eac3to - audio conversion tool

[Frogger13]

I want to UP this Issue as I'm getting the same since I moved from version 3.22 to 3.24. Happens especially when under heavy load (e.g. multiple encodings at the same time) but also spuriously when encoding one file at a time

Originally Posted by Boulder

I've been having a problem with eac3to and NeroAACEnc with later versions of eac3to. After encoding the decoded WAV, the log says that NeroAACEnc seems to be stuck (but I've no clue as to why) and it starts to encode the file again. I don't remember at which point the problem started to occur but it hasn't been there for long.

eac3to v3.24
command line: c:\utils\eac3to\eac3to.exe "e:\temp\dvd-rip\star trek tng\4x17 T81 3_2ch 384Kbps DELAY 0ms.ac3" "f:\temp\captures\tng_4x17_audio.m4a" -quality=0.42 -normalize -down2
------------------------------------------------------------------------------
AC3, 5.1 channels, 0:43:41, 384kbps, 48kHz, dialnorm: -27dB
Disabling DRC for Nero (E-)AC3 decoding...
Removing AC3 dialog normalization...
Decoding with DirectShow (Nero Audio Decoder 2)...
DirectShow reports 5.1 channels, 24 bits, 48kHz
Downmixing multi channel audio to stereo...
Writing WAV...
Creating file "f:\temp\captures\tng_4x17_audio.m4a.pass1.wav"...
Starting 2nd pass...
Reading WAV...
Reducing depth from 64 to 32 bits...
Encoding AAC <0.42> with NeroAacEnc...
Applying 3,6dB gain...
The original audio track has a constant bit depth of 64 bits.
The processed audio track has a constant bit depth of 32 bits.
The Nero AAC encoder seems to be stuck...  <ERROR>
[NeroAacEnc] Processed 0 seconds...
[NeroAacEnc] Processed 1 seconds...
[NeroAacEnc] Processed 2 seconds...
[NeroAacEnc] Processed 3 seconds...
[NeroAacEnc] Processed 4 seconds...
[NeroAacEnc] Processed 5 seconds...
[NeroAacEnc] Processed 6 seconds...
[NeroAacEnc] Processed 7 seconds...
[NeroAacEnc] Processed 8 seconds...
[NeroAacEnc] Processed 9 seconds...
[NeroAacEnc] Processed 10 seconds...
[NeroAacEnc] Processed 11 seconds...
[NeroAacEnc] Processed 12 seconds...
[NeroAacEnc] Processed 13 seconds...
[NeroAacEnc] Processed 14 seconds...
[NeroAacEnc] Processed 15 seconds...
[NeroAacEnc] Processed 16 seconds...
[NeroAacEnc] Processed 17 seconds...
[NeroAacEnc] Processed 18 seconds...
[NeroAacEnc] Processed 19 seconds...
[NeroAacEnc] Processed 20 seconds...
[NeroAacEnc] Processed 21 seconds...
.
.
.
.
[NeroAacEnc] Processed 2531 seconds...
[NeroAacEnc] Processed 2532 seconds...
[NeroAacEnc] Processed 2533 seconds...
[NeroAacEnc] Processed 2534 se
Aborted at file position 2012946500.  <ERROR>

[Thunderbolt8]

whats the status on the framerate recognition thing in case of movies which consist of more than 1 .m2ts file? is it considered a problem if that value is estimated wrong at the stage of parsing? got a case here in which both .m2ts files have the same framerate, but the complete playlist is nevertheless estimated wrongly.

edit: for the same movie I got problems with the subtitles of those combined .m2ts files. its ok with another program, but this problem does not occur at all when just using the subs from the main .m2ts file alone.

[mastrandrea]

I'm trying to slow down a 4 channel ac3, but eac3to return this error:

The AC3 encoder received a non-supported data format (float, 4, 64, -).
Aborted at file position 262144.

Am I doing somethig wrong?

Here's the mediainfo log for this particular track:

Audio
Format                           : AC-3
Format/Info                      : Audio Coding 3
Mode extension                   : CM (complete main)
Duration                         : 1h 29mn
Bit rate mode                    : Constant
Bit rate                         : 384 Kbps
Channel(s)                       : 4 channels
Channel positions                : Front: L C R, Side: C
Sampling rate                    : 48.0 KHz
Bit depth                        : 16 bits
Stream size                      : 245 MiB (100%)

Thanks in advance!

[Snowknight26]

Originally Posted by mastrandrea

Am I doing somethig wrong?

Apart from not searching beforehand, no.

http://forum.doom9.org/showthread.ph...46#post1412746

[tebasuna51]

Use eac3to and a 'pipe' to Aften.exe

eac3to input stdout.wav -slowdown | Aften -b 384 -readtoeof 1 - output.ac3

[Thunderbolt8]

when using -slowdown, to which fps rate does eac3to actually slow down, to 23.976 or precisely to 24/1.001? and when having normal source blu-rays with 23.976 audio and video, is it 24/1.001 here as well or just 23.976?

(basically Im asking when trying to combine track from different releases, if there could be a very slight sync problem in cases there are difference between source and slowed tracks due to that)

and in that respect I got a 7.1 blu-ray track here:

eac3to v3.24
command line: eac3to H:\express.flac H:\expressslowed.flac -24.000 -slowdown
------------------------------------------------------------------------------
FLAC, 7.1 channels, 1:42:14, 16 bits, 977kbps, 48kHz
Decoding FLAC...
Changing FPS from 24.000 to 23.976...
Reducing depth from 64 to 24 bits...
Encoding FLAC with libFlac...
Creating file "H:\expressslowed.flac"...
Clipping detected, a 2nd pass will be necessary.  <WARNING>
The original audio track has a constant bit depth of 16 bits.
The processed audio track has a constant bit depth of 24 bits.
Starting 2nd pass...
Decoding FLAC...
Changing FPS from 24.000 to 23.976...
Reducing depth from 64 to 24 bits...
Encoding FLAC with libFlac...
Applying -0,38dB gain...
Creating file "H:\expressslowed.flac"...
The processed audio track has a constant bit depth of 24 bits.
eac3to processing took 50 minutes, 35 seconds.
Done.

doesnt clipping usually only occur with hdtv caps? and has that anything to do with those -0,5dB gain? or why else must gain have been applied (the 7.1 source was a DTS-HD MA track with strange setup btw.)

regarding that 7.1 strange setup, I only get that note in combination with the audio of the complete blu-ray structure. when I demux that track, that information is gone. so is some kind of information now lost regarding channel setup when I want to transform that demuxed track into another audio format, would audio channels or the converted content be different compared to when using the source audio within its blu-ray structure?
if so, is it maybe to fix this with a kind of -switch which assumes the source uses such a kind of setup or maybe adding that information to the demuxed dtsma track header?

[dansrfe]

Does eac3to do timestreching or just a simple slowdown or speedup neglecting pitch etc.?

[TinTime]

Originally Posted by Thunderbolt8

doesnt clipping usually only occur with hdtv caps?

Clipping can occur when re-sampling, as is happening here.

[tebasuna51]

Originally Posted by Thunderbolt8

when using -slowdown, to which fps rate does eac3to actually slow down, to 23.976 or precisely to 24/1.001?

24/1.001

and when having normal source blu-rays with 23.976 audio and video, is it 24/1.001 here as well or just 23.976?

24/1.001

doesnt clipping usually only occur with hdtv caps?

Can occur with any lossy conversion, like here changing audio duration.

and has that anything to do with those -0,5dB gain?

Is only a info (that I request to madshi).

[Thunderbolt8]

Originally Posted by tebasuna51

24/1.001
Can occur with any lossy conversion, like here changing audio duration.

I used it on a flac file, so after a slowdown (or speedup) the audio quality, aside from the lost or gained pitch, is not lossless any more?

[TinTime]

Originally Posted by Thunderbolt8

I used it on a flac file, so after a slowdown (or speedup) the audio quality, aside from the lost or gained pitch, is not lossless any more?

No, it's been re-sampled - not (usually) a lossless procedure. The only way to keep it lossless in this case would be to play the original track slower and have a DAC that handles a (48000/1.001)kHz sample rate.

[Thunderbolt8]

has anyone already made experiences with that low volume problem of those dtsma strange setup 7.1 channels? is the volume still evenly distributed over the channels like normal, just lower now, so that cranking up the volume solves the problem? are there any other problems aside the volume thing which could come along when trying to transform to 7.1 flac? (only have a 5.1 system, so I cannot really test whether all the channels are correct)

[xkodi]

Originally Posted by Thunderbolt8

has anyone already made experiences with that low volume problem of those dtsma strange setup 7.1 channels? ...

Sonic DTS decoder correctly and bit-perfectly decodes DTS-HD MA 7.1 with "strange setup" to 5.1 channels.

so, it's your choice: get bit-perfect and correct 5.1 channels from your DTS-HD MA 7.1 with "strange setup" using Sonic DTS decoder or get 7.1 channels with using Arcsoft decoder, but all of those 7.1 channels are not-bit-perfectly decoded and since at least to me it's not clear exactly what processing of the audio data Arcsoft decoder is doing in case of "strange setup" we can even assume they are wrong.

so, at least in my opinion the safest way to go is using Sonic DTS decoder in such case of "strange setup" DTS-HD MA 7.1.

for the sake of completeness i can elaborate a little more with listing the facts:

1. "normal" (not "strange setup") DTS-HD MA 7.1 has the following channel layout:



which in the way how Microsoft (and thus eac3to) named the channels is the same as:



or DTS channel names mapped to Microsoft channel names are as follows:

DTS channel name <---> Microsoft channel name

L <---> L
R <---> R
C <---> C
LFE <---> LFE
Lsr <---> BL
Rsr <---> BR
Lss <---> SL
Rss <---> SR

2. "strange setup" DTS-HD MA 7.1 has the following channel layout:



or using the way how Microsoft (and thus eac3to) named the channels then DTS channel names mapped to Microsoft channel names are as follows:

L <---> L
R <---> R
C <---> C
LFE <---> LFE
Lsr <---> BL
Rsr <---> BR
Ls <---> i don't know how it's called in Microsoft terms
Rs <---> i don't know how it's called in Microsoft terms

3. DTS-HD MA 5.1 has the following channel layout:



which is the same as "strange setup" DTS-HD MA 7.1 layout with missing "Lsr" (or "BL" in MS terms) and "Rsr" (or "BL" in MS terms)

4. with using Sonic DTS decoder for "normal" (not "strange setup") DTS-HD MA 7.1 you get the following 5.1 channels:

L <---> L
R <---> R
C <---> C
LFE <---> LFE
Lsr <---> not decoded
Rsr <---> not decoded
Lss <---> SL
Rss <---> SR

which even that are bit-perfectly decoded are not the correct 5.1 channel configuration according to 3., because you're missing "Ls" and "Rs" channels that are used in DTS-HD MA 5.1 channel configuration as it's shown in 3., i.e. you can't convert correctly "normal" (not "strange setup") DTS-HD MA 7.1 to 5.1 channels with just using Sonic DTS decoder and actually you don't even need that, because Arcsoft gives 7.1 channels bit-perfectly decoded in case of "normal" (not "strange setup") DTS-HD MA 7.1.

5. with using Sonic DTS decoder for "strange setup" DTS-HD MA 7.1 you get the following 5.1 channels:

L <---> L
R <---> R
C <---> C
LFE <---> LFE
Lsr <---> not decoded
Rsr <---> not decoded
Ls <---> decoded, but i don't know how it's called in Microsoft terms
Rs <---> decoded, but i don't know how it's called in Microsoft terms

or you have 5.1 channels bit-perfectly decoded and also they are:

L, R, C, LFE, Ls, Rs

which is the correct configuration for DTS-HD MA 5.1 channels according to 3.

so, in very short: DTS-HD MA 7.1 with "strange setup" can be decoded bit-perfectly to correct 5.1 channels and i'm not sure why "eac3to" doesn't inform the user about it and even do that by default, because at least in my opinion it's much proper way than using Arcsoft. also, because of those facts it's correct to say that you can think in case of DTS-HD MA 7.1 with "strange setup" that you actually have DTS-HD MA 5.1, because that is what information you can extract correctly and bit-perfectly from it.

[nautilus7]

Thanks, xkodi. So you can use sonic to get (correct) 5.1 ch and arcsoft to get the (not correct) extra 2 ch and combine them in an audio editor to reach 7.1 ch.

[Midzuki]

Originally Posted by xkodi

Sonic DTS decoder correctly and bit-perfectly decodes DTS-HD MA 7.1 with "strange setup" to 5.1 channels.

4 sharing that info.

...
or get 7.1 channels with using Arcsoft decoder, but all of those 7.1 channels are not-bit-perfectly decoded and since at least to me it's not clear exactly what processing of the audio data Arcsoft decoder is doing in case of "strange setup" we can even assume they are wrong.

FWIW, and for the time being, I have confirmed that ArcSoft doesn't decode stereo lossless-DTS correctly.

Ls <---> i don't know how it's called in Microsoft terms
Rs <---> i don't know how it's called in Microsoft terms

Ls == Back-Left
Rs == Back-Right

Besides, Lsr and Rsr "do not exist" in the Wave-Format-Extensible "universe"
( and if they existed, they would be called
"Back Left of Center" && "Back Right of Center" ).

[Thunderbolt8]

is it really bit-perfect, meaning that the information from those 2 other channels is then added somewhere to the front or side channels of 5.1? or is that information from those 2 channels dropped (because you said "not decoded", which would not be bit-perfect then compared to the source information)?

and in case its added to the 5.1 channels, is there any way to measure how good it would sound compared to the same sound recorded and distributed to 5.1 only in the first place (so how accurate is the distribution of those 2 non decoded channel to the 5.1 channels)?

[xkodi]

Originally Posted by nautilus7

So you can use sonic to get (correct) 5.1 ch and arcsoft to get the (not correct) extra 2 ch and combine them in an audio editor to reach 7.1 ch.

yes, if you want 7.1 channels that's one way to lower the error, but still you won't get completely correct 7.1 channels as in case if you go for just 5.1 channels.

one other possibility to lower error maybe is to use "Ls" and "Rs" to calculate from them new "Lsr" and "Rsr", because according to:



they seem on very close positions to each other and "tebasuna51" is very good in making such calculation. however, it's just an idea, because i don't know if new "Lsr" and "Rsr" made in such way will be better than what Arcsoft is decoding for "Lsr" and "Rsr".

Originally Posted by Midzuki

Ls == Back-Left
Rs == Back-Right

Besides, Lsr and Rsr "do not exist" in the Wave-Format-Extensible "universe"
( and if they existed, they would be called
"Back Left of Center" && "Back Right of Center" ).

thanks, it's very confusing, because it seems to me in 5.1 channels configuration:

DTS Ls == MS BL
DTS Rs == MS BR

but in 7.1 channel configuration:

DTS Lsr == MS BL
DTS Rsr == MS BR

or maybe i'm wrong again - at least that's what i can conclude from the picture in my initial post - i took all of those pictures from DTS and MS.

Originally Posted by Thunderbolt8

is it really bit-perfect, meaning that the information from those 2 other channels is then added somewhere to the front or side channels of 5.1? or is that information from those 2 channels dropped (because you said "not decoded", which would not be bit-perfect then compared to the source information)?

i'm not sure that i understand what you're asking, but let me try to explain it again in different way:

* "strange setup" DTS-HD MA 7.1 decoded with Arcsoft: all 7.1 channels are decoded, but not any of those 7.1 channels is decoded bit-perfectly, which means not lossless and thus it's not correct, because DTS-HD MA is lossless. how wrong Arcsoft decodes them is not clear at least to my knowledge, because it's not clear what kind of processing of the audio data Arcsoft does in the case of "strange setup" DTS-HD MA 7.1.

* "strange setup" DTS-HD MA 7.1 decoded with Sonic: only 5.1 channels are decoded from the initial 7.1 channels, but those 5.1 channels are decoded:

- in bit-perfectly way, i.e. lossless
- as correct set of channels for 5.1 configuration, because what you get of decoding with Sonic is: (L, R, C, LFE, Lsr, Rsr, Ls, Rs) are reduced to (L, R, C, LFE, Ls, Rs) and so there is no any loss of information for 5.1 channel setup

or as i tried to summarize it my initial post "strange setup" DTS-HD MA 7.1 decoded with Sonic is way to convert that to 5.1 channel configuration without any loss of audio data for 5.1 channel setup. so, if you have movie with "strange setup" DTS-HD MA 7.1 better think of it the audio is DTS-HD MA 5.1, because there is no way to get more of it on computer.

i don't if this way of explaining is more clear or not.

Originally Posted by Thunderbolt8

and in case its added to the 5.1 channels, is there any way to measure how good it would sound compared to the same sound recorded and distributed to 5.1 only in the first place (so how accurate is the distribution of those 2 non decoded channel to the 5.1 channels)?

if you're talking about what "nautilus7" there is no way to know, because we don't know what Arcsoft is doing wrong, that's why the safest way is just get 5.1 which are at least bit-perfect/lossless and correct set of channels for 5.1 - for example even if assume we can get all channels of "strange setup" DTS-HD MA 7.1 decoded bit-perfectly then on 5.1 channel system "Lsr" and "Rsr" will be stripped and that's clear form DTS channel layout in 2. and 3. in my initial post.

[Thunderbolt8]

so its basically like theres 2 kind of channel setup information stored for the same audio content, one for 7.1 strange setup and one for normal 5.1, and sonic chooses the 5.1 setup then?

my 2nd question basically was whether that XX channel information was added arbitrarily to those 5.1 channels. but if theres another mask available which also takes all the information into consideration from the same source sound, only this time distribution is 5.1 instead of 7.1, then I guess it should sound as if that source sound was recorded to 5.1 channels only in the first place.

can I use the same procedure also for normal 7.1 dtsma tracks, those without strange setup? or is there nothing like 5.1 channel distribution information available for sonic then?

btw. ive had 24-bit 7.1 tracks which then got transformed to 16-bit 5.1 with sonic, is it still lossless?

Originally Posted by Midzuki

FWIW, and for the time being, I have confirmed that ArcSoft doesn't decode stereo lossless-DTS correctly.

hm where? does at least sonic do it correctly?

[Midzuki]

Just for the record, these are the only "7.1 DTS channel layouts" which are compatible with the Wave-Format-Extensible definitions:

L, R, C, LFE, Ls, Rs, Lw, Rw

L, R, C, LFE, Ls, Rs, Cs, Ch
(Cs == Back Center, Ch == Top Front Center)

L, R, C, LFE, Ls, Rs, Cs, Oh
(Oh == Top Center)

L, R, C, LFE, Ls, Rs, Lh, Rh
(Lh == Top Front Left, Rh == Top Front Right)

Argh!

[Anacletus]

Originally Posted by Frogger13

I want to UP this Issue as I'm getting the same since I moved from version 3.22 to 3.24. Happens especially when under heavy load (e.g. multiple encodings at the same time) but also spuriously when encoding one file at a time

I'm also experiencing this problem lately on many of my audio encodes, sources varies from flac to dts/dtshd.. neroaacenc abort the encoding after a while.. any clues?