more on this messy code. :scared:

>very nice
thanks ! I am not very cool with GUI, I prefer to swim in C++. :D
I hope this one is easy to use.

[DSPguru]

Originally posted by hypercube
dist10 is very buggy. I worked very hard to debug this release.
(translate trashed C into clean C++, port from linux to win32,
correct some random crash, etc...). Perhaps mike didn't want to
work any more on this messy code. :scared: wtf are you talking about ? mike did a hell of a job!
in fact, tooLame is one of the finset mp2 encoder around.
it just that he stopped offering multichannel encoding.

[hypercube]

hey, keep cool ! :(

wait a minute, don't tell me that your 'mike' is one from
the dist10 authors ???!!!

I suppose mike is the tooLame author, and I am sure
he did a hell of a job because dist10 is not its code
and it is very dirty.

I say dist10 code is dirty, not tooLame code is dirty !

>he stopped offering multichannel encoding
strange idea... you know why ?

[DSPguru]

Originally posted by hypercube
I suppose mike is the tooLame author, and I am sure
he did a hell of a job because dist10 is not its code
and it is very dirty.Mike Cheng. yes.
sorry, but your comments about him brought some arrogant smell into my nose.
when i talked about sound quality, i wasn't talking about features, i was talking about psychoacoustic tunings.
mike had no c/c++ problems with dist10 when he founded lame (and later, toolame), he had a problem with its quality!

[hypercube]

sorry for my poor english :(

>mike had no c/c++ problems with dist10
there is two main bugs in dist10:
- one in the table loader, wrong values are read from files. this bug
cause random effects in encoding result.
- another one in the multichannel encoding process causing random crash.

your guy seems to be very strong in DSP.
(http://mikecheng.d2.net.au/)

it should be good I contact him to improve hypercube transcoder
(one day...for instance I don't have the time to work on audio
quality)

[DSPguru]

imho, the best thing would be, if you had integrated the multichannel musicin v2.0 part inside toolame v0.2i.

[hypercube]

may be, one day...:D

[hypercube]

dear DSPGuru,

there is guy who tell me this:

------------------------
For Example If you have a movie DIVX in NTSC (23.9 Frames/second) and you
convert it to a SVCD but in PAL, the frame rate its 25/s, therefor the
movie will be shorter. For example a movie with 1 hour and 3 minutes will
be only of 1 hour in PAL. So the audio must be shorter to be synchronized

In the Audio Program BeSweet theres an option to do this NTSC -> Pal in the
OTA presets options
-------------------------------


could you explain to me what is this AUDIO convertion ? (NTSC -> Pal)

[DSPguru]

look for the source-code of WLA. (Wave Length Adjust).

[hypercube]

WLA author say:

So the main usage of this program is to adjust an audio sample when
the sample rate is a little bit off. Lets say that you grab a video
sequence of your favorite homevideo with your video grapper. The frame
rate was ok at 25.000 fps (PAL) but the audio was sampled at 44095
samples/sec instead of the intended rate of 44100 sample/sec.

nothing related to a NTSC / PAL audio convertion !

perhaps this is not a good name for this functonnality in besweet...
or I don't understand something... :(

NTSC and PAL SVCD have the same audio frequency: 44100 isn't it ?

[DSPguru]

hi hypercube,
please pay your attention to the FRC presets in BeSwewetGUI - it has the same syntax like in WLA.

[hypercube]

may be my question have been lost in all this threads... so
I clarify what I want...
I just want to know what does it mean "NTSC / PAL audio convertion"
It does not make sens to me.
I seen Besweetgui:

NTSC -> PAL (23.976 to 25.000)
NTSC -> PAL (29.970 to 25.000)
PAL -> NTSC (25.000 to 23.976)
PAL -> NTSC (25.000 to 29.970)
NTSC -> NTSC (23.976 to 29.970
NTSC -> NTSC (29.970 to 23.976)

I understand technically what theses ratio are. The main goal of this, is to expand/stretch the audio track, isn't it ?
BUT, In my opinion, there is no relashionship between video
framerate and audio frequency. So why talking about NTSC and PAL
here ?
such transformation is logicial with video track, but not with audio tracks ?

the need of expand/stretch the audio track occur only when
capturing video with bad audio card (not perfect 44100Khz)
but never in DVD ripping, isn't it ?

arrrrgl, my english is not good :scared: