Welcome to Doom9's Forum, THE in-place to be for everyone interested in DVD conversion.

Before you start posting please read the forum rules. By posting to this forum you agree to abide by the rules.

 

Go Back   Doom9's Forum > General > Audio encoding

Reply
 
Thread Tools Search this Thread Display Modes
Old 9th September 2009, 14:04   #1001  |  Link
raquete
Registered User
 
Join Date: Nov 2004
Posts: 741
oh, i don't knew about aften -longhelp option.
EncWAVtoAC3 have one kind of "online help" when you let the mouse over the option but dissapear fast and you have to take the mouse there again...between this blinkings moments when the help appear and dissapear, me, with my magnific english forgot what i read and need to restart the lecture thousands times...lol
better is print the -longhelp!
i will do one short list(really short) with specifics questions.
thanks tebasuna.
__________________
search raquete
raquete is offline   Reply With Quote
Old 9th September 2009, 14:18   #1002  |  Link
wisodev
Registered User
 
Join Date: Nov 2006
Posts: 134
Quote:
Originally Posted by raquete View Post
wisodev
when and if you have time, can you do a "how to" explaining each parameter of advanced mode/options and advanced options in EncWAVtoAC3?
(for boreds users like me )
thanks.
Tip: Each advanced option has a tool tip if you mouse over it in program window, even each item in adv. options list has tool tip based on Aften -longhelp text.

Edit: Ups I was to slow
wisodev is offline   Reply With Quote
Old 9th September 2009, 16:45   #1003  |  Link
OAKside
Registered Lurker
 
OAKside's Avatar
 
Join Date: Oct 2006
Posts: 12
Quote:
Originally Posted by raquete View Post
OAKside, what i upload for you was EncWAVtoAC3-4.1 with Aften R832 and works perfectly, right? (or not? )
now is ok as wisodev reupload "everything" but i'm curious about what i send for you.
Thanks raquete. EncWAVtoAC3 does not include the various Aften EXEs (only DLLs), which I wanted to test in a few other apps.
(AVStoDVD, MeGUI, etc.) I didn't want to bug you about it again. I appreciate the help and response, guys.

I noticed Aften R843, happily tried out the EXE (MeGUI) and that's when I noticed the weird/slow behavior compared to R762.
So I was looking for a slightly older build. (It's worth mentioning that I haven't reproduced weird R843 behavior again, though. )
OAKside is offline   Reply With Quote
Old 9th September 2009, 20:02   #1004  |  Link
raquete
Registered User
 
Join Date: Nov 2004
Posts: 741
all that i have is here, if you're interested in any or in all, i upload for you:
dates here appear as day/month/year

aften_20060709(09/07/2006), aften_parallel(22/11/2006), aften-0.0.8.tar(19/07/2008), aften-0.01(11/07/2006), aften-0.02.tar(11/07/2006), aften-0.02(15/07/2006), aften-0.03(20/07/2006), aften-0.03-dev(23/04/2006), aften-0.03-dev new(30/07/2006), aften-0.03-dev(05/08/2006), aften-0.04(05/08/2006), Aften-0.04_rev13(06/08/2006), Aften-0.04_rev27(09/08/2006), aften-0.04kurtniose(06/08/2006), aften-0.04-win32-i586(05/08/2006), aften-0.04-win32-i686(05/08/2006, aften-0.05(09/01/2007), Aften-0.05_rev185(23/10/2006), Aften-0.05_rev211(28/10/2006), aften-0.05_rev214(30/10/2006), aften-0.05-win32, aften-current.tar(20/07/2006), aften-daily-080606.tar(06/08/2006), AftenGui 1.0,1.1,1.2,1.3,1.4.
another folder(maybe can repeat some versions): aften_rev382(18/02/2007), aften_rev467(18/03/2007), aften-0.03(09/01/2007), aften-0.04(09/01/2007), aften-0.04-win32-i686(Justin)(10/01/2007), aften-0.05_rev216(09/01/2007), aften-0.05-win32(Justin)(10/01/2007), aften-svn-r216-win32-bin(Wisodev)(10/01/2007), aften-svn-r224-bin(Wisodev)(10/01/2007), aften-svn-r224-bin(09/01/2007), aften-svn-r433-win32-bin(27/02/2007).

EncWAVtoAC3: seems that have all versions.(16 versions)
__________________
search raquete
raquete is offline   Reply With Quote
Old 2nd October 2009, 12:59   #1005  |  Link
Paulcat
Registered User
 
Join Date: Jan 2005
Location: Great White North
Posts: 326
Inexplicable Behaviour

Due to a strange failure in my BeSweet/BeLight combo, the wav file I was trying to encode to an ac3 file ended up with no sound. I needed a quick fix and went to FFMpegGUI (0.3) and to WavToAC3Enc (4.1). I tested both with some small wav files (Windows Startup Sound!) and only the FFMpeg version of the ac3 file would play (Using AC3 Filter 1.65, the latest one, and Media Player Classic 6.4.9.1).

Since I was pressed for time, I had to use FFMpeg to do my main file and build my dvd, but it is quite old (adding newer versions of FFMpeg.exe to the directory caused the same error in playing the WavToAC3Enc converted files, namely the "Cannot render the file" error from MPC)

BeSweet/BeLight always worked for me before, I don't know if there was some strange error in the wav file or whether the files on my HD got corrupted, but I still need a way to do the conversion.

I'm assuming that WavToAC3Enc is producing genuine ac3 files, what can I use to play them on my pc? I don't want to go through the trouble to create a physical dvd and find out after that it won't play correctly in my set top!

Paul
Paulcat is offline   Reply With Quote
Old 2nd October 2009, 20:34   #1006  |  Link
raquete
Registered User
 
Join Date: Nov 2004
Posts: 741
play your file with winamp(using AC3 Filter, can be found in winamp or AC3 filter homepages)
__________________
search raquete
raquete is offline   Reply With Quote
Old 6th October 2009, 13:03   #1007  |  Link
Paulcat
Registered User
 
Join Date: Jan 2005
Location: Great White North
Posts: 326
So this is just an issue with MPC? I've never had an issue before...
Paulcat is offline   Reply With Quote
Old 6th October 2009, 23:20   #1008  |  Link
tebasuna51
Moderator
 
tebasuna51's Avatar
 
Join Date: Feb 2005
Location: Spain
Posts: 5,150
Quote:
Originally Posted by Paulcat View Post
... I needed a quick fix and went to FFMpegGUI (0.3) and to WavToAC3Enc (4.1). I tested both with some small wav files (Windows Startup Sound!) and only the FFMpeg version of the ac3 file would play ...
Please upload your converted ac3 files.
__________________
BeHappy, AviSynth audio transcoder, in Doom9 forums. NicAudio, BassAudio, audio decoders.
tebasuna51 is offline   Reply With Quote
Old 6th November 2009, 18:27   #1009  |  Link
Vincent Vega
Registered User
 
Join Date: Feb 2004
Location: the tiny bubble of Creation
Posts: 174
an extra frame in dts->ac3 transcoding

hi everyone, i just used megui .1051 (aften r832 sse, nicaudio 2009-06-25, all the other updates too) to convert 5.1 dts to 5.1 ac3 (default aften settings, except bitrate 448).

as i understand the theoretical math for my audio stream should be:
dts: 763872 frames x 512 samples/frame = 391102464 (2:15:47)
to
ac3: 254624 frames x 1536 samples/frame = 391102464 (2:15:47)

now valex's ac3file filter property page shows one extra frame in the encoded ac3 stream (254625) and duration 2:15:48

is this normal or not? heres ac3file screenshots in attach:
Attached Images
 
Vincent Vega is offline   Reply With Quote
Old 6th November 2009, 21:03   #1010  |  Link
tebasuna51
Moderator
 
tebasuna51's Avatar
 
Join Date: Feb 2005
Location: Spain
Posts: 5,150
Each time you encode to ac3 (commercial encoders also) a initial silence of 256 samples is padded before the source.
For a 48 KHz samplerate audio means a delay of 5.333 ms and the output is 5.333 ms long needing a frame more.

You can avoid this delay usin the Aften parameter -pad 0
__________________
BeHappy, AviSynth audio transcoder, in Doom9 forums. NicAudio, BassAudio, audio decoders.
tebasuna51 is offline   Reply With Quote
Old 6th November 2009, 23:03   #1011  |  Link
Vincent Vega
Registered User
 
Join Date: Feb 2004
Location: the tiny bubble of Creation
Posts: 174
thank you very much tebasuna51 for the explanation. then i have one more question: now to do the best to sync this ac3back to video i'd need a negative audio delay of 6ms and mkvmerge does allow delays of multiples of 2ms. at the same time from what i remember audio delays can only work if they are multiples of frame size (32ms in this case), so telling mmg to use -6ms delay will only result in some unpredictable behavior right?
Vincent Vega is offline   Reply With Quote
Old 6th November 2009, 23:53   #1012  |  Link
tebasuna51
Moderator
 
tebasuna51's Avatar
 
Join Date: Feb 2005
Location: Spain
Posts: 5,150
Forget the delay or use BeHappy instead MeGUI to put the -pad 0 parameter
__________________
BeHappy, AviSynth audio transcoder, in Doom9 forums. NicAudio, BassAudio, audio decoders.
tebasuna51 is offline   Reply With Quote
Old 7th November 2009, 02:43   #1013  |  Link
Terrachild
Registered User
 
Join Date: Aug 2009
Posts: 10
Help with settings please

I'm new to Aften, I'm using "WAV to AC3 Encoder 4.1" as the GUI, and boy are there a lot of settings.

I'm making 5.1 surround files.

If I want to get the best sound possible, what should I be changing. I don't care how slow the encoder is, just best quality.

Thanks.
Terrachild is offline   Reply With Quote
Old 7th November 2009, 03:10   #1014  |  Link
tebasuna51
Moderator
 
tebasuna51's Avatar
 
Join Date: Feb 2005
Location: Spain
Posts: 5,150
Use 640 Kb/s, -readtoeof 1 (for big files) and defaults.
__________________
BeHappy, AviSynth audio transcoder, in Doom9 forums. NicAudio, BassAudio, audio decoders.
tebasuna51 is offline   Reply With Quote
Old 7th November 2009, 16:51   #1015  |  Link
Vincent Vega
Registered User
 
Join Date: Feb 2004
Location: the tiny bubble of Creation
Posts: 174
hi tebasuna51, -pad 0 and behappy did the trick, thanx a lot.

perhaps you could kindly help me figure out one more thing: i converted this 5.1 dts to 5.1 ac3, i used Norm. to 100% and encoder defaults. now im playing these files using ac3filter and downmixed to stereo (old ac97 alc101 on this machine ).

with exactly the same ac3filter settings and same mixer matrix, dts sounds significantly louder, why this loudness difference? actually i've had the same issue with all my previous dts->ac3 megui encodes.
Vincent Vega is offline   Reply With Quote
Old 7th November 2009, 17:50   #1016  |  Link
Abradoks
Registered User
 
Join Date: Mar 2008
Posts: 71
Objective comparison of AC3 encoders

Source: DTS 1510 kbit/s 48kHz 6ch (Pirates of the Caribbean: At World's End)
Decoding into WAV (PCM 24 bit) — eac3to (libav).
SNR measurement (bigger is better) and padding detection — CompAudio
PEAQ measurement (bigger is better) — PQevalAudio
Padding — delaycut
WAV splitting — BeSplit
Spectrograms — Sound Forge

aften r832 (from MeGUI)
Dolby Digital Encoder 6.2.2 (Sonic Foundry Soft Encode 1.0.19): output.DD6448.ac3 (448 kbps), output.DD6224.ac3 (224 kbps)
Dolby Digital Encoder 7.0 (Sony Vegas Pro 9.0.563): output.DD7448.ac3 (448 kbps), output.DD7224.ac3 (224 kbps)

Difference between aften_wide and aften_spec:
- aften (wide) encodes full frequency range: -w 60
- aften (spec) encodes same frequency range as DD encoders:
448 kbps -> 20.34 kHz -> w=48
224 kbps -> 9.05 kHz -> w=8

Preprocessing in Dolby Encoder was disabled:


Comparison:
Code:
Name		|  SNR	| PEAQ

448 kbps, ch L

DD7		| 26.297| -0.226
DD6		| 26.268| -0.222
aften		| 20.653| -0.478
aften (spec)	| 18.504| -0.288
aften (wide)	| 19.368| -1.539

Aften gets less points than proprietary Dolby Digital Encoder (here referred to as "DD") both in SNR and PEAQ.
DD7 has slightly better results than DD6.

Average frequency graphs:

As you can see, aften has bigger frequency cutoff by default than DD.
With "-w 48" (aften_spec) it matches DD and more than that, it saves high frequency average energy closer to original.
But in all three cases aften has higher energy on low frequencies then original track.

DD6 vs DD7:


Time-frequency spectrograms
Original:		aften_wide:	aften_spec:		DD7:		DD6:



448 kbps, ch C

DD7  		| 26.039| -0.211
DD6		| 26.003| -0.208
aften		| 20.349| -0.379
aften (spec)	| 18.274| -0.263
aften (wide)	| 19.108| -0.975

Average frequency graphs:


Time-frequency spectrograms
Original:		aften_spec:	DD7:



224 kbps, ch L

DD7  		| 21.187| -2.266
DD6		| 20.845| -2.275
aften		| 15.009| -2.584
aften (spec)	| 14.830| -2.596
aften (wide)	| 6.6604| -3.527

Average frequency graphs:



224 kbps, ch C

DD7		| 19.303| -2.227
DD6		| 19.213| -2.239
aften		| 14.542| -2.601
aften (spec)	| 14.378| -2.583
aften (wide)	| 6.5879| -3.139


448 kbps, all channels (SNR only)

aften L		| 20.653 ||  DD7 L	| 26.297
aften R		| 20.525 ||  DD7 R	| 26.490
aften C		| 20.349 ||  DD7 C	| 26.039
aften SL	| 21.137 ||  DD7 SL	| 25.819
aften SR	| 21.003 ||  DD7 SR	| 25.541
aften LFE	| 22.176 ||  DD7 LFE	| -0.039484 ||  DD6 LFE	| 29.818

When running tests on all channels this strange problem with LFE was discovered.
I have rechecked everything several times, but DD7 was still getting bad results on LFE.
As you see DD6 hasn't this problem. More tests required to check this out.

Average frequency graphs (LFE):


Time-frequency spectrograms (LFE)
Original:		aften_spec:	DD7:
Commandline:
Code:
aften.exe -b 448 -pad 0 -w 60 -s on -acmod 7 -lfe 1 -readtoeof 1 input.wav output.aften448_wide.ac3
aften.exe -b 448 -pad 0 -w 48 -s on -acmod 7 -lfe 1 -readtoeof 1 input.wav output.aften448_spec.ac3
aften.exe -b 448 -pad 0 -acmod 7 -lfe 1 -readtoeof 1 input.wav output.aften448.ac3
aften.exe -b 224 -pad 0 -w 60 -s on -acmod 7 -lfe 1 -readtoeof 1 input.wav output.aften224_wide.ac3
aften.exe -b 224 -pad 0 -w 8 -s on -acmod 7 -lfe 1 -readtoeof 1 input.wav output.aften224_spec.ac3
aften.exe -b 224 -pad 0 -acmod 7 -lfe 1 -readtoeof 1 input.wav output.aften224.ac3

C:\comparison\eac3to\eac3to.exe output.aften448_wide.ac3 output.aften448_wide.wavs
del output.*.R.wav output.*.LFE.wav output.*.SL.wav output.*.SR.wav
C:\comparison\eac3to\eac3to.exe output.aften448_spec.ac3 output.aften448_spec.wavs
del output.*.R.wav output.*.LFE.wav output.*.SL.wav output.*.SR.wav
C:\comparison\eac3to\eac3to.exe output.aften448.ac3 output.aften448.wavs
del output.*.R.wav output.*.LFE.wav output.*.SL.wav output.*.SR.wav
C:\comparison\eac3to\eac3to.exe output.aften224_wide.ac3 output.aften224_wide.wavs
del output.*.R.wav output.*.LFE.wav output.*.SL.wav output.*.SR.wav
C:\comparison\eac3to\eac3to.exe output.aften224_spec.ac3 output.aften224_spec.wavs
del output.*.R.wav output.*.LFE.wav output.*.SL.wav output.*.SR.wav
C:\comparison\eac3to\eac3to.exe output.aften224.ac3 output.aften224.wavs
del output.*.R.wav output.*.LFE.wav output.*.SL.wav output.*.SR.wav
C:\comparison\eac3to\eac3to.exe output.DD6448.ac3 output.DD6448.wavs
del output.*.R.wav output.*.LFE.wav output.*.SL.wav output.*.SR.wav
C:\comparison\eac3to\eac3to.exe output.DD6224.ac3 output.DD6224.wavs
del output.*.R.wav output.*.LFE.wav output.*.SL.wav output.*.SR.wav
C:\comparison\eac3to\eac3to.exe output.DD7448.ac3 output.DD7448.wavs
del output.*.R.wav output.*.LFE.wav output.*.SL.wav output.*.SR.wav
C:\comparison\eac3to\eac3to.exe output.DD7224.ac3 output.DD7224.wavs
del output.*.R.wav output.*.LFE.wav output.*.SL.wav output.*.SR.wav

delaycut.exe -start 5.33 -out input.L.pad256.wav input.L.wav
delaycut.exe -start 5.33 -out input.C.pad256.wav input.C.wav
delaycut.exe -start 13.25 -out input.L.pad636.wav input.L.wav
delaycut.exe -start 13.25 -out input.C.pad636.wav input.C.wav

C:\comparison\AFsp\CompAudio.exe input.L.wav output.aften448_wide.L.wav > output.aften448_wide.L.comp.txt
C:\comparison\AFsp\CompAudio.exe input.C.wav output.aften448_wide.C.wav > output.aften448_wide.C.comp.txt
C:\comparison\AFsp\CompAudio.exe input.L.wav output.aften448_spec.L.wav > output.aften448_spec.L.comp.txt
C:\comparison\AFsp\CompAudio.exe input.C.wav output.aften448_spec.C.wav > output.aften448_spec.C.comp.txt
C:\comparison\AFsp\CompAudio.exe input.L.wav output.aften448.L.wav > output.aften448.L.comp.txt
C:\comparison\AFsp\CompAudio.exe input.C.wav output.aften448.C.wav > output.aften448.C.comp.txt
C:\comparison\AFsp\CompAudio.exe input.L.wav output.aften224_wide.L.wav > output.aften224_wide.L.comp.txt
C:\comparison\AFsp\CompAudio.exe input.C.wav output.aften224_wide.C.wav > output.aften224_wide.C.comp.txt
C:\comparison\AFsp\CompAudio.exe input.L.wav output.aften224_spec.L.wav > output.aften224_spec.L.comp.txt
C:\comparison\AFsp\CompAudio.exe input.C.wav output.aften224_spec.C.wav > output.aften224_spec.C.comp.txt
C:\comparison\AFsp\CompAudio.exe input.L.wav output.aften224.L.wav > output.aften224.L.comp.txt
C:\comparison\AFsp\CompAudio.exe input.C.wav output.aften224.C.wav > output.aften224.C.comp.txt
C:\comparison\AFsp\CompAudio.exe -l 256: input.L.pad256.wav output.DD6448.L.wav > output.DD6448.L.comp.txt
C:\comparison\AFsp\CompAudio.exe -l 256: input.C.pad256.wav output.DD6448.C.wav > output.DD6448.C.comp.txt
C:\comparison\AFsp\CompAudio.exe -l 256: input.L.pad256.wav output.DD6224.L.wav > output.DD6224.L.comp.txt
C:\comparison\AFsp\CompAudio.exe -l 256: input.C.pad256.wav output.DD6224.C.wav > output.DD6224.C.comp.txt
C:\comparison\AFsp\CompAudio.exe -l 636: input.L.pad636.wav output.DD7448.L.wav > output.DD7448.L.comp.txt
C:\comparison\AFsp\CompAudio.exe -l 636: input.C.pad636.wav output.DD7448.C.wav > output.DD7448.C.comp.txt
C:\comparison\AFsp\CompAudio.exe -l 636: input.L.pad636.wav output.DD7224.L.wav > output.DD7224.L.comp.txt
C:\comparison\AFsp\CompAudio.exe -l 636: input.C.pad636.wav output.DD7224.C.wav > output.DD7224.C.comp.txt

C:\comparison\AFsp\PQevalAudio.exe input.L.wav output.aften448_wide.L.wav > output.aften448_wide.L.PEAQ.txt
C:\comparison\AFsp\PQevalAudio.exe input.C.wav output.aften448_wide.C.wav > output.aften448_wide.C.PEAQ.txt
C:\comparison\AFsp\PQevalAudio.exe input.L.wav output.aften448_spec.L.wav > output.aften448_spec.L.PEAQ.txt
C:\comparison\AFsp\PQevalAudio.exe input.C.wav output.aften448_spec.C.wav > output.aften448_spec.C.PEAQ.txt
C:\comparison\AFsp\PQevalAudio.exe input.L.wav output.aften448.L.wav > output.aften448.L.PEAQ.txt
C:\comparison\AFsp\PQevalAudio.exe input.C.wav output.aften448.C.wav > output.aften448.C.PEAQ.txt
C:\comparison\AFsp\PQevalAudio.exe input.L.wav output.aften224_wide.L.wav > output.aften224_wide.L.PEAQ.txt
C:\comparison\AFsp\PQevalAudio.exe input.C.wav output.aften224_wide.C.wav > output.aften224_wide.C.PEAQ.txt
C:\comparison\AFsp\PQevalAudio.exe input.L.wav output.aften224_spec.L.wav > output.aften224_spec.L.PEAQ.txt
C:\comparison\AFsp\PQevalAudio.exe input.C.wav output.aften224_spec.C.wav > output.aften224_spec.C.PEAQ.txt
C:\comparison\AFsp\PQevalAudio.exe input.L.wav output.aften224.L.wav > output.aften224.L.PEAQ.txt
C:\comparison\AFsp\PQevalAudio.exe input.C.wav output.aften224.C.wav > output.aften224.C.PEAQ.txt
C:\comparison\AFsp\PQevalAudio.exe -l 256: input.L.pad256.wav output.DD6448.L.wav > output.DD6448.L.PEAQ.txt
C:\comparison\AFsp\PQevalAudio.exe -l 256: input.C.pad256.wav output.DD6448.C.wav > output.DD6448.C.PEAQ.txt
C:\comparison\AFsp\PQevalAudio.exe -l 256: input.L.pad256.wav output.DD6224.L.wav > output.DD6224.L.PEAQ.txt
C:\comparison\AFsp\PQevalAudio.exe -l 256: input.C.pad256.wav output.DD6224.C.wav > output.DD6224.C.PEAQ.txt
C:\comparison\AFsp\PQevalAudio.exe -l 636: input.L.pad636.wav output.DD7448.L.wav > output.DD7448.L.PEAQ.txt
C:\comparison\AFsp\PQevalAudio.exe -l 636: input.C.pad636.wav output.DD7448.C.wav > output.DD7448.C.PEAQ.txt
C:\comparison\AFsp\PQevalAudio.exe -l 636: input.L.pad636.wav output.DD7224.L.wav > output.DD7224.L.PEAQ.txt
C:\comparison\AFsp\PQevalAudio.exe -l 636: input.C.pad636.wav output.DD7224.C.wav > output.DD7224.C.PEAQ.txt

Measurement SNR of different channels:
delaycut.exe -start 13.25 -out input.R.pad636.wav input.R.wav
delaycut.exe -start 13.25 -out input.SL.pad636.wav input.SL.wav
delaycut.exe -start 13.25 -out input.SR.pad636.wav input.SR.wav
delaycut.exe -start 13.25 -out input.LFE.pad636.wav input.LFE.wav
C:\comparison\eac3to\eac3to.exe output.aften448.ac3 output.aften448.all.wavs
C:\comparison\eac3to\eac3to.exe output.DD7448.ac3 output.DD7448.all.wavs
C:\comparison\AFsp\CompAudio.exe input.L.wav output.aften448.all.L.wav > output.aften448.all.L.comp.txt
C:\comparison\AFsp\CompAudio.exe input.R.wav output.aften448.all.R.wav > output.aften448.all.R.comp.txt
C:\comparison\AFsp\CompAudio.exe input.C.wav output.aften448.all.C.wav > output.aften448.all.C.comp.txt
C:\comparison\AFsp\CompAudio.exe input.SL.wav output.aften448.all.SL.wav > output.aften448.all.SL.comp.txt
C:\comparison\AFsp\CompAudio.exe input.SR.wav output.aften448.all.SR.wav > output.aften448.all.SR.comp.txt
C:\comparison\AFsp\CompAudio.exe input.LFE.wav output.aften448.all.LFE.wav > output.aften448.all.LFE.comp.txt

C:\comparison\AFsp\CompAudio.exe -l 636: input.L.pad636.wav output.DD7448.all.L.wav > output.DD7448.all.L.comp.txt
C:\comparison\AFsp\CompAudio.exe -l 636: input.R.pad636.wav output.DD7448.all.R.wav > output.DD7448.all.R.comp.txt
C:\comparison\AFsp\CompAudio.exe -l 636: input.C.pad636.wav output.DD7448.all.C.wav > output.DD7448.all.C.comp.txt
C:\comparison\AFsp\CompAudio.exe -l 636: input.SL.pad636.wav output.DD7448.all.SL.wav > output.DD7448.all.SL.comp.txt
C:\comparison\AFsp\CompAudio.exe -l 636: input.SR.pad636.wav output.DD7448.all.SR.wav > output.DD7448.all.SR.comp.txt
C:\comparison\AFsp\CompAudio.exe -l 636: input.LFE.pad636.wav output.DD7448.all.LFE.wav > output.DD7448.all.LFE.comp.txt

C:\comparison\eac3to\eac3to.exe output.DD6448.ac3 output.DD6448.all.wavs
C:\comparison\AFsp\CompAudio.exe -l 256: input.LFE.pad256.wav output.DD6448.all.LFE.wav > output.DD6448.all.LFE.comp.txt

Padding detection:
BeSplit.exe -core( -input input.L.wav -prefix input.L.cut -type wav ) -split( 0 1 )
BeSplit.exe -core( -input output.aften448.L.wav -prefix output.aften448.L.cut -type wav ) -split( 0 1 )
BeSplit.exe -core( -input output.DD6448.L.wav -prefix output.DD6448.L.cut -type wav ) -split( 0 1 )
BeSplit.exe -core( -input output.DD7448.L.wav -prefix output.DD7448.L.cut -type wav ) -split( 0 1 )
C:\comparison\AFsp\CompAudio.exe -d 0:1030 input.L.cut01.wav output.aften448.L.cut01.wav > output.aften448.L.cut01.comp.txt
C:\comparison\AFsp\CompAudio.exe -d 0:1030 input.L.cut01.wav output.DD6448.L.cut01.wav > output.DD6448.L.cut01.comp.txt
C:\comparison\AFsp\CompAudio.exe -d 0:1030 input.L.cut01.wav output.DD7448.L.cut01.wav > output.DD7448.L.cut01.comp.txt
Conclusion:
1) There is a simple method of objective audio comparison that anybody can use.
2) Aften has much worse SNR than DD. Even 224 kbps AC3 encoded with DD beats aften 448 kbps in SNR.
3) When using PEAQ metric (which takes into account some psychoacoustic effects) aften is closer to DD, but still worse.
4) If you want to have frequency cutoff similar to DD encoders, you should use w=48 for 448 kbps and w=8 for 224 kbps.
5) Aften saves high frequency average energy closer to original. But it has higher energy on low frequencies then original track.
6) Aften has better SNR on surround channels than on front. With DD it is inversely.
7) DD7 has very bad SNR on LFE channel. More tests should be done on this issue.

P.S. It was originally written in Russian, so feel free to correct spelling.

Last edited by Abradoks; 8th November 2009 at 23:52.
Abradoks is offline   Reply With Quote
Old 7th November 2009, 19:28   #1017  |  Link
yesgrey
Registered User
 
Join Date: Sep 2004
Posts: 1,295
Quote:
Originally Posted by Abradoks View Post
Objective comparison of AC3 encoders
Why have you used a lossy encoded track (DTS 1510kbps) as the source for the tests? I think it turns the results less meaningfull...
yesgrey is offline   Reply With Quote
Old 8th November 2009, 01:46   #1018  |  Link
honai
Guest
 
Posts: n/a
Actually, the results are thoroughly invalidated by that.

Also, even though the Blu-ray of At World's End contains a lossless LPCM track, it's 16 bit and therefore not suitable for this test, either.
  Reply With Quote
Old 8th November 2009, 02:37   #1019  |  Link
jruggle
Registered User
 
Join Date: Jul 2006
Posts: 276
Quote:
Originally Posted by Abradoks View Post
Difference between aften_wide and aften_spec:
- aften (wide) encodes full frequency range: -w 60
- aften (spec) encodes same frequency range as DD encoders:
448 kbps -> 20.34 kHz -> w=48
224 kbps -> 9.05 kHz -> w=8
That is actually very good to know. I don't have access to a DD encoder. I would not be opposed to changing the defaults if that's what users would prefer. What are the other cutoffs for different bit rates and channel layouts?

edit: also, could you provide some small snippets of the DD-encoded files so that I can see what encoding features and params they use to get the better PEAQ and SNR values?

Last edited by jruggle; 8th November 2009 at 02:41.
jruggle is offline   Reply With Quote
Old 8th November 2009, 18:39   #1020  |  Link
Abradoks
Registered User
 
Join Date: Mar 2008
Posts: 71
Quote:
Originally Posted by jruggle View Post
What are the other cutoffs for different bit rates and channel layouts?
Code:
3/2 (L, C, R, l, r)
Bitrate			w (48 kHz)	w (44.1 kHz)	w (32 kHz)
640, 576, 512, 448	48		56		60
384			40		44		60
320			32		36		60
256			20		24		44
224			8		12		28

2/2 (L, R, l, r); 3/1 (L, C, R, l)
Bitrate				w (48 kHz)	w (44.1 kHz)	w (32 kHz)
640, 576, 512, 448, 384, 320	48		56		60
256				32		36		60
224				24		28		48
192				12		16		32

2/1 (L, R, l); 3/0 (L, C, R)
Bitrate				w (48 kHz)	w (44.1 kHz)	w (32 kHz)
640, 576, 512, 448, 384, 320	48		56		60
256				40		44		60
224				32		36		60
192				24		28		48
160				16		20		40
128				0		4		20

2/0 (L, R)
Bitrate						w (48 kHz)	w (44.1 kHz)	w (32 kHz)
640, 576, 512, 448, 384, 320, 256, 224, 192	48		56		60
160						32		36		60
128						24		28		52
112						20		24		44
96						0		4		20

1/0 (C)
Bitrate								w (48 kHz)	w (44.1 kHz)	w (32 kHz)
640, 576, 512, 448, 384, 320, 256, 224, 192, 160, 128, 112, 96	48		56		60
80								32		36		60
64								16		20		40
56								12		16		32
Quote:
Originally Posted by jruggle View Post
also, could you provide some small snippets of the DD-encoded files so that I can see what encoding features and params they use to get the better PEAQ and SNR values?
Source wavs, 224.DD6.ac3, 448.DD6.ac3. It's one minute (00:16:00-00:17:00) from previously mentioned source. Also, I've noticed that encoding with DD takes much more time compared to aften. 224.DD6.ac3 — 15 seconds, 448.DD6.ac3 — 24 seconds. aften -b 448 takes about 3 seconds.
jruggle, you can send me any sample you want to be encoded with DD plugin.


yesgrey3, honai, are you kidding? There is no difference what source do you take if it is complex enough for current encoding. And that 1510 kbps DTS is hard enough to encode in 448 kbps AC3, as you can see from SNR. Also I was trying to make a real life comparison, so I didn't try to encode any synthetic tracks. Nevertheless you can run test by yourself on any other source. It's not hard if you have DD encoder. I don't think that you will get different results.
Abradoks is offline   Reply With Quote
Reply

Thread Tools Search this Thread
Search this Thread:

Advanced Search
Display Modes

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off

Forum Jump


All times are GMT +1. The time now is 03:53.


Powered by vBulletin® Version 3.8.9
Copyright ©2000 - 2017, vBulletin Solutions, Inc.