eac3to - audio conversion tool [Archive] - Page 249

Bigmango

22nd August 2013, 06:41

kukushka

22nd August 2013, 13:31

Same problem here.

probably it's worth to give it a try with sonic too. but if every possible decoder decode same distorted original, it probably was there initially..

Bigmango

22nd August 2013, 16:03

While waiting to be able to post here for the first time, I posted over at avsvorums (http://www.avsforum.com/t/1482702/oblivion-blu-ray-official-avsforum-review/150#post_23626101) about a problem I noticed trying to decode the DTS HD MA sound from the Obilivion blu ray. In short, the center channel becomes distorted in loud bits. Since that posting, I have realised that specifying '-core' option to eac3to fixes the problem but of course its not "lossless", which my decode of the MA track clearly isn't either in this case.

I have uploaded a sample (https://dl.dropboxusercontent.com/u/60598588/oblivion_distort.zip) .dtshd file from 8:10-8:20 of the film which exhibits the problem. Also included with the sample are my eac3to decodes of the centre channel with and without "-core" as well as the libav decode, which also exhibits the distiortion. And of course all the logs and eac3to -test results.

I get the same results (bit-for-bit) using eac3to with the Arcsoft decoder versions 1.1.0.0, 1.1.0.7 and 1.1.0.8. I also hear the distortion when decoding the .dtshd directly in mplayer, vlc and audacity (which I think use the same code as libav anyway).

EDIT: On second thought, the distortion is there in core track as well. I don't know how I didn't hear that before.

After some more testing with the full movie, this seems to be a problem with the ffmpeg decoder.

BTW, as you are saying the tiny distortion you are getting in your sample is in the original track as well. But, starting at the same time up to 20 seconds later (not in your sample) there often is A LOT of sound clipping (and in other parts of the movie too).

I seem to be getting this with ffmpeg based decoders, and then not every time. PowerDVD plays it fine (also the FLAC).

So, as far as I can see, this is a bug in ffmpeg.

nevcairiel

22nd August 2013, 16:31

How is it a bug in ffmpeg if the original problem was the ArcSoft decoder causing the issue? :)

Bigmango

22nd August 2013, 17:00

How is it a bug in ffmpeg if the original problem was the ArcSoft decoder causing the issue? :)

I first thought Arcsoft was the issue (I played my Oblivion with FLAC backup on XBMC), but after some more testing the ffmpeg based players (tested XBMC, MPC-HC have the problem) seem to be the only ones having the playback problem (the tested XBMC which is playing back the FLAC as PCM on a hardware receiver is on a totaly different system than the tested MPC-HC, totaly different hardware and drivers - both have the problem.).

Furthermore, PowerDVD plays the eac3to+Arcsoft transcoded FLAC fine whitout the clipping that's sometimes produced by the ffmpeg players, and Oblivion is the only movie having this issue (in a media library of ~300 movies).

So, if PowerDVD plays the (Arcsoft decoded) FLAC fine but 2 different ffmpeg players on 2 different systems don't, ffmpeg is the issue.

nevcairiel

22nd August 2013, 17:16

So you're saying ffmpegs FLAC decoder is the issue, right? Not actually related to DTS directly at all?
I tried listening to the FLAC center channel samples posted above, and i couldn't really hear any obvious distortions, but maybe i just dont know what to listen for (or my player is fixed)

Sparktank

22nd August 2013, 23:11

Regarding distortion: I've been meaning to start a new thread with heavy investigation into this.
Intense testing with libav/arcsoft/sonic/etc.

I didn't observe any (for Oblivion) using Foobar2000 with the center-FLAC.
I can't remember if it was apparent in MPC-HC (FFDshow), but I don't remember there being any.
The distortion wasn't immensely apparent for me as it was for you.

Even opening in Audacity it didn't show any clipping (marked in red). It does show that it's cut off at .8 ("The vertical scale displays amplitude when showing the waveform" from Audacity Help)
http://i.imgur.com/KNnn2lO.png

Using PaulStretch (to slow down the audio by repeating samples numerous times) shows clipping when slowed down to 10x the normal rate.
http://i.imgur.com/ICQa6Od.png

Which is where you hear the perceived distortion ("perceived" as in it's not apparent to everyone).

----

This could be a result of various setups of software/hardware. (why some can hear it and some cannot)
I've disabled all "effects" that my soundcard has to offer (ie, "loudness equalization", etc).
I've even disabled any normalization or regain effects in FFDShow when playing back the FLAC in MPC-HC.
Also MPC-HC has all its internal effects disabled to play back the audio with the lest altering/manipulation.

I haven't done any testing with the elementary stream (dts-hd ma) though.
I have Arcsoft 1.0.0.9 and can do a test to compare but I don't have any detailed knowledge or software to really examine the differences between the other versions.
But Foobar2000 has a "compare two files" to detect any offset in volume, etc.

----

I observed the same thing with Tron Legacy. Although, with TL, the distortion is much more apparent than that of Oblivion.
After looking at the Oblivion samples (which played fine with barely any distortion in Foobar), I went to rip my BD of TL to test for distortion.

It seems that depending on the player it (Tron Legacy) did have severe bass distortion.
I observed distortion on different instances. Notably when playing back on MPC-HC with FFShow.

Playing back a converted WAV (5.1 (down6) or 7.1) on Foobar doesn't have any distortion but playing the WAVs on MPC-HC did however (FFDShow again).
I have not tried using LAV filters though or other.
Downmixing (either 7.1 or converted 5.1) to 2.0 (-downStereo; -7.xxdb automatically applied with 2nd pass) did playback on all media players without any distortion.
---
I'm still playing around with different setups to test with the most variables.
I'm even running through SoX for "stat" and "stats" (though that info is quite useless to me until I learn it).

But the distortion seemed very conditional.
I just need to remember to document everything I did to see where distortion was obvious (and not, in case of Foobar??)

Though I'll be creating a new thread for this with mega details.
I'm really curious about some movies having this weird distortion anomaly (fact: some hear it; some don't).

X-Men: First Class has it at times as well.
I can hear it when watching the movie on my system (but since I upgraded my BD player, it might be a different experience now).

Sorry for the off-topic details, will create a new thread.
This is highly intriguing.

madshi

22nd August 2013, 23:18

If the problem is in the ffmpeg decoder then using madFlac should work without issues?

nevcairiel

23rd August 2013, 01:36

The ffmpeg FLAC decoder is perfectly lossless, so its more likely some other combination of the playback chain.

Bigmango

23rd August 2013, 04:51

So you're saying ffmpegs FLAC decoder is the issue, right? Not actually related to DTS directly at all?

Yes, or something else in ffmpeg.

I tried listening to the FLAC center channel samples posted above, and i couldn't really hear any obvious distortions, but maybe i just dont know what to listen for (or my player is fixed)

Yes, these samples don't have the distortion. There is some kind of "stuttering" at one point , but it is also present in the original DTS track.

What I am however experiencing, is sound "clipping" (some very high and loud fast "pops" or "ticks") which happens several times starting at around 08:05 to 08:30 in the movie (and at some other places in the movie too when there's some high pitched sound).

edit: I think this is what Sparktank explains above.

Bigmango

23rd August 2013, 04:59

The ffmpeg FLAC decoder is perfectly lossless, so its more likely some other combination of the playback chain.

If this is really the case (?) then it should be something else in the ffmpeg playback chain as Openelec (XBMC) playing back as PCM on a hardware receiver and MPC-HC are both having the issue, on totally different hardware (+ Openelec is Linux and MPC-HC is on Windows).

Bigmango

23rd August 2013, 05:06

Which is where you hear the perceived distortion ("perceived" as in it's not apparent to everyone).

It comes out very lound and clear on my 7.1 home theater system (pay back with Openelec XBMC mediacenter).

On windows MPC-HC with headphones it is less noticeable, but still there. However, the strange thing is that sometimes the clipping is there, and sometimes it isn't. This needs more testing...

tebasuna51

23rd August 2013, 11:03

I make some test over the dtshd sample included. And I want post my opinion about this.

1) Decoded to wavs directly (ArcSoft 1.1.0.0 and libav) result the same output than decode the Flacs included in the sample.
For me don't exist problems related to Flac encoder/decoders.

2) Decoding the "core" with ArcSoft produce the same (more or less) output than decoding with ffmpeg/libav (or libdts with NicAudio).
Then I can't suspect problems with ffmpeg/libav decoder.

3) There are clip (see attached image) in the wave form.
This can be listened like distort sound, more or less clear depending on audio system chain.
But the clip isn't at 0dB, that suggest the problem was in the original source and isn't a decoder problem.

4) But there are a problem that I never see. The lossless output from DTS-MA decoded with ArcSoft is 3dB loud than the decoded core (ArcSoft or libav).
Not only on clip parts but over the full sample.

I checked all my DTS-MA samples and always the volume is the same decoding lossless or core.

The attached image is from my decoded wavs but is the same than flacs included in the sample.

Anakunda

23rd August 2013, 15:29

Hello, may I get some help how to properly setup eac3to for conversion 7.1 channels audio, whenever I try to convert 8-channel track (DtsHD master audio) to AAC it always produces useless 8channel digital noise. All 6-channel tracks convert nicely so the problem must have to do something with the channels layout. Maybe wrong channels mapping? Or is that encoder bug? Is Arcsoft decoder decoding the audio improperly?
Using pipe, the only way I know:

C:\Users\Anakunda\AppData\Local\Temp>benice "I:\video\(workset)\MARVELS_IRON_MAN_3_BD_BLUEBIRD\english71.dtshd" -d -55 --down16
Executing eac3to "I:\video\(workset)\MARVELS_IRON_MAN_3_BD_BLUEBIRD\english71.dtshd" stdout.wav -55ms -down16 -normalize | D:\media\AAC\qaac.exe --ignorelength --threading --verbose --no-delay --rate keep --quality 2 --tvbr 100 -o "I:\video\(workset)\MARVELS_IRON_MAN_3_BD_BLUEBIRD\english71.m4a" -

eac3to v3.27
command line: D:\media\eac3to\eac3to.exe "I:\video\(workset)\MARVELS_IRON_MAN_3_BD_BLUEBIRD\english71.dtshd" stdout.wav -55ms -normalize
------------------------------------------------------------------------------
DTS Master Audio, 7.1 channels, 24 bits, 48kHz
(core: DTS, 5.1 channels, 24 bits, 1509kbps, 48kHz)
Decoding with ArcSoft DTS Decoder...
Applying RAW/PCM delay...
Writing WAV...
Creating file "stdout.pass1.wav"...
The original audio track has a constant bit depth of 24 bits.
Starting 2nd pass...
Reading WAV...
Reducing depth from 64 to 24 bits...
Writing WAV...
Applying -0.09dB gain...
Creating file "stdout.wav"...
The processed audio track has a constant bit depth of 24 bits.
eac3to processing took 49 minutes, 59 seconds.
Done.

qaac 2.19, CoreAudioToolbox 7.9.8.3
english71.m4a
Prepend 960 samples (0.02 seconds)
Using default channel layout.
Output layout: 7.1 (C Lc Rc L R Ls Rs LFE)
Enable threading
AAC-LC Encoder, TVBR q100, Quality 96
2:10:32.839 (4.6x)
375976304/-1 samples processed in 28:09.990
Overall bitrate: 665.123kbps
7812/7812 chunks written (optimizing)
Encoding done.

tebasuna51

23rd August 2013, 21:22

Hello, may I get some help how to properly setup eac3to for conversion 7.1 channels audio, whenever I try to convert 8-channel track (DtsHD master audio) to AAC it always produces useless 8channel digital noise.
Without problems here with a simple channels test DTS-MA 7.1.

The ArcSoft (1.1.0.0) output is correct (FL FR FC LF BL BR SL SR) the qaac (2.19) Output layout: 7.1 (C Lc Rc L R Ls Rs LFE) is the default like you can see here (https://github.com/nu774/qaac/wiki/Multichannel--handling), but is played correctly in my 7.1 system with mpc-hc, and decoded correctly with NeroAacDec.

Try without --threading to see if there are problems in your system with pipe and multithread with big files.
Try also decode the DTS-MA to wav and encode the wav file.

Anakunda

23rd August 2013, 23:21

Try without --threading to see if there are problems in your system with pipe and multithread with big files.

Thanks dude, don't know what's --threading exactly for but it seems without it the output now is perfect :cool:
For some reason with 6channel audio qaac with threading support worked correctly.

carlmart

28th August 2013, 14:33

It's been a long time since I could use eac3to. I wonder what might be happening.

I have already installed both eac3to and put the GUI on the same folder. But when I load a file as I remember I did, something seems to go wrong because it doesn't accept any command any longer.

Sorry if this question is so primary.

Snowknight26

28th August 2013, 20:50

eac3to doesn't come with a GUI.

What's the question?

Sparktank

28th August 2013, 21:57

the GUI on the same folder

Which GUI? There's a couple.
Try downloading a fresh copy of the GUI and installing it in the existing folder.

Maybe you lost some dependencies (MSV? etc).

Or when was the last time you used it?
eac3to is now 3.27 and one of the GUI's had to update (for reasons I can't remember) so try downloading a newer/updated version?

carlmart

29th August 2013, 15:35

Sorry for not including more details.

GUI was < yr_eac3to_more_gui>, supposedly latest version.

eac3to is January 6, 2013 version, so it is 3.27

How do I check my "dependencies"? What's MSV?

LigH

29th August 2013, 15:41

Possibly "Microsoft Visual ...(C++ Runtime Libraries)"...

Sparktank

29th August 2013, 19:03

yes, some GUI's might need Microsoft Visual.
I'm not sure, I don't use GUI's.

But that GUI seems to have stopped development. It may not be compatible with v3.27.

Switch to something more recent and still active like BDClown or UsEac3to v1.0.0.

Clown BD: (from first post in this thread)
http://forum.slysoft.com/showthread.php?t=25818

UsEac3to v1.0.0: a different GUI for eac3to
http://forum.doom9.org/showthread.php?t=145574

All other GUI's seem to have died off in 2010.

laserfan

30th August 2013, 23:17

While waiting to be able to post here for the first time, I posted over at avsvorums (http://www.avsforum.com/t/1482702/oblivion-blu-ray-official-avsforum-review/150#post_23626101) about a problem I noticed trying to decode the DTS HD MA sound from the Obilivion blu ray. In short, the center channel becomes distorted in loud bits. Since that posting, I have realised that specifying '-core' option to eac3to fixes the problem but of course its not "lossless", which my decode of the MA track clearly isn't either in this case.

I have uploaded a sample (https://dl.dropboxusercontent.com/u/60598588/oblivion_distort.zip) .dtshd file from 8:10-8:20 of the film which exhibits the problem. Also included with the sample are my eac3to decodes of the centre channel with and without "-core" as well as the libav decode, which also exhibits the distiortion. And of course all the logs and eac3to -test results.

I get the same results (bit-for-bit) using eac3to with the Arcsoft decoder versions 1.1.0.0, 1.1.0.7 and 1.1.0.8. I also hear the distortion when decoding the .dtshd directly in mplayer, vlc and audacity (which I think use the same code as libav anyway).

EDIT: On second thought, the distortion is there in core track as well. I don't know how I didn't hear that before.
I just racked this one up and my Arcsoft TMT 3 which uses (I believe) dtsdecoderdll.dll 1.1.0.7 has serious problems with the DTS-MA audio. But on my setup it's not as you describe, but rather a problem with the background music and afaict not the center channel. In any case it is a problem throughout.

Using instead my Media Player Classic HC, version 1.6.8.7417, playback of the soundtrack is perfect. I dunno how to determine what decoding method is being used in MPC-HC.

An interesting problem; I have never seen anything like it before, so I wonder what is different about this disc.

EDIT: Bah, I played with this some more today, and Arcsoft TMT 3 (mine's 3.0.1.185) does indeed play this disc fine--I don't know what the magic is though. I had looked at it a half dozen times, and fiddled with settings, but today I fiddled with settings again, both the Audio Sound Card/Output Mode settings and the Audio Effect mode, and it just started sounding correct. I can't seem to "go back" i.e. change settings back and get the bad audio again. Sorry for the bandwidth--I don't know what the problem might have been before.

Sparktank

30th August 2013, 23:43

dtsdecoderdll.dll 1.1.0.7

Have you tried 0 or 8? (or even 9)

While playing back in MPC-HC you can go the Menu Bar > Play > Filters and see what is being used to decode (most likely FFDShow; core audio for DTS-HD MA).

The previous comments from others largely suggests the process chain has a huge effect on the perceivable distortion yet it does seem to be on the disc itself.
Looking at other threads about similar problems, it seems that hardware also has different ways of experiencing the clipping/distortion.

If you read the thread about Tron: Legacy, you'll see endless posts about software VS hardware and even the probability of getting just a bad disc from a "bad batch".

Although, I've yet to see any hard evidence supporting any such claims of "bad batches" where the audio is malformed.
In most cases of "bad batches", the disc itself deteriorates and makes the entire disc unreadable.

Or just the transfer itself is poorly mastered (in the case of the UK Boxset for David Lynch where many customers complained and received a newly transferred boxset free of charge upon producing a receipt within a constrained time limit).

Tron Legacy and X-Men First Class exhibit similar clipping/distortion, from what I've seen.

A most recent extraction of a sample from Tron: Legacy (using Arcsoft versions 0, 8 and 9) indicate the clipping/distortion is on the movie itself, and while others experience it, I don't believe it to be secluded to a single disc or a small batch.

laserfan

31st August 2013, 13:12

Have you tried 0 or 8? (or even 9)
No. I do have TMT5 and 6 also but they are not installed atm. Can I just swap-in to TMT3 the newer dlls?

To be clear, while I've posted in the eac3to thread, I have not attempted (yet anyway) to convert in any way the DTS-MA soundtrack. I posted merely to say that 1.) Oblivion ripped w/o errors, so whatever problems the disc might exhibit are not due to disc defects, and 2.) whatever problems one might have with the sound are due to the player/playback software, cuz MPC-HC plays it perfectly.

I've yet to play the disc on any of my set-top players, but fully expect the sound to be perfect.

When I buy a disc, I tend to try ripping it onto my PC in order to confirm it is not defective (and before committing 2+ hours to it in my HT).

IanD

1st September 2013, 04:03

Is there an option in eac3to to combine a TrueHD soundtrack and AC3 soundtrack into a .thd+ac3 combination?

I have a situation where a .thd+ac3 soundtrack was accidentally demuxed into its separate components and TSmuxer does not seem to want to mux a standalone TrueHD track.

At the moment, eac3to decodes the TrueHD track to AC3 and uses that to create a .thd+ac3, but I would prefer to use the original AC3.

tebasuna51

1st September 2013, 13:31

@IanD
I think is not possible now.

Stereodude

7th September 2013, 18:49

Are there any known issues with ArcSoft DTS Decoder - version 1.1.0.0 when handling DTS Master Audio, English, 6.1 channels, 24 bits, 48kHz (core: DTS-ES, 6.1 channels, 24 bits, 1509kbps, 48kHz) with v3.27 of eac3to?

I tried a -downDpl -mixlfe and -downDpl and got a stereo wave file in both cases with a ~6dB channel imbalance between L and R and the file doesn't sound anything like what I hear when playing the disc in ArcSoft TotalMedia Theatre 5 having it do a 2 channel mixdown.

DPLII mixdown w/ LFE:
http://imageshack.us/a/img822/985/oe80.png

DPLII mixdown no LFE:
http://imageshack.us/a/img835/360/a5nm.png

I extracted each track to a separate .wav using .wavs and eac3to is correct that the Back Center channel is empty (as shown in the log files below). However, there isn't any left/right volume imbalance in the remaining 6 individual wave files, so I'm unsure why the 2 channel mixdown is off. :confused:

Individual channels - click for full size:
http://imageshack.us/a/img28/3706/og7p.th.png (http://imageshack.us/a/img28/3706/og7p.png)

"LOGS":
C:\HDTV Tools\eac3to>eac3to.exe i: 1) 3: e:\TempBD\CSH\DTSMA_w_LFE.wav -downDpl -mixLFE
M2TS, 1 video track, 2 audio tracks, 1:35:56, 24p /1.001
1: Chapters, 20 chapters
2: h264/AVC, 1080p24 /1.001 (16:9)
3: DTS Master Audio, English, 6.1 channels, 24 bits, 48kHz
(core: DTS-ES, 6.1 channels, 24 bits, 1509kbps, 48kHz)
4: RAW/PCM, English, 2.0 channels, 24 bits, 48kHz
a03 Extracting audio track number 3...
a03 Remapping channels...
a03 Decoding with ArcSoft DTS Decoder...
a03 Downmixing multi channel audio to stereo...
a03 Reducing depth from 64 to 24 bits...
a03 Writing WAV...
a03 Clipping detected, a 2nd pass will be necessary.
a04 Creating file "e:\TempBD\CSH\stereo.wav"...
a03 Creating file "e:\TempBD\CSH\DTSMA_w_LFE.wav"...
a03 Original audio track, L+R+C+LFE+BL+BR: constant bit depth of 24 bits.
a03 Original audio track, BC: no audio data.
a03 The processed audio track has a constant bit depth of 24 bits.
a03 Starting 2nd pass...
a03 Extracting audio track number 3...
a03 Remapping channels...
a03 Decoding with ArcSoft DTS Decoder...
a03 Downmixing multi channel audio to stereo...
a03 Reducing depth from 64 to 24 bits...
a03 Writing WAV...
a03 Applying -14.42dB gain...
a03 Creating file "e:\TempBD\CSH\DTSMA_w_LFE.wav"...
a03 The processed audio track has a constant bit depth of 24 bits.
Video track 2 contains 138004 frames.
eac3to processing took 25 minutes, 5 seconds.
Done.

C:\HDTV Tools\eac3to>eac3to.exe i: 1) 3: e:\TempBD\CSH\DTSMA.wav -downDpl
M2TS, 1 video track, 2 audio tracks, 1:35:56, 24p /1.001
1: Chapters, 20 chapters
2: h264/AVC, 1080p24 /1.001 (16:9)
3: DTS Master Audio, English, 6.1 channels, 24 bits, 48kHz
(core: DTS-ES, 6.1 channels, 24 bits, 1509kbps, 48kHz)
4: RAW/PCM, English, 2.0 channels, 24 bits, 48kHz
a03 Extracting audio track number 3...
a03 Remapping channels...
a03 Decoding with ArcSoft DTS Decoder...
a03 Downmixing multi channel audio to stereo...
a03 Reducing depth from 64 to 24 bits...
a03 Writing WAV...
a03 Clipping detected, a 2nd pass will be necessary.
a03 Creating file "e:\TempBD\CSH\DTSMA.wav"...
a03 Original audio track, L+R+C+LFE+BL+BR: constant bit depth of 24 bits.
a03 Original audio track, BC: no audio data.
a03 The processed audio track has a constant bit depth of 24 bits.
a03 Starting 2nd pass...
a03 Extracting audio track number 3...
a03 Remapping channels...
a03 Decoding with ArcSoft DTS Decoder...
a03 Downmixing multi channel audio to stereo...
a03 Reducing depth from 64 to 24 bits...
a03 Writing WAV...
a03 Applying -9.71dB gain...
a03 Creating file "e:\TempBD\CSH\DTSMA.wav"...
a03 The processed audio track has a constant bit depth of 24 bits.
Video track 2 contains 138004 frames.
eac3to processing took 25 minutes, 7 seconds.
Done.

DoctorM

7th September 2013, 19:59

Each version of ArcSoft has it's own issues:
I tested 5 different versions of dtsdecoderdll.dll along with LAV Audio, here are results.

Versions tested: 1.1.0.0, .1, .5, .7, .8.

Samples tested:
1) Lossy DTS 2.0, 2.1, 5.1, 6.0, 6.1 configs, 16/48, 24/48 and 24/96 bitdepth/sample rate;
2) DTS-MA 1.0, 2.0, 2.1, 5.1, 6.1, 7.1 configs including 7.1 "strange setup" config (eac3to term), 16/48, 24/48, 24/96 variants.
My results differ from results that some people posted in eac3to thread, I guess there may be some eac3to specific problems.

lossy DTS
1) 1.1.0.0 and 1.1.0.1 always decode lossy DTS as 24 bit, and their decoding results differ from each other.
2) 1.1.0.5 and up decode in proper bitdepth.
3) Versions 1.1.0.5 and up decode lossy DTS identically.
4) 1.1.0.1 and 1.1.0.5 decode 24 bit lossy DTS identically.
5) Decoding of 16 bit lossy DTS was changed from 1.1.0.0 to 1.1.0.1 and from 1.1.0.1 to 1.1.0.5.
6) 1.1.0.7 and 1.1.0.8 decode 6.0 without back center channel.

The conclusion for lossy: I assume the best version for lossy DTS is 1.1.0.5. It decodes all configurations I've tested, in proper bitdepth, and decoding algorithm didn't change since this version (except 6.0 bug in .7 and .8).

DTS-MA
All versions output identical streams except mono case. Mono decoding was broken in 1.1.0.5 and partially fixed in 1.1.0.8 (outputs as stereo). Note that MA 7.1 16/48 "strange setup" is decoded as 24/48 by all versions.

The conclusion for lossless: the best versions are 1.1.0.0 and 1.1.0.1.

Of course, to be sure that decoding is done properly some "reference proper" decoder is needed. I don't have it.

EDIT: Forgot to mention about 2.1 bug. None of the versions can decode LFE channel in 2.1 files.

Stereodude

7th September 2013, 19:59

Frankly, I'm not sure what's going on... I took a closer look at the channels that came out when I decoded with .wavs. The supposed LFE channel is full range and sounds like a back center channel.

As a sanity check I played the Blu-Ray back from my HTPC through my Pioneer Elite receiver which supports HDMI bitstreaming. It sees a 6.1 DTS-HD MA audio track and there is content in all 6.1 channels, no empty channels.

eac3to tells me I have the ArcSoft DTS decoder (1.1.0.0)

C:\HDTV Tools\eac3to>eac3to.exe -test
eac3to (v3.27) is up to date
Nero Audio Decoder (Nero 7) works fine
ArcSoft DTS Decoder (1.1.0.0) works fine

Stereodude

7th September 2013, 20:19

Each version of ArcSoft has it's own issues:How are you determining the revision? When I right click on ASAudioHD.ax bring up the properties and click on the version tab I see numbers like

1.3.2.87
1.3.3.127
1.3.3.283

eac3to tells me "ArcSoft DTS Decoder (1.1.0.0) works fine" for all 3 of those versions. :confused:

DoctorM

7th September 2013, 20:21

Check the version of dtsdecoderdll.dll. "Eac3to -test" should report the correct version as well.

Stereodude

7th September 2013, 20:31

Check the version of dtsdecoderdll.dll. "Eac3to -test" should report the correct version as well.Oh, okay. I was checking the wrong file.

I between my various computers I have 1.1.0.0 and 1.1.0.8. I apparently have two different 1.1.0.8 versions though since the file sizes are different and they are not binary identical.

Stereodude

7th September 2013, 21:05

Do I just need to change out the .dll to change decoder versions? I'm getting 7 identical channels of corrupted / garbled audio when trying the 1.1.0.8 dll's. :confused:

Since I can't get 1.1.0.8 to work, I'll just report the results I got from 1.1.0.0

With just the DTS-ES, 6.1 channels, 24 bits, 1509kbps, 48kHz core you don't get the back center channel. If you decode the full DTS-HD MA, the channel that was LFE channel (with the core) turns into a full range channel. It seems to be the back center, but I'm can't confirm if it's been correct decoded or not. However, the LFE channel vanishes.

tebasuna51

7th September 2013, 21:15

...
I tried a -downDpl -mixlfe and -downDpl and got a stereo wave file in both cases with a ~6dB channel imbalance between L and R and the file doesn't sound anything like what I hear when playing the disc in ArcSoft TotalMedia Theatre 5 having it do a 2 channel mixdown.

Is not a problem with ArcSoft version, try adding -phaseShift parameter to -downDpl ( -mixlfe not recommended).

Stereodude

7th September 2013, 21:29

Is not a problem with ArcSoft version, try adding -phaseShift parameter to -downDpl ( -mixlfe not recommended).Well, ArcSoft / eac3to can't even correctly decode the 6.1 DTS-HD MA audio so I certainly wouldn't exactly expect it to downmix it correctly.

dansrfe

7th September 2013, 23:16

Is vorbis encoding/decoding support within eac3to in the foreseeable future?

Sparktank

8th September 2013, 04:57

Well, ArcSoft / eac3to can't even correctly decode the 6.1 DTS-HD MA audio so I certainly wouldn't exactly expect it to downmix it correctly.

Have you even tried his suggestion to use -phaseshift?
(omitting -mixlfe)

If you want to keep the LFE channel, try making a 2.1 (LFE separate channel not mixed into FL+FR channels) audio file instead.

tebasuna51

8th September 2013, 11:29

Well, ArcSoft / eac3to can't even correctly decode the 6.1 DTS-HD MA audio so I certainly wouldn't exactly expect it to downmix it correctly.

I don't have problems decoding 6.1 with ArcSoft 1.1.0.0 like you have.

Try with this test channels: http://www.sendspace.com/file/s04pgo

tebasuna51

8th September 2013, 11:43

Is vorbis encoding/decoding support within eac3to in the foreseeable future?

To encode you can use the pipe option:

eac3to input stdout.wav | OggEnc2 -q 3 --ignorelength -o %_.ogg -

To decode we need than madshi add the option, libav can decode vorbis.

Stereodude

8th September 2013, 13:03

Have you even tried his suggestion to use -phaseshift?
(omitting -mixlfe)

If you want to keep the LFE channel, try making a 2.1 (LFE separate channel not mixed into FL+FR channels) audio file instead.No, I haven't tried it yet. I'm still working to get it to decode all the channels instead of leaving some of them empty. Once it can decode all the channels I'll worry about the downmix.

I don't have problems decoding 6.1 with ArcSoft 1.1.0.0 like you have.

Try with this test channels: http://www.sendspace.com/file/s04pgo
That test file decodes fine.

Try this one:
http://www.sendspace.com/file/vw8tb3

tebasuna51

8th September 2013, 17:48

Try this one:
http://www.sendspace.com/file/vw8tb3

You are right, decoding this sample the BC channel is empty and the LFE channel are full of high frequencies.

Now the question is if there are a decoder than can decode correctly this sample. Try other ArcSoft versions or Sonic.

If don't found any maybe we can conclude than this dtshd was wrong encoded.

Decoding the core with Foobar2000 or NicDtsSource the LFE channel is low volume but with high frequencies also. I think something is wrong in this sample.

Stereodude

8th September 2013, 18:50

You are right, decoding this sample the BC channel is empty and the LFE channel are full of high frequencies.

Now the question is if there are a decoder than can decode correctly this sample. Try other ArcSoft versions or Sonic.

If don't found any maybe we can conclude than this dtshd was wrong encoded.

Decoding the core with Foobar2000 or NicDtsSource the LFE channel is low volume but with high frequencies also. I think something is wrong in this sample.Did you try the core with eac3to? When i did, the BC channel is empty but the LFE seemed okay.

The sample may be unusual, but it does decode fine on my receiver via HDMI. It shows DTS Master Audio with 6 channels + LFE. When I bypass all bass redirection and processing I get a full range signal in all 6 channel (none are empty), plus a solid LFE channel.

As I recall there were several variants of DTS-ES. One has the 6th channel matrixed into the two rear channels. The other had a discrete 6th channel. I haven't played the core via SPDIF to a receiver yet to see which this is.

I couldn't get 1.1.0.8 working on my system and don't have other versions. I'm not sure if I have Sonic.

tebasuna51

8th September 2013, 19:16

Did you try the core with eac3to? When i did, the BC channel is empty but the LFE seemed okay.
Yes, I decode the core with ArcSoft and libav and both output a BC empty and a LFE like Foobar2000 or NicDtsSource, low volume but with high frequencies, not normal.

The sample may be unusual, but it does decode fine on my receiver via HDMI. It shows DTS Master Audio with 6 channels + LFE. When I bypass all bass redirection and processing I get a full range signal in all 6 channel (none are empty), plus a solid LFE channel.
Do you have a 6.1 speaker system?

If you have 5.1 or 7.1 maybe the receiver make some mix to side/back channels.

You can't know if LFE have high frequencies, only low frequencies was send to subwoofer speaker.

As I recall there were several variants of DTS-ES. One has the 6th channel matrixed into the two rear channels. The other had a discrete 6th channel. I haven't played the core via SPDIF to a receiver yet to see which this is.
This sample is signaled like 6.1 discrete (Extended Coding Flag set to 1)

Nico8583

8th September 2013, 19:22

Hi,
I have a problem when I try to extract video streams from "MONSTERS INC 3D".
I have this message one time for each video stream : [v02] The video bitstream framerate field doesn't seem to match the timestamps. <WARNING>
And several warnings like that : [v02] Video has a gap of 14 frames at playtime 0:00:00. <WARNING>
On playlist info, it is indicated "48p /1.001" but streams are 24p/1001.
I have also tried with the new TsMuxer and no problems.
Anyone can help me ?
Thanks !

Stereodude

8th September 2013, 22:56

Yes, I decode the core with ArcSoft and libav and both output a BC empty and a LFE like Foobar2000 or NicDtsSource, low volume but with high frequencies, not normal.I went back and double checked... Libav and ArcSoft give me a real LFE channel with the core, not a full range channel at a low volume.

Here's an mp3 of each (I increased the volume by 10dB).

Libav (core):
http://www.sendspace.com/file/xnr773

ArcSoft v1.1.0.0 (core):
http://www.sendspace.com/file/6kxkst

Do you have a 6.1 speaker system?

If you have 5.1 or 7.1 maybe the receiver make some mix to side/back channels.No, I have a 7.1 speaker system. The receiver mirrors the BC into both of the rear speakers. It does not mix the sides into the rears. I used Pure Stream Direct when testing which disables any channel redirection, bass management, or processing of the audio when I tested.

This sample is signaled like 6.1 discrete (Extended Coding Flag set to 1)ArcSoft TMT 5 says it's ES Discrete when output via SPDIF. Apparently my Pioneer Elite receivers don't display whether it's Discrete or Matrix. My old Sony ES unit did.

Edit: Sonic 4.3.0.169 does the same thing as ArcSoft 1.1.0.0. The core leaves the BC channel empty, but decodes the LFE. The full DTS-MA HD stream seems to put the BC into the LFE channel.

I wonder if the file is incorrectly flagged as discrete but is actually matrix.