* BeSweet does not work as expected

[LigH]

Dear friends,

I wanted to use BeSweet to automate the conversion between AC3 and MP3. Because the BeSweet GUI is using so much resources (like any VB application) and sometimes won't start when other programs are already running, I wanted to write a batch file to call it with the required parameters. But I probably misunderstood something...

Maybe the command line is too long? But I even changed the order - same result. At least, no MP3 file is created, because the LAME part is not executed, instead a WAV file is written. Executing from the command line didn't help as well. I'm not sure if I can move everything to a list file, but this would probably not be what I want, because that would require the input file to be included there?

I'll post the source here, maybe someone can find out what's wrong. I checked it with all BeSweet versions I have (0.95, 0.96, 1.0). Also, in 1.0 the boost limit factor is reported to be 0.0!


@ECHO OFF
IF "%1"" == "" GOTO NoParams
IF "%2"" == "" GOTO NoParams
IF "%3"" == "" GOTO NoParams
ECHO ON
H:\Programme\BeSweet\BeSweet -core( -input "%1" -output "%2" -2ch ) -azid( -f1 -n1 ) -ssrc( --rate 44100 ) -boost( /l=0.95 /b=5 ) -lame( -m j --abr %3 -b 80 -B 320 )
@ECHO OFF
GOTO End
:NoParams
ECHO Syntax: %0 input.ac3 output.mp3 bitrate
:End

[LigH]

You also might want to have a look at the output:


J:\VIDEO_TS\old>H:\Programme\BeSweet\BeSweet -core( -input "AC3TEST.AC3" -output "AC3TEST.MP3" -2ch ) -azid( -f1 -n1 ) -lame( -m j --abr 160 -b 80 -B 320 ) -ssrc( --rate 44100 ) -boost( /l=0.95 /b=5 )
BeSweet v1.0 by DSPguru.
--------------------------
Using azid.dll v1.7.1 (b698) by Midas (midas@egon.gyaloglo.hu).
Using ssrc.dll v1.28 by Naoki Shibata (shibatch@geocities.co.jp).
Manual Dynamic-Compression algorithm by LigH (author of WaveBooster).

[00:00:00:000] +------- BeSweet -----
[00:00:00:000] | Input : AC3TEST.ac3
[00:00:00:000] | Output: AC3TEST.MP3
[00:00:00:000] | Floating-Point Process: No
[00:00:00:000] +-------- AZID -------
[00:00:00:000] | Output Stereo mode: Dolby surround compatible
[00:00:00:000] | Total Gain: 0.0dB, Compression: None
[00:00:00:000] | LFE levels: To LR 0.0dB, To LFE 0.0dB
[00:00:00:000] | Center mix level: BSI
[00:00:00:000] | Surround mix level: BSI
[00:00:00:000] | Dialog normalization: Yes
[00:00:00:000] | Rear channels filtering: Yes
[00:00:00:000] +-------- SSRC -------
[00:00:00:000] | Source Sample-Rate: 48.0KHz
[00:00:00:000] | Dest. Sample-Rate: 44.1KHz
[00:00:00:000] | Attenuation : 0.0db
[00:00:00:000] +-------- BOOST ------
[00:00:00:000] | Boost Factor : 5.0
[00:00:00:000] | Limit Factor : 0.00
[00:00:00:000] +---------------------
[00:00:11:616] transcoding ...

[DSPguru]

hi LigH,
-core( -2ch ) is a switch that tells BeSweet to create a stereo wave.
-i found that the report of LimitFactor value was wrong. but the bug was only in the report, and it didn't affect the boosting process.

[LigH]

Indeed, you are right: without -2ch it works fine; I must have missed that it implies always wave output. Thanks!

[DSPguru]

i will fix the LimitFactor report on v1.1b3 .

btw,
you might want to consider adding -g max option to your batch file..

[LigH]

That probably won't change much with such a high amplification - but I'll need to check it with all my DVDs. All I need is to add a few report lines in my WaveBooster and compare the calculated values for the 2-pass mode.

The main reason why I'm not using '-g max' is the strange fact that it often froze my PC. I'm not sure which reason there is, maybe it's my CPU, my mainboard, my memory... Without that, 1-pass mode runs fine.

[Amerzone]

@ DSPguru

You know, I was always wondering why you don't include in the zip archive a text file describing the -core switches of BeSweet, instead of having the user to launch BeSweet from the command line to get info about them?

[user]

Hi,

is it technically possible (for you) to implement the use of the exe cutables of eg. ssrc and toolame instead of dll's , even in fullautomatic BeSweet mode ?

So that as much switches as possible could be used in automatic besweet mode ?

[sQuid_66]

Any idea why I get a CRC error during transcoding?

It works fine for a while, or fine on smaller files.

Other times it just gives a CRC error over and over and the output MP3 skips like crazy.

[DSPguru]

@Amerzone
was lazy i guess.. :( will do it in the future.

@user
that's basicly means we drop the no-intermediate-file idea.
if you like, you can use BeSweet to create 48khz/16bit/stereo wav, and then feed it to ssrc and then toolame.

@sQuid_66
probably something went wrong with your rip process, and the ac3 stream is corrupted.

[user]

"So that as much switches as possible could be used in automatic besweet mode ?"

Hi DSPGuru,

because of that reason I wrote "as much switches as possible".

I know that in ssrc the switch ssrc/2pass is not possible in automatic mode, 'cause of non intermediate file.

Would it be possible to implement in besweet auto mode the toolame.exe keeping nonintermediate files ?

[DSPguru]

BeSweet does offer two-pass without intermediate-file.
about toolame - no can do.

[user]

Offers BS now 2pass ssrc in dll in automatic mode ?
Some weeks ago I thought not. Should I update ?!

[DSPguru]

BeSweet's two-pass isn't based on SSRC's two-pass.
btw, i'm now wokring on a smarter two-pass engine (end-to-end calculation & find of total gain for a List), this mode would probably be slower than the one offered today, but will give higher Gain to the track without loosing quality :D.

[user]

Hi,

Some time ago I tried besweet on ac3 with 2.0, not the usual 5.1.

If you use azid in dos-box then you will know its behaviour in sorting max. gain values.

With 5.1 ac3 you will have normally really the highest value on top, then no problems by 20db - highest detected gain value and so on.

But with 2.0 ac3 there are often max gains of eg. 14 db. Would be no problem but sorting:
As "highest" gain value there is written 9.9 decreasing in list 9.8, 9.7 and so on down to 2.0

Then in the middle of the sort file:
really highest gain:
eg. 14.1 db
then sorted decreasing again , down to 0.0 db

If you use azid.exe manually , then you know the problem and you will look somewhere in the middle of the sort txt file to find really highest gain value.

But I remember that Besweet did not "know" it and took 9.9 db ( first value in txt file and normally the really highest...) .
That resulted of cause in bad clipping, because really highest was eg 14 db, so that besweet amplified too much (because besweet/azid thought that highest would be only 9.9 db...)

[DSPguru]

hi user :)
10q 4 the info.

i believe this won't happen with the new normalization process.

[sQuid_66]

Originally posted by DSPguru

@sQuid_66
probably something went wrong with your rip process, and the ac3 stream is corrupted.


I re-tried everything.... several different configs, rebooting, then retrying, etc.

On one of the DVD tracks it died at the same point, everytime, no matter what... about 9 mins into it.

:confused:

[coreano]

HI THERE, CAN ANYBODY HELP ME?

I CAN´T DOWNLOAD THE BESWEET GUI FROM ANY OF THE LINKS ABOVE,
CAN ANYBODY EMAIL THAT FILE PLEASE. I WILL APRECIATE.
clopez.arancia@zetafwd.com

THANKS

[mancer]

No need to scream there ;)