BeHappy - AviSynth based audio transcoding tool (UPD 19-07-2006)

[tebasuna51]

Originally Posted by idbirch2

What version of Aften are you using?

edit: I forgot, here's the output from this test:

Channels=6, BitsPerSample=16 int, SampleRate=48000Hz
Aften.exe -v 0 -b 384 -m 1 -readtoeof 1 -cmix 0 -smix 0 -dsur 0 -dnorm 31 -dynrng 5 - "C:\test2.ac3"
...
Aften: A/52 audio encoder
Version 0.06

Please use the last shon3i package

For Aften you need rev449 at least, because with 2h21m 6 channnls you need "-readtoeof 1" not supported by v0.06.

[idbirch2]

Thanks very much for your help. I did originally start out using build 449 but I must have been doing something else wrong then as having gone back to 449, BH is now working

[tauka]

hi guys, i need a little help. i have the newest behappy, and i try to convert an ac3 file from pal to ntsc 6ch wav, its 2h24min long.. the problem is this: i open the wav with any prog, it shows 2:04:17.. in delaycut in the info actually the correct length is shown (2h31min), but in the target file the 2:04:17 again.. what can be the problem? the wav is actually around 4,8gig.. any suggestion? is it possible at all to convert such a long track with behappy?
thanks

update: well, i actually managed to get the correct length, but only if i load in the six separate mono wavs and stretch those.. strange.. but the ntsc track isnt good, it goes out of sync at the end, but its correct in the beginning.. with adobe audition i get with time stretch perfect timing.. :s

[tebasuna51]

Originally Posted by tauka

hi guys, i need a little help. i have the newest behappy, and i try to convert an ac3 file from pal to ntsc 6ch wav, its 2h24min long.. the problem is this: i open the wav with any prog, it shows 2:04:17..

Then don't trust in these prog's.

in delaycut in the info actually the correct length is shown (2h31min), but in the target file the 2:04:17 again.. what can be the problem? the wav is actually around 4,8gig.. any suggestion?

Wav files have a header field with a 4GB limit (or 2:04:17 for wav 16 bit int, 6 channel, 48 KHz) but you have more data until 4.8 GB or 2h31m.
Edit: see also this link

Don't worry Aften can encode this long track with the parameter: -readtoeof 1.
If you want encode to aac, NeroAacEnc also support long files with -ignorelength parameter.

is it possible at all to convert such a long track with behappy?

Of course, and don't need the huge intermediate wav file. Just I make a test with an ac3 7846.336 seconds. ( 2 h. 10 m. 46.336 s.) and timestretch 25 -> 23.976 directly to ac3 and the result is 8181.472 seconds. ( 2 h. 16 m. 21.472 s.) with less than a frame error (32 ms) than 7846.336*25/23.976 = 8181,448

[thuongshoo]

Hi ! Thank all for continueing to developt BeHappy. I love all new feature.
I used to use OggdropXP. It can estimate bitrate in Q mode. I hope that a new version will has this feature.
Bye!

[thuongshoo]

tebasuna1! You seem to be the one which do last correct. Can you send a copy of newest version of BeHappy to me ?
Thank you!

[tebasuna51]

Originally Posted by thuongshoo

tebasuna1! You seem to be the one which do last correct. Can you send a copy of newest version of BeHappy to me ?

Last links resumed:
2007-03-24 Last Shon3i BeHappy package

2007-03-31 Chumbo mod for Timestretch and others v0.1.10

2007-04-01 BeHappy mod to support new enc_aacPlus.exe and MP4mux


Don't forget last Aften 0.07

[thuongshoo]

Oh! Thanks tebasuna51!
I didn't say clearly. I like source code
These pic decribe my words

[tebasuna51]

Last BeHappy sources.

Use http://www.imageshack.us/ to share pic's and avoid the "Attachments Pending Approval".

[tebasuna51]

Originally Posted by thuongshoo

I didn't say clearly. I like source code
These pic decribe my words

Sorry I can't reproduce your problem (I haven't Visual C++), I only obtain this:

[NoX1911]

I'm trying to convert AC3 to AAC 6ch. The source is a DVB stream (live concert). I don't know if it has DRC. How can i figure that out (metadata)?

If i use 'NicAC3Source (DRC)' what compression rate will be used? Most standalone dvd players have 3 profiles (low, med, high).

[tebasuna51]

Originally Posted by NoX1911

I'm trying to convert AC3 to AAC 6ch. The source is a DVB stream (live concert). I don't know if it has DRC. How can i figure that out (metadata)?

You need read the header like in http://forum.doom9.org/showthread.ph...860#post993860
Read the next posts also.

If i use 'NicAC3Source (DRC)' what compression rate will be used? Most standalone dvd players have 3 profiles (low, med, high).

The attenuation values supplied in the ac3 stream are used without correction. I don't know what is: low, med, high. Maybe half, full, double?

[NoX1911]

Ok.. just hoped there would be something like this where you could additionally modify the DRC (or change metadata/Line Mode Profile of AC3).

[tebasuna51]

Ac3Filter have a DRC function independent of the DRC data in ac3 stream. NicAudio is a decoder to apply only the DRC data in the stream, if you want addicional/different compression you can use SoxFilter.

[NoX1911]

I think its not a different compressor. It is intended by Dolby to have impact on DRC like this to optimize volume for different environments (small speakers (high compression/flat sound), big speakers (no compression/full dynamic)). That's the way 'dialogue normalization' works. There are official dolby whitepapers linked in the faq sticky. I think i read it there...

My settop dvd player has some static values (low, med, high) for DRC that do the same what Ac3Filter does (i think) so this must be intended by dolby.

The second screenshot i posted above holds some metadata for 'line mode profile'. These could be preset values for that additional compression and may be used by Nic's DRC method.

But i'm not sure on that...

Edit:
solved... neroaacenc.exe has to be moved to 'encoder' folder.

Beside of that BeHappy doesn't work with my scenario. I use BeHappy with aac update.
NicAc3Source(DRC) to load my Ac3 file. If i output as avs script it works fine with mpc. If i enqueue and start the batch process following error msg appears in the log:

Translation:
Das System kann die angegebene Datei nicht finden = File not found

Starting job ggg.ac3->ggg.mp4
Found Audio Stream
Channels=2, BitsPerSample=32 float, SampleRate=48000Hz
encoder\neroAacEnc.exe -ignorelength -q 0.3  -if - -of "F:\ggg.mp4"
Error: System.ApplicationException: Can't start encoder: Das System kann die angegebene Datei nicht finden ---> System.ComponentModel.Win32Exception: Das System kann die angegebene Datei nicht finden
   bei System.Diagnostics.Process.StartWithCreateProcess(ProcessStartInfo startInfo)
   bei System.Diagnostics.Process.Start()
   bei BeHappy.Encoder.createEncoderProcess(AviSynthClip x)
   --- Ende der internen Ausnahmestapelüberwachung ---
   bei BeHappy.Encoder.createEncoderProcess(AviSynthClip x)
   bei BeHappy.Encoder.encode()

solved... neroaacenc.exe has to be moved to 'encoder' folder.

[tebasuna51]

Originally Posted by NoX1911

My settop dvd player has some static values (low, med, high) for DRC that do the same what Ac3Filter does (i think) so this must be intended by dolby.

Not exactly. Your standalone, and soft decoders like PowerDVD, Azid, ffdshow, NicAudio, ... read the gain/attenuation value put in the ac3 stream for each block (5.33 ms) and apply the full value or a part only, don't calculate the DRC needed analyzing the sound decoded.
AC3Filter analyze the sound and calculate the DRC required.

The second screenshot i posted above holds some metadata for 'line mode profile'. These could be preset values for that additional compression and may be used by Nic's DRC method.

These presets are used at encoding time. The encoder calculate the appropriate gain/attenuation in function of these presets and the actual volume and put this value in the ac3 stream for each block (5.33 ms). The decoders don't need calculate, only read the values proposed by the encoder.

[thuongshoo]

This is a new version of BeHappy
http://www.box.net/shared/5tyvsczuxo

[alwa]

Originally Posted by thuongshoo

This is a new version of BeHappy
http://www.box.net/shared/5tyvsczuxo

@thuongshoo:You should list your changes.
I've noticed chages in OggVorbisEncoder.cs and CodingTechnologiesAAC.cs, please document them.

Maybe the mod should have it's own repository, something like a project page on sf.net.

[tebasuna51]

In CodingTechnologiesAAC.cs only size parameters to avoid the problem in post #568.

In OggVorbisEncoder.cs size parameters and:

if (Quality <= 4) ApproximateBitrate = (double)(Quality + 2)*16 + 32;
if ((Quality > 4) && (Quality <= 8)) ApproximateBitrate = (double)(Quality - 4) * 32 + 128;
if (Quality > 8 && Quality <= 9) ApproximateBitrate = (double)(Quality - 8) * 64 + 256;
if (Quality > 9) ApproximateBitrate = (double)((Quality - 9)) * 179.8 + 320;
rbtnVBR.Text = string.Format("Variable Bitrate Q={0} approximate {1} kbs", Quality,ApproximateBitrate);

I don't use ogg files and don't know if is correct or not, but maybe the number of channels need to be considered.

[thuongshoo]

I've noticed chages in OggVorbisEncoder.cs and CodingTechnologiesAAC.cs, please document them.

@alwa:I'm sorry!
CodingTechnologiesAAC.cs: I resized and changed position of component to avoid the problem in post #568
OggVorbisEncoder.cs: added "ApproximateBitrate" value.

I don't use ogg files and don't know if is correct or not, but maybe the number of channels need to be considered.

@Tebasuna51: I tested 2 file and feel Ok. Everyone please re-test

It is quite right if I use oggenc2 while oggenc show a lower bitrate.
I'm using MediaInfo to observer information of media file
This feature come from OggDropXPd


Link to download oggenc2 ( I prefer oggdropXPd V.1.8.9 using libVorbis v1.1.2 with IMPULSE_TRIGGER_PROFILE Option)
and OggDropXP
http://www.rarewares.org/ogg.html
Good luck!