Aften 0.0.8 is out

[Atak_Snajpera]

Do you have Vista or XP? I have Vista.

I tried with wavi and I get this
http://img74.imageshack.us/img74/5738/new2iu8.th.png (http://img74.imageshack.us/my.php?image=new2iu8.png)

[Kurtnoise]

add "-readtoeof 1" in your aften command line...

[Atak_Snajpera]

Ok It's working now but with -readtoeof 1 progress is not working :(

[tebasuna51]

Ok It's working now but with -readtoeof 1 progress is not working :(

You cheat me with the size. :cool:
Your file is 8304.448 sec (2:18:24.448) and not 15:36.

The 4GB limit for 16bit-6chan-48KHz is 2:04:16.540 then the length in wav header (only 4 bytes) is 2:18:24.448 - 2:04:16.540 = 14:07.908 and Aften stop encoding at this time.

To force Aften continue encoding disregarding the length in wav header is need -readtoeof 1, but now Aften don't know the real total length and can't show the progress.

[Atak_Snajpera]

You cheat me with the size.
I didn't have time to send whole 450 MB file :)

To force Aften continue encoding disregarding the length in wav header is need -readtoeof 1, but now Aften don't know the real total length and can't show the progress.
BePipe works again when I add -readtoeof 1 and I get progress as well (from Bepipe not from Aften).

Thanks for help guys!

[Blue_MiSfit]

Gaaagh! The sourceforge page seems to be having issues downloading the latest versions of Aften, including the super-awesome-ohmigod-cantbelieveit E-AC3 encoding build!

Does anyone have a mirror available?

~MiSfit

[NormanBates]

@Blue_MiSfit:
I have downloaded the newest eac3 experimental build (r733) for you and uploaded it. The download link is:
http://www.sendspace.com/file/i52sr1

The latest revision (r762) is here:
http://www.sendspace.com/file/g05ada

Hope this helps you.

[Blue_MiSfit]

Thanks! We should update the main OP somehow

~MiSfit

[Adub]

Uh, has development stalled on this project? I am curious, as I have gotten into ac3 encoding recently, and I know that this was THE standard for a while. Plus there were rumors of EAC3 support, so I am wondering if the author could comment. His blog hasn't been updated in months and the svn code base is for the most part several months old.

Anyone able to comment?

[jruggle]

Development is temporarily on hold, but certainly not dead. I do intend to continue development at some point. I will definitely add E-AC-3 support, and that will likely be my first focus when I have the time.

[Adub]

Excellent!! Good to hear from you jruggle and I look forward to your future work!

[ACrowley]

For anyone interested here are available for download (https://sourceforge.net/project/showfiles.php?group_id=183195&package_id=212610&release_id=583753) updated Aften (http://aften.sourceforge.net/) binaries at revision 762 and here is available for download (https://sourceforge.net/project/showfiles.php?group_id=158644&package_id=219726&release_id=583773) new release of WAV to AC3 Encoder (http://www.thefrontend.net/EncWAVtoAC3/index.html) at version 1.2 (includes Aften R762).

Thanks,
wisodev

I like your Wave 2 Ac3 Gui for Aften..

2.0 is the latest Version ,right ? Is there a newer Version/updates ?

The Site "http://www.thefrontend.net/EncWAVtoAC3/" is down

[wisodev]

I like your Wave 2 Ac3 Gui for Aften..

2.0 is the latest Version ,right ? Is there a newer Version/updates ?

The Site "http://www.thefrontend.net/EncWAVtoAC3/" is down

The site will be no longer maintained! The development will be continued and there will be new releases. All the old pages are not valid any more, please use new project page : http://code.google.com/p/wavtoac3encoder/

[wisodev]

Aften binaries at revision 786 are available for download.

http://code.google.com/p/wavtoac3encoder/

[nautilus7]

Thanks a lot! Could you point where i can find a changelog?

[wisodev]

Thanks a lot! Could you point where i can find a changelog?

This is the official changelog: http://aften.svn.sourceforge.net/viewvc/aften/Changelog

[Yobbo]

Thank you Wisodev for the ongoing development of this!

I downloaded the new version of WavToAC3Encoder (2.1), but the Aften engine says R762. I downloaded the new binaries (R786), so do I just pop the new libaften.dll's into the WavToAC3Encoder subfolders to replace the old libaftens??

Actually I already did this but WavToAC3Encoder still says "R762"?

How do I keep this the most up-to-date please?

Thanks again! (I use it a lot!).

[wisodev]

Thank you Wisodev for the ongoing development of this!

I downloaded the new version of WavToAC3Encoder (2.1), but the Aften engine says R762. I downloaded the new binaries (R786), so do I just pop the new libaften.dll's into the WavToAC3Encoder subfolders to replace the old libaftens??

Actually I already did this but WavToAC3Encoder still says "R762"?

How do I keep this the most up-to-date please?

Thanks again! (I use it a lot!).
Well it should work with this version of Aften, if you want to see new version number in main window of program you should edit configuration files manually as I do when updating Aften libraries (depending on release these settings are located in file: EncWAVtoAC3.win32 or EncWAVtoAC3.win64).

Example for win32 release:
Aften R786=libaftendll_x86\libaften.dll
Aften R786 SSE=libaftendll_x86_SSE\libaften.dll
Aften R786 SSE2=libaftendll_x86_SSE2\libaften.dll
Aften R786 SSE3=libaftendll_x86_SSE3\libaften.dll
PS. As far I know there where some changes in functionality between this releases so please be careful and test before doing any work. I will update GUI to include this changes, so for now it is recommend to use Aften R762.

[ACrowley]

ive a question about DD EX encoding. The Back Center Channel is greyed out/not avaibele In your Mux Wizard?
Is it possible to unlock the Back Center Channel for 6.1 (Matrix encoded ofcourse) DD EX encoding ?

[tebasuna51]

ive a question about DD EX encoding. The Back Center Channel is greyed out/not avaibele In your Mux Wizard?
Is it possible to unlock the Back Center Channel for 6.1 (Matrix encoded ofcourse) DD EX encoding ?

Aften can't encode 7 channels (6.1) if you see the channel config the max channels are 5.1
When you select 3/1 the SL-SR are greyed and appear the S
(only 1 surround channel).

If you have 3 surround channels and previously matrix the SC in SL and SR you can use Aften to encode and set the flag in header:
'Dolby Surround EX mode' to 'Dolby Surround EX encoded'

I think you can use Sox to matrix the SC in SL-SR:

sox -m -v 0.5858 SL.wav -v 0.4141 SC.wav new-SL.wav
sox -m -v 0.5858 SR.wav -v 0.4141 SC.wav new-SR.wav

[madshi]

I think you can use Sox to matrix the SC in SL-SR:

sox -m -v 0.5858 SL.wav -v 0.4141 SC.wav new-SL.wav
sox -m -v 0.5858 SR.wav -v 0.4141 SC.wav new-SR.wav
Don't you need to apply a phase shift for matrixing?

[tebasuna51]

Don't you need to apply a phase shift for matrixing?

I think is not necesary.

In dpl downmix the FC channel is mixed (and recovered) with FL-FR without phase shift.

The phase shift can be necesary in dpl if FL and SL (and FR with SR) are dependent and the mix between can produce problems.

BTW, you can test using:
sox -m -v 0.5858 SL.wav -v 0.4141 SC.wav new-SL.wav
sox -m -v -0.5858 SR.wav -v 0.4141 SC.wav new-SR.wav

[ACrowley]

Aften can't encode 7 channels (6.1) if you see the channel config the max channels are 5.1
When you select 3/1 the SL-SR are greyed and appear the S
(only 1 surround channel).

If you have 3 surround channels and previously matrix the SC in SL and SR you can use Aften to encode and set the flag in header:
'Dolby Surround EX mode' to 'Dolby Surround EX encoded'

I think you can use Sox to matrix the SC in SL-SR:

sox -m -v 0.5858 SL.wav -v 0.4141 SC.wav new-SL.wav
sox -m -v 0.5858 SR.wav -v 0.4141 SC.wav new-SR.wav

Ok, thx. However, i use AAC or dts es 6.1 discrete to reencode 7Ch Tracks.
But its the same Problem.. we have no (100%) proper Method to decode DD EX to 6.1/7ch waves ,also not for dts.
But this is the aften encoder thread ......:)

By the Way : Ive tested wave2ac3 2.1 x264 with rev786 SSE3Wow! The Perfomance is great. ~30% faster compared with DD Encoder Pro v7
The Quality is "fully" identical to the off. Dolby Digital Encoder Pro v7. Maybe i will switch to Aften instead of DD Enc Pro 7 from SonyVegas 8. Aften has no disadvantage

[wisodev]

By the Way : Ive tested wave2ac3 2.1 x264 with rev786 SSE3Wow! The Perfomance is great. ~30% faster compared with DD Encoder Pro v7
The Quality is "fully" identical to the off. Dolby Digital Encoder Pro v7. Maybe i will switch to Aften instead of DD Enc Pro 7 from SonyVegas 8. Aften has no disadvantage

Did you try Aften on multi-core CPU? Aften is even faster with that kind of systems!

Tip:
To get even more performance set the:
"Fast bit allocation" to "Faster encoding"
and
"Fast exponent strategy decision" to "Faster encoding"

there is probably some quality impact with these settings.

[ACrowley]

Did you try Aften on multi-core CPU? Aften is even faster with that kind of systems!

Tip:
To get even more performance set the:
"Fast bit allocation" to "Faster encoding"
and
"Fast exponent strategy decision" to "Faster encoding"

there is probably some quality impact with these settings.

Im encoding on a quadcore !6600 @ 3.2ghz..
No, i dont use FastBitAllocation/FastEx.Startegiy.D.....Its fast enough.

[menlvd]

why encoding (wavtoac3) don't work with latest build 826 and 786
get - [2008-11-06, 20:43:48.828] Encoder Error: Failed to initialize encoder.

[wisodev]

why encoding (wavtoac3) don't work with latest build 826 and 786
get - [2008-11-06, 20:43:48.828] Encoder Error: Failed to initialize encoder.

Please read this post: http://forum.doom9.org/showthread.php?p=1207826#post1207826

[raquete]

wisodev, thanks so much for the news!

[wisodev]

WAV to AC3 Encoder version 2.2 was released (changelog (http://wavtoac3encoder.googlecode.com/svn/trunk/doc/Changes.txt)).

Aften Win32 and x64 binaries svn revision 832 where released.

Download all from here. (http://code.google.com/p/wavtoac3encoder/)

[Yobbo]

:thanks:

[raquete]

thank you wisodev. :)

[wisodev]

WAV to AC3 Encoder version 2.3 is available for download (http://code.google.com/p/wavtoac3encoder/) (changes (http://wavtoac3encoder.googlecode.com/svn/trunk/doc/Changes.txt)).

[totya]

WAV to AC3 Encoder version 2.3 is available for download (http://code.google.com/p/wavtoac3encoder/) (changes (http://wavtoac3encoder.googlecode.com/svn/trunk/doc/Changes.txt)).

Thanks!

[Boulder]

Is there a changelog regarding the various Aften SVN builds after v0.0.8?

[wisodev]

Is there a changelog regarding the various Aften SVN builds after v0.0.8?

http://aften.svn.sourceforge.net/viewvc/*checkout*/aften/Changelog

for changes after 0.08 please check point "version SVN : current"

i don't think there is changelog for each SVN build, you may try checking same revision of Changelog

[Boulder]

Thanks :)

[raquete]

wisodev,
thanks so much one more time, wav2ac3enc stills champion!

[TFM_TheMask]

What does the new mode of operation mean. And specific what does Transcoding do?

[jruggle]

What does the new mode of operation mean. And specific what does Transcoding do?
What exactly are you referring to?

Transcoding would be AC3-to-AC3, with essentially a decode/re-encode. Aften is setup to be able to handle this in the future, but does not currently do so. The main obstacle is that the decoder that would be used is currently under GPL. If/when it is ever relicensed to LGPL then it will be ported over to Aften from FFmpeg.

[wisodev]

I have been working on Avisynth scripts support in WAV to AC3 Encoder, finally I have added some preliminary support for .avs scripts. Please find attached the version 3.0 BETA 1 of WAV to AC3 Encoder, it really needs a lot of testing before final release.

Changes in v3.0 BETA 1:
- Added preliminary Avisynth (*.avs files) scripting support (not supported in 'Multiple mono input' and in 'One per file' mode).
- Added *.avs filter to open dialog.
- Double-click on *.avs script in files list to show AVS File Properties window.

DOWNLOAD: WAV to AC3 Encoder 3.0 BETA 1 (http://wavtoac3encoder.googlecode.com/files/EncWAVtoAC3-3.0_BETA_1-Win32-bin.zip)

[TFM_TheMask]

What exactly are you referring to?

Transcoding would be AC3-to-AC3, with essentially a decode/re-encode. Aften is setup to be able to handle this in the future, but does not currently do so. The main obstacle is that the decoder that would be used is currently under GPL. If/when it is ever relicensed to LGPL then it will be ported over to Aften from FFmpeg.

Ok :thanks:. I also saw in the todo list that you want to implement an upmix feature. Is this something that is coming soon or not? Maybe you can take a look at the upmix feature of the AC3Filter (DirectShow).

[jruggle]

Ok :thanks:. I also saw in the todo list that you want to implement an upmix feature. Is this something that is coming soon or not? Maybe you can take a look at the upmix feature of the AC3Filter (DirectShow).
I'm sorry, but it's not high on my priority list. The highest priority for Aften is improving the core encoder, which mostly involves channel coupling, better bit allocation, and improved DRC. After that, it's E-AC3, so unless I suddenly find myself with massive amounts of free time or some volunteer developers, it likely will be some time before I get to the less important items on the list.

[Boulder]

How's the development going on regarding those important items?

[TFM_TheMask]

I'm sorry, but it's not high on my priority list. The highest priority for Aften is improving the core encoder, which mostly involves channel coupling, better bit allocation, and improved DRC. After that, it's E-AC3, so unless I suddenly find myself with massive amounts of free time or some volunteer developers, it likely will be some time before I get to the less important items on the list.

Then for the time being I'll stick to Ac3Filter. Keep up the good work.:thanks:

[madshi]

I'm not using the latest build, so maybe it has been fixed in the meanwhile, but maybe you Aften guys can double check?

If I encode a 44.1khz AC3 file, the framesize toggles between two different values. So far, so good. The frames with the smaller framesize are fully correct. However, the frames with the bigger framesize have an incorrect 2nd CRC (the first CRC is correct), as far as I can say. Both delaycut and eac3to don't like the 44.1khz files created by Aften because of this problem.

Could you please check that?

(I've tried both 448kbps and 640kbps, 5.1 channels).

[jruggle]

I'm not using the latest build, so maybe it has been fixed in the meanwhile, but maybe you Aften guys can double check?

If I encode a 44.1khz AC3 file, the framesize toggles between two different values. So far, so good. The frames with the smaller framesize are fully correct. However, the frames with the bigger framesize have an incorrect 2nd CRC (the first CRC is correct), as far as I can say. Both delaycut and eac3to don't like the 44.1khz files created by Aften because of this problem.

Could you please check that?

(I've tried both 448kbps and 640kbps, 5.1 channels).
I can't duplicate the problem with the latest build, but I don't have delaycut or eac3to. liba52 and libavcodec both check the CRCs and they aren't having any problems with a 44.1kHz 5.1-ch file I just made. Could you check again with the latest version from SVN?

[madshi]

I can't duplicate the problem with the latest build, but I don't have delaycut or eac3to. liba52 and libavcodec both check the CRCs and they aren't having any problems with a 44.1kHz 5.1-ch file I just made. Could you check again with the latest version from SVN?
Edit: I've just checked with wisodev's "WAV to AC3 Encoder 3.0 BETA 1" and the problem doesn't occur with that one. So I guess it's fixed in SVN. I'll update to the latest SVN next week.

Thanks!

[nautilus7]

I'll update to the latest SVN next week.At last! :p

[gizzin]

How's the development going on regarding those important items?

I'm curious myself :)

[jruggle]

How's the development going on regarding those important items?
I have channel coupling working locally, but it's basically just on or off with fixed parameters. Using coupling in the variable bandwidth mode will be my ultimate goal for bit allocation. And I at least have a somewhat detailed plan:

A table can be created, similar to the bandwidth bits table, in order to estimate the total number of bits for a given coupling setting. Here is an outline for how it could work:

Set bit allocation map at target quality/bitrate setting.

mantissa/exp bit estimate for each fbw bin up to target bandwidth for all channels/blocks
mantissa/exp bit estimate for each coupling band up to target bandwidth for coupling channel/all blocks
overhead bit estimate for adding coupling params
bit estimates for coupling coordinates

Start bit count at full target bandwidth w/o coupling. If it fits, goto 12.
Enable coupling

start band = min(target-1, 15)
end band = target
coord blocks = 6

Incrementally decrease start band down to target-5 while frame is still too large. If it fits, goto 12.
Reset start/end bands. Change coord blocks to match exponent strategy. If it fits, goto 12.
Incrementally decrease start band down to target-5 while frame is still too large. If it fits, goto 12.
Reset start/end bands. Change coord blocks to 1. If it fits, goto 12.
Incrementally decrease start band down to zero while frame is still too large. If it fits, goto 12.
If bandwidth priority is selected, goto 12. If quality priority is
selected, continue to step 10.
Incrementally decrease end band down to 3 while frame is still too large. If it fits, goto 12.
Turn off coupling. Set bwcode to target and incrementally decrease bwcode down to zero while frame is still too large. (current vbw mode)
Always reencode in CBR mode with final settings to get the final result.


I think the toughest part to get right about this will be the bandwidth/quality tradeoff since bandwidth also affects the overall sound quality.

As for improving DRC... I know what I need to look into (attack/decay rates), but I haven't started to do so yet. It also presents difficulties with threaded encoding which I have to explore more thoroughly.