eac3to - audio conversion tool

[odin24]

oh maaaan


Did that work?

[nautilus7]

I got the Wall E blu-ray Disc. It supposed to have DTS-HD MA 6.1 track, but eac3to reports it as 5.1 Here's the log:

eac3to wall.e.dtshd -logdts

+ DTS-Core
- frameSize 2012
- DTS-ES +
- channelNo 5
- lfe 1
- channelDescr 5.1
- samplingRate 48000
- bitDepth 24
- bitrate 1536000
- samplesPerFrame 512
- copyHistory 1

+ DTS-HD
- fullSize 2776
- headerSize 32
- refClockCode 1/48000
- frameDurationCode 1
- activeMasks [1], [[1]]
+ Asset [0]
- fullSize 2744
- headerSize 14
- corePackets Core
- extSubStrPackets XLL
- bitResolution 24
- maxSampleRate 48000
- totalNumChannels 6
- activeSpeakers C L R Ls Rs LFE ($f)

DTS Master Audio, 5.1 channels, 24 bits, 48khz
(core: DTS-ES, 5.1 channels, 24 bits, 1536kbps, 48khz)

Both Arcsoft and Sonic decoder decode it as 5.1, so i guess it's really 5.1ch, right?

I made a sample (http://www.sendspace.com/file/a7dbxu) in case you want to have a look.

[madshi]

I got the Wall E blu-ray Disc. It supposed to have DTS-HD MA 6.1 track, but eac3to reports it as 5.1.

Both Arcsoft and Sonic decoder decode it as 5.1, so i guess it's really 5.1ch, right?

I made a sample (http://www.sendspace.com/file/a7dbxu) in case you want to have a look.
This seems to be 5.1 ES, with the back channel(s) matrixed into the surround channels. Kind of strange, haven't seen a "DTS Master Audio ES" track yet. Why didn't they go regular 6.1 or 7.1? I've no idea...

[NorthPole]

What format would the data have, you send to standard input? Generally it's a difficult thing for me to add because eac3to by design first analyzes the format of the source, and then in a 2nd step rereads the whole source file to do its work. This approach doesn't fit well to standard input where the source file can not be read twice.


How about having the ability to feed in standard wav file format. I currently have material stored in TTA format. I can decode that at the command line and pipe it into eac3to.

[flyingernst]

Did that work?

have not tried it until now...was in the college. But my PC is still busy with the mkv .... so I have to wait

[flyingernst]

Hi, here is a second project:
Want a movie in m2ts with DolbyTrueHD. It is in original splittet so I have to fix that gab thing (which I dont understand) with this method:

eac3to v2.74
command line: "C:\Program Files\Minnetonka Audio Software\SurCode DVD DTS\eac3to.exe" "K:" 1) "F:\Hancock\film.mkv"
------------------------------------------------------------------------------
M2TS, 2 video tracks, 6 audio tracks, 19 subtitle tracks, 1:42:14
1: Chapters, 16 chapters
2: h264/AVC, 1080p24 /1.001 (16:9)
3: h264/AVC, 480p24 /1.001 (20:11)
4: TrueHD/AC3, English, 5.1 channels, 48khz
(embedded: AC3, 5.1 channels, 448kbps, 48khz)
5: AC3, English, 5.1 channels, 448kbps, 48khz, dialnorm: -30dB
6: TrueHD/AC3, German, 5.1 channels, 48khz
(embedded: AC3, 5.1 channels, 448kbps, 48khz, dialnorm: -29dB)
7: AC3, Russian, 5.1 channels, 448kbps, 48khz, dialnorm: -29dB
8: AC3, Ukrainian, 5.1 channels, 448kbps, 48khz, dialnorm: -29dB
9: DTS Express, English, 2.0 channels, 24 bits, 192kbps, 48khz
10: Subtitle (PGS), English
11: Subtitle (PGS), English
12: Subtitle (PGS), German
13: Subtitle (PGS), Dutch
14: Subtitle (PGS), Danish
15: Subtitle (PGS), Estonian
16: Subtitle (PGS), Finnish
17: Subtitle (PGS), Modern Greek
18: Subtitle (PGS), Hindi
19: Subtitle (PGS), Lithuanian
20: Subtitle (PGS), Norwegian
21: Subtitle (PGS), Swedish
22: Subtitle (PGS), Turkish
23: Subtitle (PGS), English
24: Subtitle (PGS), German
25: Subtitle (PGS), Dutch
26: Subtitle (PGS), Russian
27: Subtitle (PGS), Ukrainian
28: Subtitle (PGS), English
Creating file "F:\Hancock\film - Chapters.txt"...
[v02] Extracting video track number 2...
[a04] Extracting audio track number 4...
[a05] Extracting audio track number 5...
[a06] Extracting audio track number 6...
[a07] Extracting audio track number 7...
[a08] Extracting audio track number 8...
[s10] Extracting subtitle track number 10...
[s11] Extracting subtitle track number 11...
[s12] Extracting subtitle track number 12...
[s23] Extracting subtitle track number 23...
[s21] Extracting subtitle track number 21...
[s20] Extracting subtitle track number 20...
[s26] Extracting subtitle track number 26...
[s27] Extracting subtitle track number 27...
[s14] Extracting subtitle track number 14...
[s15] Extracting subtitle track number 15...
[s19] Extracting subtitle track number 19...
[s13] Extracting subtitle track number 13...
[s25] Extracting subtitle track number 25...
[s28] Extracting subtitle track number 28...
[s22] Extracting subtitle track number 22...
[s16] Extracting subtitle track number 16...
[s17] Extracting subtitle track number 17...
[s24] Extracting subtitle track number 24...
[s18] Extracting subtitle track number 18...
[a04] Extracting TrueHD stream...
[a04] Decoding with libav/ffmpeg...
[v02] Muxing video to Matroska...
[a06] Extracting TrueHD stream...
[a06] Decoding with libav/ffmpeg...
[a04] Encoding FLAC with libFlac...
[a06] Encoding FLAC with libFlac...
[a07] Removing AC3 dialog normalization...
[a08] Removing AC3 dialog normalization...
[a05] Removing AC3 dialog normalization...
[a06] Creating file "F:\Hancock\film - 6 - TrueHD+AC3, German, 5.1 channels, 48khz.24bit.flac"...
[a04] Creating file "F:\Hancock\film - 4 - TrueHD+AC3, English, 5.1 channels, 48khz.24bit.flac"...
[a05] Creating file "F:\Hancock\film - 5 - AC3, English, 5.1 channels, 448kbps, 48khz.ac3"...
[a08] Creating file "F:\Hancock\film - 8 - AC3, Ukrainian, 5.1 channels, 448kbps, 48khz.ac3"...
[a07] Creating file "F:\Hancock\film - 7 - AC3, Russian, 5.1 channels, 448kbps, 48khz.ac3"...
[s23] Creating file "F:\Hancock\film - 23 - Subtitle (PGS), English.sup"...
[s24] Creating file "F:\Hancock\film - 24 - Subtitle (PGS), German.sup"...
[s25] Creating file "F:\Hancock\film - 25 - Subtitle (PGS), Dutch.sup"...
[libav] Lossless check failed - expected 2e, calculated 2b
[s10] Creating file "F:\Hancock\film - 10 - Subtitle (PGS), English.sup"...
[s15] Creating file "F:\Hancock\film - 15 - Subtitle (PGS), Estonian.sup"...
[s11] Creating file "F:\Hancock\film - 11 - Subtitle (PGS), English.sup"...
[s18] Creating file "F:\Hancock\film - 18 - Subtitle (PGS), Hindi.sup"...
[s22] Creating file "F:\Hancock\film - 22 - Subtitle (PGS), Turkish.sup"...
[s19] Creating file "F:\Hancock\film - 19 - Subtitle (PGS), Lithuanian.sup"...
[s13] Creating file "F:\Hancock\film - 13 - Subtitle (PGS), Dutch.sup"...
[s12] Creating file "F:\Hancock\film - 12 - Subtitle (PGS), German.sup"...
[s17] Creating file "F:\Hancock\film - 17 - Subtitle (PGS), Modern Greek.sup"...
[s21] Creating file "F:\Hancock\film - 21 - Subtitle (PGS), Swedish.sup"...
[s14] Creating file "F:\Hancock\film - 14 - Subtitle (PGS), Danish.sup"...
[s20] Creating file "F:\Hancock\film - 20 - Subtitle (PGS), Norwegian.sup"...
[s16] Creating file "F:\Hancock\film - 16 - Subtitle (PGS), Finnish.sup"...
[s27] Creating file "F:\Hancock\film - 27 - Subtitle (PGS), Ukrainian.sup"...
[libav] Lossless check failed - expected 16, calculated 33
[libav] Lossless check failed - expected 9e, calculated bd
[libav] Lossless check failed - expected 9, calculated aa
[libav] Lossless check failed - expected 69, calculated c3
[libav] Lossless check failed - expected 3, calculated 37
[libav] Lossless check failed - expected 1c, calculated 4e
[libav] Lossless check failed - expected 3e, calculated 37
[libav] Lossless check failed - expected 5a, calculated c5
[libav] Lossless check failed - expected 9f, calculated 15
[libav] Lossless check failed - expected 22, calculated 74
[libav] Lossless check failed - expected 39, calculated 3a
[libav] Lossless check failed - expected 8a, calculated 62
[libav] Lossless check failed - expected a8, calculated 86
[libav] Lossless check failed - expected e9, calculated ec
[a04] The original audio track has a constant bit depth of 16 bits.
[a04] The zero bytes were successfully removed.
[a05] Audio overlaps for 27ms at playtime 0:05:51.
[a05] Audio overlaps for 23ms at playtime 0:27:55.
[a05] Audio overlaps for 16ms at playtime 0:32:31.
[a05] Audio overlaps for 9ms at playtime 0:34:13.
[a05] Audio overlaps for 20ms at playtime 0:48:54.
[a05] Audio overlaps for 18ms at playtime 0:49:56.
[a05] Audio overlaps for 17ms at playtime 0:52:28.
[a05] Audio overlaps for 29ms at playtime 0:53:10.
[a05] Audio overlaps for 31ms at playtime 0:57:34.
[a05] Audio overlaps for 8ms at playtime 1:04:46.
[a05] Audio overlaps for 25ms at playtime 1:07:09.
[a05] Audio overlaps for 7ms at playtime 1:10:47.
[a05] Audio overlaps for 23ms at playtime 1:13:47.
[a05] Audio overlaps for 24ms at playtime 1:15:20.
[a05] Audio overlaps for 18ms at playtime 1:25:56.
[a05] The audio file was demuxed without making use of the gap/overlap information.
[a05] Please rerun the same eac3to command line. That will correct the gaps/overlaps.
[a06] The original audio track has a constant bit depth of 16 bits.
[a06] The zero bytes were successfully removed.
[a07] Audio overlaps for 27ms at playtime 0:05:51.
[a07] Audio overlaps for 23ms at playtime 0:27:55.
[a07] Audio overlaps for 16ms at playtime 0:32:31.
[a07] Audio overlaps for 9ms at playtime 0:34:13.
[a07] Audio overlaps for 20ms at playtime 0:48:54.
[a07] Audio overlaps for 18ms at playtime 0:49:56.
[a07] Audio overlaps for 17ms at playtime 0:52:28.
[a07] Audio overlaps for 29ms at playtime 0:53:10.
[a07] Audio overlaps for 31ms at playtime 0:57:34.
[a07] Audio overlaps for 8ms at playtime 1:04:46.
[a07] Audio overlaps for 25ms at playtime 1:07:09.
[a07] Audio overlaps for 7ms at playtime 1:10:47.
[a07] Audio overlaps for 23ms at playtime 1:13:47.
[a07] Audio overlaps for 24ms at playtime 1:15:20.
[a07] Audio overlaps for 18ms at playtime 1:25:56.
[a07] The audio file was demuxed without making use of the gap/overlap information.
[a07] Please rerun the same eac3to command line. That will correct the gaps/overlaps.
[a08] Audio overlaps for 27ms at playtime 0:05:51.
[a08] Audio overlaps for 23ms at playtime 0:27:55.
[a08] Audio overlaps for 16ms at playtime 0:32:31.
[a08] Audio overlaps for 9ms at playtime 0:34:13.
[a08] Audio overlaps for 20ms at playtime 0:48:54.
[a08] Audio overlaps for 18ms at playtime 0:49:56.
[a08] Audio overlaps for 17ms at playtime 0:52:28.
[a08] Audio overlaps for 29ms at playtime 0:53:10.
[a08] Audio overlaps for 31ms at playtime 0:57:34.
[a08] Audio overlaps for 8ms at playtime 1:04:46.
[a08] Audio overlaps for 25ms at playtime 1:07:09.
[a08] Audio overlaps for 7ms at playtime 1:10:47.
[a08] Audio overlaps for 23ms at playtime 1:13:47.
[a08] Audio overlaps for 24ms at playtime 1:15:20.
[a08] Audio overlaps for 18ms at playtime 1:25:56.
[a08] The audio file was demuxed without making use of the gap/overlap information.
[a08] Please rerun the same eac3to command line. That will correct the gaps/overlaps.
[s26] Creating file "F:\Hancock\film - 26 - Subtitle (PGS), Russian.sup"...
[s28] Creating file "F:\Hancock\film - 28 - Subtitle (PGS), English.sup"...
Added fps value to MKV header.
Video track 2 contains 147076 frames.
Video track 3 contains 147076 frames.
eac3to processing took 36 minutes, 20 seconds.
Done.


How can I use that methode and get out no flacs, but the TrueHD Files without GAP Problem
Thanks, Greetings,
Michael

[madshi]

How about having the ability to feed in standard wav file format. I currently have material stored in TTA format. I can decode that at the command line and pipe it into eac3to.
I'll think about that. It's not easy to implement...

Hi, here is a second project:
Want a movie in m2ts with DolbyTrueHD. It is in original splittet so I have to fix that gab thing (which I dont understand) with this method:

How can I use that methode and get out no flacs, but the TrueHD Files without GAP Problem
Thanks, Greetings,
Michael
First of all, please post logs between "[ code ]" and "[ / code ]" (without the spaces) tags. That way the logs will consume less space in your post (they'll appear in a smaller window with scroll bars).

As far as I can see from your eac3to log, the TrueHD tracks do not have any gaps/overlaps in them. eac3to only reported overlaps for the 3 AC3 tracks number 5, 7 and 8. So what you're looking for should be possible by doing this:

eac3to "K:" 1) 2: "F:\Hancock\film.mkv" 4: "F:\Hancock\Englisch.thd+ac3" 6: "F:\Hancock\Deutsch.thd+ac3"
Of course if the video MKV is already done, you can remove the "2: "F:\Hancock\film.mkv"" from the command line to save time.

Generally if there are overlaps in a TrueHD track Houston has a problem because eac3to can not properly remove overlaps from TrueHD tracks. In this specific case you're lucky cause the overlaps seem to be too small for eac3to to complain about...

[flyingernst]

well then thank you very much.
First I will have to see if TSmuxer will accept the TrueHD Files to import. Even when, I will have to remux the Movie with TSremux because of the PopCorn Hour wants it....

[Thunderbolt8]

This seems to be 5.1 ES, with the back channel(s) matrixed into the surround channels. Kind of strange, haven't seen a "DTS Master Audio ES" track yet. Why didn't they go regular 6.1 or 7.1? I've no idea...
its listed in eac3to as DTS Master Audio 5.1 with DTS-ES 5.1 core

[Jeff Flowerday]

Evan Almighty HD-DVD rip:

When I run eac3to on the rip it doesn't list the feature movie for some reason. All I get is the following:

C:\Users\FlowerdayJ\Videos\EVANALMIGHTY>eac3to .
1) WATERRISING_1.EVO+WATERRISING_2.EVO+RU_GIRAFFE.EVO+RU_ELEPHANT.EVO+RU_LION.EV
O+RU_GORILLA.EVO+RU_WOLVES.EVO+RU_EAGLE.EVO+RU_KOALA.EVO+RU_OTTER.EVO+RU_CHEETAH
.EVO+RU_SURFEIT.EVO+RU_AUSTRALIA.EVO+RU_CHIMP.EVO+RU_GRIZZLY.EVO+RU_LLAMA.EVO+RU
_PEACOCK.EVO+RU_CAMEL.EVO+RU_ORANGUTAN.EVO+RU_ARMADILLO.EVO+RU_GOAT.EVO+RU_ANACO
NDA.EVO+RU_EXITBEFORE.EVO+RU_OUTRO.EVO+RU_EXITAFTER.EVO+RU_END.EVO, 0:04:50
"Animal Round Up"
- VC-1, 1080p (16:9)
- E-AC3, 2.0, 48khz

[nautilus7]

I was playing around with the new floating point output and i wonder which offers better quality: AC3 decoded with nero (24 bit output) or AC3 decoded with libav to floating point?
What do you think? Is floating point output preferable over nero decoder? For sure the difference will be very small, but i don't have that much expensive equipment to test my self. Can anyone tell?

[Snowknight26]

What is the difference between 2 channel AC3 Surround, 2 channel AC3 Headphone and just normal 2 channel AC3?

[nautilus7]

I guess dolby surround is the old "analog" dolby surround pro logic (center and surround channels matrixed into left and right), while dolby headphne is something similar (equivalent), but specially developed for headphones.

[Jeff Flowerday]

Evan Almighty HD-DVD rip:

When I run eac3to on the rip it doesn't list the feature movie for some reason. All I get is the following:

C:\Users\FlowerdayJ\Videos\EVANALMIGHTY>eac3to .
1) WATERRISING_1.EVO+WATERRISING_2.EVO+RU_GIRAFFE.EVO+RU_ELEPHANT.EVO+RU_LION.EV
O+RU_GORILLA.EVO+RU_WOLVES.EVO+RU_EAGLE.EVO+RU_KOALA.EVO+RU_OTTER.EVO+RU_CHEETAH
.EVO+RU_SURFEIT.EVO+RU_AUSTRALIA.EVO+RU_CHIMP.EVO+RU_GRIZZLY.EVO+RU_LLAMA.EVO+RU
_PEACOCK.EVO+RU_CAMEL.EVO+RU_ORANGUTAN.EVO+RU_ARMADILLO.EVO+RU_GOAT.EVO+RU_ANACO
NDA.EVO+RU_EXITBEFORE.EVO+RU_OUTRO.EVO+RU_EXITAFTER.EVO+RU_END.EVO, 0:04:50
"Animal Round Up"
- VC-1, 1080p (16:9)
- E-AC3, 2.0, 48khz

When I go straight at the drive I get the same result.

Not sure what you use to find the information but I assume the xpl files so I've zipped them up.

http://www.flowerdayconsulting.com/Misc/vpl.zip

[jruggle]

Problem 1 (bruce-lee 1.0 ac3)
Sorry, obviously something went wrong. Here the 10mb-file:
http://rapidshare.com/files/161846249/bruce10MB.eac3


This now decodes properly with libavcodec from FFmpeg's current SVN.

Thanks madshi for contacting me directly about this file. It does use spectral extension. I had implemented the feature in the "sandbox" repository from the original Google Summer of Code project, but had not merged to FFmpeg SVN because I only had 1 sample file to test... until now!! With this sample and another (also provided by madshi) working properly, I decided to merge the changes to FFmpeg SVN.

[jruggle]

I was playing around with the new floating point output and i wonder which offers better quality: AC3 decoded with nero (24 bit output) or AC3 decoded with libav to floating point?
What do you think? Is floating point output preferable over nero decoder? For sure the difference will be very small, but i don't have that much expensive equipment to test my self. Can anyone tell?
Listening tests aside, I doubt the quality is noticeably different. The internal dynamic range for AC3 is 24-bit.

First of all, we don't know what the Nero decoder does internally. It might be completely fixed point, or maybe it uses floating point everywhere and just converts to 24-bit at the end. The lavc decoder uses 24-bit fixed point for everything up to the IMDCT and downmixing in each block. If I had a 24-bit fixed-point IMDCT implementation it would not use floating point at all. The only reason the downmixing is floating point is because it's sometimes done after the IMDCT.

[sehgal.v7]

@Madshi
Do you want sample of Iron Man Audio in TrueHD & DTS HD-MA?
I saw sometime back you needed sample of Same track in two Lossless Formats..

[madshi]

What is the difference between 2 channel AC3 Surround, 2 channel AC3 Headphone and just normal 2 channel AC3?
I guess dolby surround is the old "analog" dolby surround pro logic (center and surround channels matrixed into left and right), while dolby headphne is something similar (equivalent), but specially developed for headphones.
Yep, that's correct. Dolby Headphone I believe is 5.1 downconverted to 2.0 in such a way that it is supposed to still sound like surround when played back by a headphone...

When I go straight at the drive I get the same result.

Not sure what you use to find the information but I assume the xpl files so I've zipped them up.

http://www.flowerdayconsulting.com/Misc/vpl.zip
For whatever strange reason the "VPLST000.XPL" file only contains some extra stuff. The movie is in the "VPLST001.XPL" file. Never seen that before yet! You can "fix" the problem by deleting "VPLST000.XPL" and renaming "VPLST001.XPL" to "VPLST000.XPL", because eac3to always only checks "VPLST000.XPL".

This now decodes properly with libavcodec from FFmpeg's current SVN.
Great - thanks! :)

Listening tests aside, I doubt the quality is noticeably different. The internal dynamic range for AC3 is 24-bit.

First of all, we don't know what the Nero decoder does internally. It might be completely fixed point, or maybe it uses floating point everywhere and just converts to 24-bit at the end. The lavc decoder uses 24-bit fixed point for everything up to the IMDCT and downmixing in each block. If I had a 24-bit fixed-point IMDCT implementation it would not use floating point at all. The only reason the downmixing is floating point is because it's sometimes done after the IMDCT.
Ah, that's interesting...

@Madshi
Do you want sample of Iron Man Audio in TrueHD & DTS HD-MA?
I saw sometime back you needed sample of Same track in two Lossless Formats..
Yes, why not! Such samples never hurt...

[deathlord]

madshi

I have a question on the new ac3 ex feature. Dolby digital ex has a 6th channel built in for use in 6.1 or 7.1 setups. Can eac3to decode those tracks to 7.1?




No.


Any plans on implementing it, or do you intentionally keep these tracks at 5.1 only?

[Jeff Flowerday]

For whatever strange reason the "VPLST000.XPL" file only contains some extra stuff. The movie is in the "VPLST001.XPL" file. Never seen that before yet! You can "fix" the problem by deleting "VPLST000.XPL" and renaming "VPLST001.XPL" to "VPLST000.XPL", because eac3to always only checks "VPLST000.XPL".



Thank you! Will do.

[htpc@rischer.at]

This now decodes properly with libavcodec from FFmpeg's current SVN.

Thanks madshi for contacting me directly about this file. It does use spectral extension. I had implemented the feature in the "sandbox" repository from the original Google Summer of Code project, but had not merged to FFmpeg SVN because I only had 1 sample file to test... until now!! With this sample and another (also provided by madshi) working properly, I decided to merge the changes to FFmpeg SVN.

Great work. bruce10.eac3 was my example-track.
So, when and how can I get this changes in order to convert my track?

[Jeff Flowerday]

For whatever strange reason the "VPLST000.XPL" file only contains some extra stuff. The movie is in the "VPLST001.XPL" file. Never seen that before yet! You can "fix" the problem by deleting "VPLST000.XPL" and renaming "VPLST001.XPL" to "VPLST000.XPL", because eac3to always only checks "VPLST000.XPL".




Just an FYI, the VPLST001.XPL is formatted weird as well, eac3to wasn't showing any chapters, so I went and deleted the <ScheduledControlList> sections in both Main Movie titles and copied the <ChapterList> section from Title 1 down into Title 27 where it was missing. After doing that eac3to showed me only one Main Movie and it had chapters.

I googled the movie and some people were having trouble with the HD-DVD player freezing on chapter operations with the particular movie. This may very well explain the problems they were having.

:rolleyes:

[Boulder]

What about volume normalization, is there a chance we might see that in eac3to? Sometimes tracks that are downmixed from multichannel to stereo are rather low on volume.

[n0mag!c]

First of all it depends on the decoder. I patched libav so that it gives me the full floating point decode. But Nero and Sonic decoders internally downconvert to 24bit PCM. Now with libav eac3to by default downconverts to 24bit PCM if you ask it to create a WAV file. The reason for that is that 24bit PCM is usually plenty good enough. The main purpose of keeping things in floating point is if you want to do further processing on the data or if you want to feed the data to an encoder which wants floating point data. If that is the case, simply use the "-full" switch. Then eac3to will create floating point WAV files (if floating point data is available).
In the educational purposes could you tell please how do you perform (for example) 24 bit to 16 bit downconverting? Just throw out 8 lower bits or doing some sort of companding (applying dynamic compression to dynamic range) similar to mu-law etc?

[xkodi]

before few month i reported that:

http://forum.doom9.org/showthread.php?p=1158014#post1158014

Sonic Audio Decoder version 5.0.0.203 (from Sonic CineVision 2.5.0) doesn't decode DTS-HD extension and decodes the DTS core only and that version 4.3.0.169 (from Sonic CineVision 2.0.2) is the last one that decodes the DTS-HD extension.

now, Sonic CineVision 2.6.1 is out, it includes version 4.3.0.230 of the decoder.

Sonic has strange version numbering:

Sonic CineVision 2.0.2: Sonic Audio Decoder version 4.3.0.169
Sonic CineVision 2.5.0: Sonic Audio Decoder version 5.0.0.203
Sonic CineVision 2.6.1: Sonic Audio Decoder version 4.3.0.230

maybe, only the last 3 digits are important and they are the build number. anyway, the new version 4.3.0.230 doesn't work with the DTS-HD extension too.

btw, while testing the new version 4.3.0.230 of Sonic i found out something that looks like a bug in eac3to: the same DTS core decoded with Sonic in eac3to and GraphEdit produce different output (not byte by byte indentical), which doesn't seem right.

[odin24]

I have a DTS hi-res 7.1 track @ 2084 kb/s I am about to remux for BD playback, however my HT setup is only 5.1. Is the extra kb/s only assigned to the rear surround channels or would there be "hi res" data on the other 6 channels as well?

Basically will I be getting just the core because I have a 5.1 setup? I have an HD audio amp and it is hooked up via HDMI.

Also, when demuxing a DTS HD or DTS HR track should the output be ".dtshd/hr" and would setting it as .dts write a new dts track and leave out the HD data?

Thanks.

[nautilus7]

HD data are used for all channels.

To demux a DTS-HD track you need .dtshd or .dts or even .dtsma/.dtshr (i have never used the last though). To get the core you need the -core switch.

[odin24]

HD data are used for all channels.

To demux a DTS-HD track you need .dtshd or .dts or even .dtsma/.dtshr (i have never used the last though). To get the core you need the -core switch.

So using .dts on a DTS HD/HR would in fact demux all HD data rather than just writing a new DTS track, goo to know.

I was only concerned with the fact that my setup is 5.1, and the DTS HR track I have is 7.1... and was I going to get the extra HD data.

Thanks.

[Ryu77]

Madshi, this may have been asked before so please excuse me if that is the case...

Since Nero 9's filters are once again accessible (as opposed to Nero 8's being locked) through direct show, do you have any plans to incorporate usage of these filters into EAC3to?

I would love to be able to upgrade my version of Nero. I am still using Nero 7, simply because I want to retain full use of your fantastic application. It just seems a shame to not be able to replace an entire multimedia suite due to this.

[lithiumus]

Hey Madshi,

I was wondering if you could add an option to downsize from 8 channels to 7 channels in addition to downsizing from 8 to 6 by combining the Lsr(BL) and Rsr(BR) into Cs(BC?).

When combining this option with extracting to "wavs" would take an 8 channel 7.1 stream and downsize it to 6.1 and output 7 separate wav files which you can then manually feed into a DTS-ES encoder to create a 6.1 DTS-ES track.

What do you think?

[73ChargerFan]

Since Nero 9's filters are once again accessible...Nero 9 sounds like a dog version, according to what I've been reading. There is no BD plug-in, and the v8 plug in won't work. BD authoring is still there, as an extra $$$ plug-in.

[bmnot]

I don't have Predator. Why don't you simply retry with the latest eac3to version? There was a bug which resulted in eac3to sometimes aborting processing at the very end of movies. That was fixed in v2.74. So probably Predator should work fine now.

Works with 2.75 now. Thank you very much for the fix.

[Thunderbolt8]

Madshi, this may have been asked before so please excuse me if that is the case...

Since Nero 9's filters are once again accessible (as opposed to Nero 8's being locked) through direct show, do you have any plans to incorporate usage of these filters into EAC3to?

I would love to be able to upgrade my version of Nero. I am still using Nero 7, simply because I want to retain full use of your fantastic application. It just seems a shame to not be able to replace an entire multimedia suite due to this.
try installing both versions

[cavediver]

How do I determine which playlist contains the correct movie? I ran eac3to and got the following results: Both indicate they are the same length, but use different m2ts files.

D:\Video Ripping\eac3to>eac3to i:\dieanotherday
1) 00001.mpls, 2:12:34
[130+131+133+134+136+137+139+140+142+143+145+146+148+149+151+152+154+155+157+
158+160+161+163+164+166+167+169+170+172+173+175+176+178].m2ts
- h264/AVC, 1080p24 /1.001 (16:9)
- DTS Master Audio, English, multi-channel, 48khz
- AC3, Spanish, stereo, 48khz
- AC3, French, multi-channel, 48khz
- AC3, English, stereo, 48khz
- AC3, English, stereo, 48khz
- AC3, English, stereo, 48khz

2) 00099.mpls (angle 2), 2:12:34
[130+132+133+135+136+138+139+141+142+144+145+147+148+150+151+153+154+156+157+
159+160+162+163+165+166+168+169+171+172+174+175+177+178].m2ts
- h264/AVC, 1080p24 /1.001 (16:9)
- DTS Master Audio, English, multi-channel, 48khz
- AC3, Spanish, stereo, 48khz
- AC3, French, multi-channel, 48khz
- AC3, English, stereo, 48khz
- AC3, English, stereo, 48khz
- AC3, English, stereo, 48khz

3) 00200.mpls, 00028.m2ts, 0:51:40
- MPEG2, 480i60 /1.001 (16:9)
- AC3, English, stereo, 48khz

4) 00201.mpls, 00029.m2ts, 0:23:35
- MPEG2, 480i60 /1.001 (16:9)
- AC3, English, stereo, 48khz

5) 00202.mpls, 00030.m2ts, 0:22:39
- MPEG2, 480i60 /1.001 (16:9)
- AC3, English, stereo, 48khz

D:\Video Ripping\eac3to>

[jj666]

How do I determine which playlist contains the correct movie?

Check out some of the "different" M2TS files. It's more than likely different languages in the movie titles etc.

-jj-

[madshi]

Great work. bruce10.eac3 was my example-track.
So, when and how can I get this changes in order to convert my track?
Justin just notified me about a bug he found in the new libav AC3 decoder code. So I can't give you any date.

What about volume normalization, is there a chance we might see that in eac3to? Sometimes tracks that are downmixed from multichannel to stereo are rather low on volume.
What exactly do you mean with that? Do you mean automatic volume increase just below clipping levels?

In the educational purposes could you tell please how do you perform (for example) 24 bit to 16 bit downconverting? Just throw out 8 lower bits or doing some sort of companding (applying dynamic compression to dynamic range) similar to mu-law etc?
Any integer to integer bitdepth reduction (e.g. 24 -> 16) is performed by doing standard TPDF dithering.

btw, while testing the new version 4.3.0.230 of Sonic i found out something that looks like a bug in eac3to: the same DTS core decoded with Sonic in eac3to and GraphEdit produce different output (not byte by byte indentical), which doesn't seem right.
Is there a volume difference? You could compare output to the ArcSoft decoder to find out whether the eac3to decode or the GraphEdit decode is correct.

do you have any plans to incorporate usage of these filters into EAC3to?
Not at this time. Maybe later... AFAIK, you can install Nero 7 and Nero 9 at the same time, btw.

Any plans on implementing it
I was wondering if you could add an option to downsize from 8 channels to 7 channels in addition to downsizing from 8 to 6 by combining the Lsr(BL) and Rsr(BR) into Cs(BC?).
I'm not sure about these feature wishes. There are a lot of channel manipulation features I could add, if I knew the exact matrix or dematrix maths. Maybe I'll find a way to do this properly, maybe not. Can't say right now...

How do I determine which playlist contains the correct movie?
This movie seems to have multiple angles. This is usually used to encode different languages in a space saving way. E.g. credits in English, French, German etc. Usually angle 1 is English, but there's no guarantee for that, so you better check out the different m2ts parts yourself to find out which angle is which language (as jj666 already said).

[madshi]

I have a DTS hi-res 7.1 track @ 2084 kb/s I am about to remux for BD playback, however my HT setup is only 5.1. Is the extra kb/s only assigned to the rear surround channels or would there be "hi res" data on the other 6 channels as well?
HD data are used for all channels.
Actually it's not that simple. Everything depends on what the studio did. Here are 3 very different DTS-HD High Resolution examples:

Example 1, Basic Instinct (Blu-Ray)

+ DTS-Core
- frameSize 2012
- DTS-ES +
- channelNo 6
- lfe 1
- channelDescr 6.1
- samplingRate 48000
- bitDepth 24
- bitrate 1536000
- extAudio XCh
- samplesPerFrame 512
- copyHistory 1
+ DTS-HD
- fullSize 2040
- headerSize 28
- refClockCode 1/48000
- frameDurationCode 1
- activeMasks [1], [[1]]
+ Asset [0]
- fullSize 2012
- headerSize 10
- corePackets Core+XCh
- extSubStrPackets XBR
- bitResolution 24
- maxSampleRate 48000
- totalNumChannels 7
- activeSpeakers C L R Ls Rs LFE Cs ($1f)
DTS Hi-Res, 6.1 channels, 0:00:14, 24 bits, 3093kbps, 48khz
(core: DTS-ES, 6.1 channels, 0:00:14, 24 bits, 1536kbps, 48khz)
This has a core with 6.1 discrete channels. The DTS-HD extension contains additional bitrate for all 6.1 channels. So if you extract the core, you lose bitrate for all 6.1 channels. This example matches nautilus7's explanation.

Example 2, Herbert Grönemeyer 12 live (HD DVD)

+ DTS-Core
- frameSize 2012
- DTS-ES +
- channelNo 5
- lfe 1
- channelDescr 5.1
- samplingRate 48000
- bitDepth 24
- bitrate 1536000
- samplesPerFrame 512
- copyHistory 1
+ DTS-HD
- fullSize 1400
- headerSize 28
- refClockCode 1/48000
- frameDurationCode 1
- activeMasks [1], [[1]]
+ Asset [0]
- fullSize 1372
- headerSize 12
- corePackets Core
- extSubStrPackets XXCh+X96
- bitResolution 24
- maxSampleRate 96000
- totalNumChannels 8
- activeSpeakers C L R LFE Lsr Rsr Lss Rss ($84b)
DTS Hi-Res, 7.1 channels, 0:00:16, 24 bits, 2605kbps, 96khz
(core: DTS-ES, 5.1 channels, 0:00:16, 24 bits, 1536kbps, 48khz)
This example has a core with 5.1 discrete channels. The DTS-HD information contains a channel extension to 7.1 and a sample rate extension to 96khz. But there's absolutely no added bitrate in the DTS-HD block for the 48khz version of the 5.1 channels. So if you extract the core, all you lose is the two added channels and 96khz, but you don't lose additional bitrate for the 5.1 channels (unless you want 96khz and count the added 96khz bitrate).

Example 3, House of 1000 Corpses (Blu-Ray)

+ DTS-Core
- frameSize 2012
- DTS-ES +
- channelNo 5
- lfe 1
- channelDescr 5.1
- samplingRate 48000
- bitDepth 24
- bitrate 1536000
- samplesPerFrame 512
- copyHistory 1
+ DTS-HD
- fullSize 1740
- headerSize 32
- refClockCode 1/48000
- frameDurationCode 1
- activeMasks [1], [[1]]
+ Asset [0]
- fullSize 1708
- headerSize 13
- corePackets Core
- extSubStrPackets XBR+XXCh
- bitResolution 24
- maxSampleRate 48000
- totalNumChannels 8
- activeSpeakers C L R LFE Lsr Rsr Lss Rss ($84b)
DTS Hi-Res, 7.1 channels, 0:00:14, 24 bits, 2864kbps, 48khz
(core: DTS-ES, 5.1 channels, 0:00:14, 24 bits, 1536kbps, 48khz)
Here you can see that the core contains 5.1 discrete channels. The DTS-HD part of the track contains both a bitrate extension (XBR) and a channel extension to 7.1 (XXCh). So if you extract the core, you lose the back channels and also you lose additional bitrate for the 5.1 channels.

Basically, if you want to know what you'll lose by extract the core, you should use the undocumented "-logdts" switch and then check out which extensions are stored in the DTS-HD blocks. If you strip off XCh or XXCh extensions, you're losing additional channels. If you strip off the X96 extension, you're losing 96khz. If you strip off the XBR extension, you're losing additional bitrate for the 5.1 channels. If you strip off the XLL extension, you're losing Master Audio. XCh, XXCh and X96 extensions can be contained in either the core or in the DTS-HD blocks, while XBR and XLL extensions can only be contained in the DTS-HD blocks...

[nautilus7]

I see. Thanks

But in the last example you say:
...and also you lose additional bitdepth for the 5.1 channels.You mean bitrate, right?

[odin24]

So by looking at this example, is the only information stored in the HD extension is the extra channels (XXCh)?

Thanks for the great explanation BTW. :)

The Condemned

+ DTS-Core
- frameSize 2012
- DTS-ES -
- channelNo 5
- lfe 1
- channelDescr 5.1
- samplingRate 48000
- bitDepth 24
- bitrate 1536000
- samplesPerFrame 512
- copyHistory 1
+ DTS-HD
- fullSize 716
- headerSize 28
- refClockCode 1/48000
- frameDurationCode 1
- activeMasks [1], [[1]]
+ Asset [0]
- fullSize 688
- headerSize 11
- corePackets Core
- extSubStrPackets XXCh
- bitResolution 24
- maxSampleRate 48000
- totalNumChannels 8
- activeSpeakers C L R Ls Rs LFE Lw Rw ($40f)
DTS Hi-Res, 7.1 channels, 1:53:39, 24 bits, 2083kbps, 48khz

[nautilus7]

Yes, indeed. Only the extra channels are added by the -HD extension.

[odin24]

Great, so since my set up is 5.1 all I am getting is the core track... oh well.

Thanks for the help.

[tebasuna51]

a question regarding delaycut 1.3.0.0. when I use an .ac3 file as input then sometimes the duration of the target file is 1 ac3 frame shorter than the duration of the source file. this is shown instantly, before I even attempt to process....

Only the header of first frame are read. And the data are previsions:

We know the FrameLength (must be constant) and the FileLength then can have:

Frames = int(FileLength/FrameLength)

Each frame have 1536 samples, at 48000 Hz each frame have 32 ms then:

Duration = frames x 32 ms.

[woah!]

How do I determine which playlist contains the correct movie? I ran eac3to and got the following results: Both indicate they are the same length, but use different m2ts files.

D:\Video Ripping\eac3to>eac3to i:\dieanotherday
1) 00001.mpls, 2:12:34
[130+131+133+134+136+137+139+140+142+143+145+146+148+149+151+152+154+155+157+
158+160+161+163+164+166+167+169+170+172+173+175+176+178].m2ts
- h264/AVC, 1080p24 /1.001 (16:9)
- DTS Master Audio, English, multi-channel, 48khz
- AC3, Spanish, stereo, 48khz
- AC3, French, multi-channel, 48khz
- AC3, English, stereo, 48khz
- AC3, English, stereo, 48khz
- AC3, English, stereo, 48khz

2) 00099.mpls (angle 2), 2:12:34
[130+132+133+135+136+138+139+141+142+144+145+147+148+150+151+153+154+156+157+
159+160+162+163+165+166+168+169+171+172+174+175+177+178].m2ts
- h264/AVC, 1080p24 /1.001 (16:9)
- DTS Master Audio, English, multi-channel, 48khz
- AC3, Spanish, stereo, 48khz
- AC3, French, multi-channel, 48khz
- AC3, English, stereo, 48khz
- AC3, English, stereo, 48khz
- AC3, English, stereo, 48khz

3) 00200.mpls, 00028.m2ts, 0:51:40
- MPEG2, 480i60 /1.001 (16:9)
- AC3, English, stereo, 48khz

4) 00201.mpls, 00029.m2ts, 0:23:35
- MPEG2, 480i60 /1.001 (16:9)
- AC3, English, stereo, 48khz

5) 00202.mpls, 00030.m2ts, 0:22:39
- MPEG2, 480i60 /1.001 (16:9)
- AC3, English, stereo, 48khz

D:\Video Ripping\eac3to>

i would say 1) is the main movie as 2) is the Angle 2 version...

[madshi]

But in the last example you say:
You mean bitrate, right?
Yes, corrected it.

Great, so since my set up is 5.1 all I am getting is the core track... oh well.
It's not so bad. Look, if this wasn't a DTS-HD track, both the 5.1 channels and also the XXCh extension would be in the core. In that case the XXCh extension would eat away bitrate from the 5.1 channels. In other words: The 5.1 channels wouldn't have full 1536Kbps. Instead you would maybe have 1200Kbps or something like that for 5.1 and the rest for the additional channels. But thanks to DTS-HD the XXCh extension could be moved to the DTS-HD block, so the full bitrate of the core could be spent on 5.1. So you should have a really good sounding 5.1 core.

[Boulder]

What exactly do you mean with that? Do you mean automatic volume increase just below clipping levels?Yes, exactly that. In fact, SSRC includes this option (in 2-pass mode, of course) so maybe you could use existing code.

[madshi]

a question regarding delaycut 1.3.0.0.
I have a surprise for you:

delaycut has its own thread.

[tebasuna51]

yes, but in this case where is the missing frame? at the beginning, at the end?

You don't know until fix the full stream. You have only previsions:
Like NotFixedDelay is 0, the first header begin at first byte then a correct ac3 stream must have all frames ok, until the last truncated by a buggy video editor.
(the video has 1-2 overlapped frames each second)

The video issues isn't relevant for audio.

[madshi]

the chance of getting answer here is just higher
This is exactly how you can make me angry. I will not tolerate anyone *intentionally* misusing this thread for asking OT questions. :mad:

[Thunderbolt8]

nvm then, deleted these entries

but wheres that goddamn delaycut thread to be found? search engine is bs in that case and searching with strg+f for "delay" also didnt lead me there 20 pages in the audio section.
so at least dont be surprised questions are asked here.

heres another sample similar to that of robocop, another h264 1080i50 movie. I also have sync problems when slowing down that one, tried it with and without the gaps file. the runtime of the original video is 1:26:58, so the slowed down version should be about 90,7 minutes. with gaps file usage the length is ~1:30:05 (or similar), without gaps file 1:30:36, which also is a little off.

http://www.sendspace.com/file/fyonek

[odin24]

It's not so bad. Look, if this wasn't a DTS-HD track, both the 5.1 channels and also the XXCh extension would be in the core. In that case the XXCh extension would eat away bitrate from the 5.1 channels. In other words: The 5.1 channels wouldn't have full 1536Kbps. Instead you would maybe have 1200Kbps or something like that for 5.1 and the rest for the additional channels. But thanks to DTS-HD the XXCh extension could be moved to the DTS-HD block, so the full bitrate of the core could be spent on 5.1. So you should have a really good sounding 5.1 core.

That's a good way of looking at it, I was more dissapointed with the fact there was no HD data attatched to the core. It's no big deal though, I will enjoy it regardless.

Thanks again for the help.