Winamp FhG AAC encoder with gapless fix, Updates from the 5.62 release Options

[C.R.Helmrich]

Depends on what you call "transparent".
The encoder is indeed very efficient and every detail is clear at the highest VBR mode, but the sound is still kinda "compact" and a bit unnatural compared to the lossless files.

Transparent here means indistinguishable from the original audio, i.e. failing a blind ABX test. Please provide a sample whose VBR 5 encoding sounds different to you, so I can try it myself. And if you haven't done so, please familiarize yourself with the ABX testing concept and the Terms of Services of this forum.

Chris

Igor, yes, I tried to reduce bit-rate variance a bit, as noted in a related thread (http://www.hydrogenaudio.org/forums/index....st&p=766941)

This post has been edited by C.R.Helmrich: Oct 29 2011, 20:50

[ZinCh]

I have tried to get Winamp 5.562 standard, but I have legal problem with installation, probably its wrong build or beta release or wrong setup file :

setup inclides LGPL binaries libsndfile.dll and enc_lame.dll , but after installation I cannot find copy of LGPL license as text file (and information about where I can find source or patches for used lgpl libraries)

I was able successfuly download and install Winamp 5.562 lite version - looks like this file is legal (it didn't have LGPL parts, so it didn't have to ship copy of LGPL license text), but this edition didn't have FhG codec

please, add FhG codec dll as attach to this thread, so I can get this file from legal source

thanks

This post has been edited by ZinCh: Oct 30 2011, 11:40

[Thundik81]

can anybody help with the new .dll?

please, add FhG codec dll as attach to this thread, so I can get this file from legal source

1. Download latest standard version (http://download.nullsoft.com/winamp/client/winamp5622_full_emusic-7plus_all.exe )
2. Extract downloaded installer with 7-Zip (NSIS installer)
3. encoder is in Plugins directory

[lvqcl]

I have tried to get Winamp 5.562 standard, but I have legal problem with installation, probably its wrong build or beta release or wrong setup file
[...]
please, add FhG codec dll as attach to this thread, so I can get this file from legal source

Even Winamp 5.562 lite (and enc_fhgaac itself!) uses libmp4v2.dll (Mozilla Public License):

3.6. Distribution of Executable Versions.
You may distribute Covered Code in Executable form only if the
requirements of Section 3.1-3.5 have been met for that Covered Code,
and if You include a notice stating that the Source Code version of
the Covered Code is available under the terms of this License,
including a description of how and where You have fulfilled the
obligations of Section 3.2.

So it seems that you'll have legal problems anyway

[b66pak]

@Thundik81 thanks a lot...
_

[Anakunda]

Just curious, any chance to use this excellent codec with foobar or from command line? (without Winamp). I like this FhG thingy but dont want install Winamp just because of it.

[mudlord]

easy, someone can reverse engineer it, though fraunhofer and this forum frowns upon that.
egad!

[A_Man_Eating_Duc...]

Just curious, any chance to use this excellent codec with foobar or from command line? (without Winamp). I like this FhG thingy but dont want install Winamp just because of it.

You mean like THIS (thanks to nao for creating it)

[Anakunda]

Thanks so much.

[Gainless]

Just curious, any chance to use this excellent codec with foobar or from command line? (without Winamp). I like this FhG thingy but dont want install Winamp just because of it.

You mean like THIS (thanks to nao for creating it)

Does it work for Foobar, too?

[A_Man_Eating_Duc...]

why yes it does work with foobar

[A_Man_Eating_Duc...]

since i can't edit my previous post

CODE

--vbr 5 --ignorelength - %d

[nu774]

Hi,
I found channel mapping is incorrect for 4ch case.
From what I can see, resulting 4ch AAC has default channel layout in ISO spec (C L R Cs).
However, it seems L is mapped to C, R is mapped to L, and C is mapped to R.
(I don't know if it's winamp side or Fhg side problem).

[botface]

CODE

--vbr 5 --ignorelength - %d

These parameters aren't working for me in Foobar. I'm getting " Conversion failed: The encoder has terminated prematurely with code 0 (0x00000000); please re-check parameters". Am I overlooking something obvious?

[benski]

Hi,
I found channel mapping is incorrect for 4ch case.
From what I can see, resulting 4ch AAC has default channel layout in ISO spec (C L R Cs).
However, it seems L is mapped to C, R is mapped to L, and C is mapped to R.
(I don't know if it's winamp side or Fhg side problem).

It's a Winamp problem. There's not really good support at the moment for surround encoding. It's pending some API changes to make it work effectively.

[nu774]

It's a Winamp problem. There's not really good support at the moment for surround encoding. It's pending some API changes to make it work effectively.

Thanks. I noticed there's no way to specify channel layout in current Winamp plugin API;
Though 5ch, 5.1ch, and 7.1ch seemed to be fine.

[A_Man_Eating_Duc...]

CODE

--vbr 5 --ignorelength - %d

These parameters aren't working for me in Foobar. I'm getting " Conversion failed: The encoder has terminated prematurely with code 0 (0x00000000); please re-check parameters". Am I overlooking something obvious?

are these files all in the the same folder?

CODE

enc_fhgaac.dll
fhgaacenc.exe
libmp4v2.dll
libsndfile.dll
nsutil.dll

you'll find these in the winamp install folder

This post has been edited by A_Man_Eating_Duck: Nov 1 2011, 05:51

[Gainless]

I've done a little ABX test with a trance track, using the FhG encoder. The results are not perfect but i think it's at least excluded that I was totally guessing all the time. Test log and a sample are included in the attachement...

This post has been edited by Gainless: Nov 3 2011, 14:56

Attached File(s)

 ABX_Test___IM___Deeply_Disturbed.rar ( 4.15MB ) Number of downloads: 105

 

[C.R.Helmrich]

Thanks a lot for testing, Gainless! Unfortunately I can't hear any artifacts on this item myself. Which kind of artifact(s) did you hear, and in which passage? And I assume you used the latest Winamp version?

Chris

[Destroid]

For those who like numbers, I have quick dirty test results:

CODE

fhg aac 3.2.3 --vbr 4 115s 10.10%
fhg aac 3.2.3 --vbr 5 125s 15.60%
helix mp3 5.1 -V120 29s 13.68%
lame 3.98.4* -V3 140s 13.22%
lame 3.99.1 -V3 160s 13.07%
musepack 1.30 -q5 114s 12.09%
musepack 1.16 -q5 114s 12.42%
neroaac 1.0.7.0 -q0.5 190s 12.18%
neroaac 1.5.4.0 -q0.5 146s 12.63%
ogg 2.83 lancer -q5 78s 12.10%
ogg 2.87 lancermod -q5 106s 12.01%

*iccpatch applied (test system: Athlon 64 2.0GHz, 2GB DDR, WinXp SP3)

Comments:
- I saw 96kbps test and was intrigued by Fhg AAC as I never dealt with before (although my test is in the safe/transparent range);
- I used nao's so I could bench on commandline with all codecs (enc_fhgaac.dll from Winamp 5.622 package)
- Couldn't get Fhg AAC into close range with other codecs, quite a jump between --vbr 4 and --vbr 5, and decimal values ignored

[b66pak]

@C.R.Helmrich do you have any plans for increasing the encoding speed?

looking at this http://nyaochi.sakura.ne.jp/encoder-benchm...t-20061103.html

i can see that another fraunhofer tool (the mp3 encoder) is king of the speed...

so you have a very high standard for encoding speed to keep!
_

This post has been edited by b66pak: Nov 9 2011, 19:49

[Gainless]

Thanks a lot for testing, Gainless! Unfortunately I can't hear any artifacts on this item myself. Which kind of artifact(s) did you hear, and in which passage? And I assume you used the latest Winamp version?

Chris

I also don't hear any obvious artifacts, the aac just sounds a bit different to the wav, that's all. I guess I had some kind of "lucky" day when I made the abx test in my previous post, because most of the time I don't get any good results (< 6-10%)...

This post has been edited by Gainless: Nov 11 2011, 17:00

[IgorC]

If You don't listen the artifacts then it can be simply differences in the volume or time shifting.
It will be more safe to apply time shifting and gain to ABX sessions. Page 11 of http://www.rarewares.org/rja/ListeningTest.pdf

[Gainless]

If You don't listen the artifacts then it can be simply differences in the volume or time shifting.
It will be more safe to apply time shifting and gain to ABX sessions. Page 11 of http://www.rarewares.org/rja/ListeningTest.pdf

I'm already using DSP on the ABX tests in Foobar, why shouldn't be that enough?

[IgorC]

Foobar's ABX logs don't provide the information about DSP (Replaygain, etc.).
Your sample is ok. FhG LC-AAC doesn't put additional frames (delay) and volume is practically the same.





This post has been edited by IgorC: Nov 12 2011, 04:20