eac3to - audio conversion tool

[ACrowley]

Bad news about Nero 8

Nero 8 Filters/Decoder wont work anymore outside Showtime!

For a few weeks Madshi had Contact with a Nero Programer. The Guy told him that newer Nero Versions will not allow using any Decoder/Filter Outside Showtime!
Obviously ist Fact now with Nero 8

My HDDVD/BR Plugin works correct in Nero 8 Showtime. TrueHD,DD+ Plaback works. But not outside Showtime/Graphedit(recode.ex) or eac3to anymore .What a Crap!

So when when you want decode your Eac3/TrueHD Tracks ,dont install Nero 8 and stay with Nero7 until we have a Solution

[maya]

Bad news about Nero

Nero Filters/Decoder wont work anymore outside Showtime!

For a few weeks Madshi had Contact with a Nero Programer. The Guy told him that newer Nero Versions will not allow using any Decoder/Filter Outside Showtime!
Obviously ist Fact now with Nero 8

My HDDVD/BR Plugin works correct in Showtime. TrueHD,DD+ Plaback works. But not outside Showtime/Graphedit(recode.ex) or eac3to anymore .What a Crap!

So when when you want decode your Eac3/TrueHD Tracks ,dont install Nero 8 and stay with Nero7 until we have a Solution

Thanks for letting me know, I just updated my Nero 7 Ulra edition, but I will make sure I stay away from Nero 8.

[madshi]

All these files will be VC-1 since they are my HD DVD rips. But what tools do I need to get this to work, and how do I use them if not self-explanatory? I have no experience with command line prompts by the way, but if I knew what the prompts were supposed to be then I could certainly input them. Thanks in advance.
Install the Haali Media Splitter package. The tool "gdsmux" (shipping with that package) can directly remux EVO into MKV. Once you have your MKV, you can use the tool "mkvtoolnix" to add audio tracks to the MKV. Maybe you can already do that with gdsmux, I've not used that yet. mkvtoolnix has a nice GUI, though. However, muxing the EVO file to MKV should be done with gdsmux and not with mkvtoolnix.

[maya]

Install the Haali Media Splitter package. The tool "gdsmux" (shipping with that package) can directly remux EVO into MKV. Once you have your MKV, you can use the tool "mkvtoolnix" to add audio tracks to the MKV. Maybe you can already do that with gdsmux, I've not used that yet. mkvtoolnix has a nice GUI, though. However, muxing the EVO file to MKV should be done with gdsmux and not with mkvtoolnix.

Thanks, I will give this a shot and report back

[ACrowley]

Install the Haali Media Splitter package. The tool "gdsmux" (shipping with that package) can directly remux EVO into MKV. Once you have your MKV, you can use the tool "mkvtoolnix" to add audio tracks to the MKV. Maybe you can already do that with gdsmux, I've not used that yet. mkvtoolnix has a nice GUI, though. However, muxing the EVO file to MKV should be done with gdsmux and not with mkvtoolnix.

yes, but its important to remux the mkv from gdsmux with mkvmerge and 23.976 Timecode
Because remuxed VC1/Evo output mkv from gdsmux has 29.97 FPS (as like as HaaliMediaSplitter on VC1 what is wrong)!
You must remux VC1/Evo in any case with mkvmerge and 23.976 Timecode
So a VC1/Evo Remux "must" be done with gdsmux and mkvmerge.

[madshi]

yes, but its important to remux the mkv from gdsmux with mkvmerge and 23.976 Timecode
Because remuxed VC1/Evo output mkv from gdsmux has 29.97 FPS (as like as HaaliMediaSplitter on VC1 what is wrong)!
You must remux VC1/Evo in any case with mkvmerge and 23.976 Timecode
So a VC1/Evo Remux "must" be done with gdsmux and mkvmerge.
Yep, that's true, forgot about that one.

[TheSof]

yes, but its important to remux the mkv from gdsmux with mkvmerge and 23.976 Timecode
Because remuxed VC1/Evo output mkv from gdsmux has 29.97 FPS (as like as HaaliMediaSplitter on VC1 what is wrong)!
You must remux VC1/Evo in any case with mkvmerge and 23.976 Timecode
So a VC1/Evo Remux "must" be done with gdsmux and mkvmerge.

That's worked fine for VC1 (reclock reports 23.976) but with AVC I just can't get reclock to get the FPS. There's no stutter so I suppose it matters not.

[madshi]

That's worked fine for VC1 (reclock reports 23.976) but with AVC I just can't get reclock to get the FPS. There's no stutter so I suppose it matters not.
I can't even get AVC properly remuxed into MKV. gdsmux stalls for me. Demuxing and remuxing works but audio sync is totally lost that way for me. (All tested with Equilibrium). So currently I have my h264 HD DVD movies still in EVO files. But I think we're a bit out of topic now...

[TheSof]

I can't even get AVC properly remuxed into MKV. gdsmux stalls for me. Demuxing and remuxing works but audio sync is totally lost that way for me. (All tested with Equilibrium). So currently I have my h264 HD DVD movies still in EVO files. But I think we're a bit out of topic now...

Also sorry for going OT, but I had same problems with Equilibrium, and gave up. A real nightmare disc. However, had no problems with Blue Ray AVC "The Prestige", as long as I used TSRemux to extract the avc1 and then use mkvextract. All fine except no framerate in reclock.

It's a good job most HDDVDs are vc1 (shivers at the thought of Equilibrium).

Back on topic, can someone please confirm that using eac3to is the only way to bypass DRC with the nero filter? And that by using the nero filter in a directshow player (mpc, zoomplayer, theatretek etc.) will use DRC?

[madshi]

Back on topic, can someone please confirm that using eac3to is the only way to bypass DRC with the nero filter? And that by using the nero filter in a directshow player (mpc, zoomplayer, theatretek etc.) will use DRC?
Confirmed. Well, if there's enough interest, I could probably add a tweak to switch off DRC for use in 3rd party media players. This tweak would only work for E-AC3, though. It would *NOT* work for conventional AC3.

[TheSof]

Confirmed. Well, if there's enough interest, I could probably add a tweak to switch off DRC for use in 3rd party media players. This tweak would only work for E-AC3, though. It would *NOT* work for conventional AC3.

I would be very interested in such a tweak.

I'm now going down the analog out route and using nero in TT or MPC would be ideal.

[ACrowley]

I would be very interested in such a tweak.
I'm now going down the analog out route and using nero in TT or MPC would be ideal.

this DRC Tweak is only eac3to internal and not possible for Realtime Playback. You can only Remove Dialnorm from a DD+ File

For DShow Playback of DD+:
-only in Nero Showtime it self you can switch of DRC
-Nero Decoder in a Dshow Player like mpc/WMp/Zooomplayer applies DRC+DialNorm
-SonicAudioDecoder applies DRC+DolbyGain,Sonic System Settings have no Effect. in eac3to and Dshow Players

Conclusion:
-DD+ Playback with a Dshow Player like mpc isnt possible without DRC/Dialnorm
-To decode DD+ to wave/etc eac3to+ Nero is the one and only Way to avoid DRC/Dialonorm
-TrueHD Decoding is still with DialNorm

But ffmpeg EAC3 is a chance for realtime Playback without DRC/Dialnorm. Hopefully a proper ffdshow Build is coming soon

[tvjunky]

Does the AC3 file you get from eac3to play with the correct length? If so, it's very unprobable that eac3to is at guilt here. Maybe the aac 5.1 encoder generally doesn't like tracks longer than 124 minutes? I've no idea. Btw, what you're doing is not really optimal. You're decoding E-AC3 which is fine. Then you're reencoding to a lossy codec (AC3). Which is fine if AC3 is your final aim. But if you want to end up with another format you should skip the AC3 step to keep audio quality as high as possible. You should probably ask eac3to to give you a wav file and then convert that wav file to aac by using BeSweet.

Thanks for your reply, madshi (and for this great tool!)

I have tested it again on underworld yesterday. The wav-file and the ac3-file eac3to created worked fine in all players. But converting these files to aac still doesn't work with besweet. Yesterday i tried it with neroaacenc.exe and neroaacenc_sse2. Always the same result -> audio length is 124 minutes. Still confused about this problem ...

[madshi]

this DRC Tweak is only eac3to internal and not possible for Realtime Playback.
Please reread the part where TheSof quoted me! ;) In short: I could make the DRC tweak work for realtime playback, if there's enough interest.

[madshi]

I have tested it again on underworld yesterday. The wav-file and the ac3-file eac3to created worked fine in all players. But converting these files to aac still doesn't work with besweet. Yesterday i tried it with neroaacenc.exe and neroaacenc_sse2. Always the same result -> audio length is 124 minutes. Still confused about this problem ...
This sounds like a problem with the aac encoder or BeHappy to me. I don't think eac3to is at guilt. So that means there's nothing I can do to help you. Please ask for help in the BeHappy or aac thread.

[tebasuna51]

This sounds like a problem with the aac decoder or BeHappy to me. I don't think eac3to is at guilt. So that means there's nothing I can do to help you. Please ask for help in the BeHappy or aac thread.

I think tvjunky use BeSweet, not BeHappy.

With BeHappy you can open the ac3 with NicAc3Source() or the big_wav with RaWavSource() (http://forum.doom9.org/showthread.php?p=1037550#post1037550). And after encode with NeroAacEnc.

If you have the big_wav (>4GB) don't need BeHappy because can use a direct command line.

But, always remember use the parameter -ignorelength with NeroAacEnc (or -readtoeof 1 with Aften) when the wav file is > 4GB.

[madshi]

I think tvjunky use BeSweet, not BeHappy.
Ah yes, sorry.

[tvjunky]

THANKS, tebasuna51! "-ignorelength" was the missing link!!

[Nikos]

In short: I could make the DRC tweak work for realtime playback, if there's enough interest.

Yes, i want realtime playback without DRC and Dial. norm.

[ACrowley]

Please reread the part where TheSof quoted me! ;) In short: I could make the DRC tweak work for realtime playback, if there's enough interest.

Ah, sry....nice!

[VempX]

how to encode .pcm file to flac?
I use TsRemux demux a audio track from m2ts.I name it test.pcm.but when I convert the .pcm file to flac by eac3to I get a error flac.
I demux the pcm by dsmux,then I get a correct pcm.I could play it by foobar2000.But eac3to can't convert it.
who can help me?

[madshi]

how to encode .pcm file to flac?
I use TsRemux demux a audio track from m2ts.I name it test.pcm.but when I convert the .pcm file to flac by eac3to I get a error flac.
What do you mean with "error flac"? I need more details. Please give me the full eac3to + flac output text and a description of what is wrong. TsRemux + eac3to should work fine.

[VempX]

I uploaded the .pcm file and encoded .flac file.

you can download it in here

http://www.yousendit.com/transfer.php?action=download&ufid=470C8DE25BDEDC4E

this is log
=================
E:\>"F:\eac3to\eac3to.exe" "E:\test.pcm" "E:\test.flac" -24 -big
RAW, 5.1 channels, 0:00:06, 24 bits, 48khz
Remapping channels. Please wait...
Converting the raw file to flac. Please wait...

Done.
=============================

[yonta]

F:\>eac3to\eac3to.exe Innocence.LPCM16.8ch.pcm innocence.eac3to.ac3 -8 -16 -down6 -640
RAW, 7.1 channels, 0:01:00, 16 bits, 48khz
Converting the raw file to wav. Please wait...
Converting the wav file to ac3. Please wait...

Aften: A/52 audio encoder
Version 0.0.8
(c) 2006-2007 Justin Ruggles, Prakash Punnoor, et al.

input format: RAW Signed 16-bit little-endian 48000 Hz stereo
output format: 48000 Hz stereo (2/0)

SIMD usage: MMX SSE SSE2 SSE3
Threads: 2

progress: 100% | q: 306.6 | bw: 60.0 | bitrate: 640.0 kbps

Done.

innocence.eac3to.ac3 is just noise and 4 times bigger than the correct 5.1ch file.


tried converting the eac3to-generated wav file to ac3 using aften.exe.
F:\>eac3to\aften\aften.exe Innocence.LPCM16.8ch.wav innocence.aften.ac3 -b 640

Aften: A/52 audio encoder
Version 0.0.8
(c) 2006-2007 Justin Ruggles, Prakash Punnoor, et al.

input format: WAVE Signed 16-bit little-endian 48000 Hz 5.1-channel
output format: 48000 Hz 3/2 + LFE

SIMD usage: MMX SSE SSE2 SSE3
Threads: 2

progress: 100% | q: 255.5 | bw: 57.9 | bitrate: 640.0 kbps

innocence.aften.ac3 seems OK

converting to the other formats than ac3 works correctly.

[madshi]

I uploaded the .pcm file and encoded .flac file.
That's one strange PCM file. Where is it from? From a Blu-Ray? Can you please demux it with xport and post here what xport reports over the LPCM attributes? It would be helpful if you could upload a sample of the m2ts file. From a quick check it could a 32bit mono PCM file. Never seen such a Blu-Ray PCM track yet!!

[madshi]

F:\>eac3to\eac3to.exe Innocence.LPCM16.8ch.pcm innocence.eac3to.ac3 -8 -16 -down6 -640
RAW, 7.1 channels, 0:01:00, 16 bits, 48khz
Converting the raw file to wav. Please wait...
Converting the wav file to ac3. Please wait...

Aften: A/52 audio encoder
Version 0.0.8
Could you please try the Aften build that is shipping with eac3to? Does that work? Also, which eac3to version are you using?

input format: RAW Signed 16-bit little-endian 48000 Hz stereo
Yeah, the "stereo" is the problem here.

Can you please rerun the same process and then press "Ctrl-C" while Aften is running? That should leave you with the wav file that eac3to was feeding to Aften. Can you please check the properties of the wav file? Is it properly flagged as a 6 channel wav file? Or is it incorrectly flagged as stereo?

[VempX]

er........I demux the pcm from a m2ts file by TsRemux.
This is the mpa(5.1 LPCM) which demuxed by xport.
http://www.yousendit.com/transfer.php?action=download&ufid=5E1BDD3C0DAFE50F

[yonta]

Could you please try the Aften build that is shipping with eac3to? Does that work? Also, which eac3to version are you using?

I'm using v1.14 and using the aften.exe that comes with eac3to:
F:\>eac3to\eac3to.exe f:\Innocence.LPCM16.8ch.pcm e:\innocence.eac3to.aften.from.eac3to.ac3 -16 -down6 -640 -8
RAW, 7.1 channels, 0:01:00, 16 bits, 48khz
Converting the raw file to wav. Please wait...
Converting the wav file to ac3. Please wait...

Aften: A/52 audio encoder
Version SVN
(c) 2006-2007 Justin Ruggles, Prakash Punnoor, et al.

invalid RIFF id in wav header
invalid wav file: -b
"aften" reported error code "1". A valid *.wav file was created sucessfully, though.


Can you please rerun the same process and then press "Ctrl-C" while Aften is running? That should leave you with the wav file that eac3to was feeding to Aften. Can you please check the properties of the wav file? Is it properly flagged as a 6 channel wav file? Or is it incorrectly flagged as stereo?

mediainfo says;
Audio #0
Codec : PCM
Codec/Family : PCM
Codec/Info : Microsoft PCM
Bit rate : 4608 Kbps
Channel(s) : 6 channels
Sampling rate : 48 KHz
Resolution : 16 bits

This wav file plays OK.
I tried converting this wav file to ac3 directly using aften.exe and it worked.(second command line in the first post)
That's why this problem is weird.

[Rectal Prolapse]

In short: I could make the DRC tweak work for realtime playback, if there's enough interest.

madshi, that would be wonderful!

[Creator1]

Bad news about Nero 8

Nero 8 Filters/Decoder wont work anymore outside Showtime!

For a few weeks Madshi had Contact with a Nero Programer. The Guy told him that newer Nero Versions will not allow using any Decoder/Filter Outside Showtime!
Obviously ist Fact now with Nero 8

My HDDVD/BR Plugin works correct in Nero 8 Showtime. TrueHD,DD+ Plaback works. But not outside Showtime/Graphedit(recode.ex) or eac3to anymore .What a Crap!

So when when you want decode your Eac3/TrueHD Tracks ,dont install Nero 8 and stay with Nero7 until we have a Solution

Thanks for the tip!

A trick for people who really want Nero 8... you can install Nero 7 in a virtual PC image and upgrade to Nero 8 on your real Windows :)

[madshi]

er........I demux the pcm from a m2ts file by TsRemux.
This is the mpa(5.1 LPCM) which demuxed by xport.
http://www.yousendit.com/transfer.php?action=download&ufid=5E1BDD3C0DAFE50F
Sorry, I wasn't clear. When you run "xport" it will tell you how many channels and which bitdepth etc the LPCM file has. Can you please post here what xport is posting to the screen?

[madshi]

invalid RIFF id in wav header
invalid wav file: -b
"aften" reported error code "1". A valid *.wav file was created sucessfully, though.
Can you please redownload eac3to just to make sure that the files are not corrupt? This sounds extremely strange to me. I mean Aften doesn't complain when you start it manually - but it does complain when being started by eac3to? That's weird.

Also, would it be possible for me to get a small sample of that problematic source file? Please before sending the sample check if the problem can be reproduced with the sample. Thanks!

[madshi]

I've a goody for you guys. Check out the last post of the delaycut thread:

http://forum.doom9.org/showthread.php?t=71545

I've updated delaycut to v1.3.0.0 with full E-AC3 support!!!

[yonta]

Can you please redownload eac3to just to make sure that the files are not corrupt? This sounds extremely strange to me. I mean Aften doesn't complain when you start it manually - but it does complain when being started by eac3to? That's weird.

Also, would it be possible for me to get a small sample of that problematic source file? Please before sending the sample check if the problem can be reproduced with the sample. Thanks!

redownloaded and binary-compared and the files are bit-identical.
I tested several times and this problem is reproduced everytime.
sample here (http://www.sendspace.com/file/pnfj69).

[VempX]

Sorry, I wasn't clear. When you run "xport" it will tell you how many channels and which bitdepth etc the LPCM file has. Can you please post here what xport is posting to the screen?

This is report of xport
=================================
D:\[DVDISO] AIR BDBOX VOL.1\BDMV\STREAM>xport -h 00003.m2ts 1 3 4
xport Transport Stream Demuxer 0.98
program = 1, video channel = 3, audio channel = 4
Program Number = 0 (0x0000), Program Map PID = 31 (0x001f)
Program Number = 1 (0x0001), Program Map PID = 256 (0x0100)
program descriptor = 0x05, 0x04, 0x48, 0x44, 0x4d, 0x56
program descriptor = 0x88, 0x04, 0x0f, 0xff, 0xff, 0xfe
ES descriptor for stream type 0x02 = 0x05, 0x08, 0x48, 0x44, 0x4d, 0x56, 0xff, 0
x02, 0x44, 0x3f
ES descriptor for stream type 0x81 = 0x05, 0x04, 0x41, 0x43, 0x2d, 0x33
ES descriptor for stream type 0x81 = 0x81, 0x04, 0x08, 0x3c, 0x0e, 0x00
ES descriptor for stream type 0x80 = 0x05, 0x08, 0x48, 0x44, 0x4d, 0x56, 0xff, 0
x80, 0x31, 0x7f
ES descriptor for stream type 0x81 = 0x05, 0x04, 0x41, 0x43, 0x2d, 0x33
ES descriptor for stream type 0x81 = 0x81, 0x04, 0x08, 0x3e, 0x04, 0x00
Audio PID = 4355 <0x1103>, type = 0x80
ES descriptor for stream type 0x80 = 0x05, 0x08, 0x48, 0x44, 0x4d, 0x56, 0xff, 0
x80, 0x61, 0x7f
LPCM Audio Mode = 3/2+lfe
LPCM Audio Bits/sample = 16
LPCM Audio Sample Rate = 48000
ts rate = unspecified
============================

[ACrowley]

I've a goody for you guys. Check out the last post of the delaycut thread:

http://forum.doom9.org/showthread.php?t=71545

I've updated delaycut to v1.3.0.0 with full E-AC3 support!!!

WOW THX..your my man :)

1. I can only load a EAC3 via drag&drop. The Open Browser recognize no ec3/eac3/ddp extension
2. i run 2 unclean EAC3 Track trough delaycut with "fix". One fixed EAC3 wirks perfect after fixing.
But the other fixed EAC3 cant be used aynmore, Nero and Sonic Decoder aborts Plaback

[madshi]

But the other fixed EAC3 cant be used aynmore, Nero and Sonic Decoder aborts Plaback
Try the different fix methods (silence, skip, ...). Do they all fail? How does the delaycut log look like? But this problem might be better handles in the delaycut thread?

[ACrowley]

Try the different fix methods (silence, skip, ...). Do they all fail? How does the delaycut log look like? But this problem might be better handles in the delaycut thread?

yes...will going into delaycut thread for it

As i say via PN.

After fix the problematic eac3 works fine in eac3to ,but Playback with Nero or SonicDecoder not..
I will test another fix Method
But can skip/silence cause async Problems when a few frames are filled with silence for exmaple ?

[madshi]

After fix the problematic eac3 works fine in eac3to ,but Playback with Nero or SonicDecoder not..
I will test another fix Method
But can skip/silence cause async Problems when a few frames are filled with silence for exmaple ?
"fix" is bad IMHO. "skip" has the biggest danger of asnyc problems. "silence" should be similar to "fix" in terms of audio sync. Even "fix" can result in audio sync problems.

[ACrowley]

"fix" is bad IMHO. "skip" has the biggest danger of asnyc problems. "silence" should be similar to "fix" in terms of audio sync. Even "fix" can result in audio sync problems.

Ok THX

The problematic EAC3 5.1 640Kbps is from Troy HDDVD.
It works with "Silence" :)

The Reencoded ac3 from "fix" Output has a loud cracking Noise, while the rencode from "silence" Output is clean.
I will better use "Silence" from now on for eac3 and ac3

[Jedi_Vader20]

Hiya guys.

I recently imported the HDDVD for Nine Inch Nails' Beside You In Time.

I ripped the audio from the disc, intending to convert it for use on my PSP. However, when I use eac3to to convert the EVO to a WAV, I get a file that looks the right size, but is only 9m35s long. Converting to FLAC/AC3/DTS gets me the right file size.

Am I doing something wrong?

[madshi]

Hiya guys.

I recently imported the HDDVD for Nine Inch Nails' Beside You In Time.

I ripped the audio from the disc, intending to convert it for use on my PSP. However, when I use eac3to to convert the EVO to a WAV, I get a file that looks the right size, but is only 9m35s long. Converting to FLAC/AC3/DTS gets me the right file size.

Am I doing something wrong?
WAV files cannot really handle files > 4GB. Some programs even stumble with WAV files over 2GB. Other programs can be configured to try to work around the WAV problem and they work alright with such big files.

[ACrowley]

Hiya guys.

I recently imported the HDDVD for Nine Inch Nails' Beside You In Time.

I ripped the audio from the disc, intending to convert it for use on my PSP. However, when I use eac3to to convert the EVO to a WAV, I get a file that looks the right size, but is only 9m35s long. Converting to FLAC/AC3/DTS gets me the right file size.

Am I doing something wrong?

On huge interleaved waves (+4gb )the Runtime is often wrong.
But the mono waves and encoded Output from it is fine, no Problem

Do you need AC3 for your PSPS ?

[tebasuna51]

The Reencoded ac3 from "fix" Output has a loud cracking Noise, while the rencode from "silence" Output is clean.
I will better use "Silence" from now on for eac3 and ac3

The "Silence" method is always the most sure (works always, maintain the sync) but can also have clicks when play.

The "Fix" method work fixing the CRC and leaving errors in data, most the times work better than "Silence" with inaudible effects, if errors are not important, but sometimes the errors are important and can crash the decoders or produce big cracking noise like you say.

[madshi]

The "Silence" method is always the most sure (works always, maintain the sync)
I think there are different kinds of corruption. The usual kind is probably that the length of the audio file is correct but that some of the data is corrupt. In that case "Silence" should maintain the sync. However, if the source container is corrupt (or if the demuxing routine has a bug) there could be additional invalid data in the source file. As a result the final repaired audio file could be too long. In that case "Skip" might maintain sync while "Silence" could lose sync.

[Jedi_Vader20]

On huge interleaved waves (+4gb )the Runtime is often wrong.
But the mono waves and encoded Output from it is fine, no Problem

Do you need AC3 for your PSPS ?

No, wanted to downmix it to two channel and encode to AAC.

When I go to try and convert, all programs only handle the first 9 minutes and 35 seconds.

How do I split into six mono wavs?

[honai]

Does DTS-core extraction from DTS-HD implicitly remove Dial.Norm. from the stream?

[madshi]

No, wanted to downmix it to two channel and encode to AAC.

When I go to try and convert, all programs only handle the first 9 minutes and 35 seconds.
Check the previous posts from tebasuna51. He explained how to call the AAC decoder to workaround this problem. There's a special parameter to make big WAV files work.

[madshi]

Does DTS-core extraction from DTS-HD implicitly remove Dial.Norm. from the stream?
Yes...

[Schotenhüter]

Getting "Dump" Instance failed

whats that:stupid:

[madshi]

This is report of xport
=================================
D:\[DVDISO] AIR BDBOX VOL.1\BDMV\STREAM>xport -h 00003.m2ts 1 3 4
xport Transport Stream Demuxer 0.98
program = 1, video channel = 3, audio channel = 4
Program Number = 0 (0x0000), Program Map PID = 31 (0x001f)
Program Number = 1 (0x0001), Program Map PID = 256 (0x0100)
program descriptor = 0x05, 0x04, 0x48, 0x44, 0x4d, 0x56
program descriptor = 0x88, 0x04, 0x0f, 0xff, 0xff, 0xfe
ES descriptor for stream type 0x02 = 0x05, 0x08, 0x48, 0x44, 0x4d, 0x56, 0xff, 0
x02, 0x44, 0x3f
ES descriptor for stream type 0x81 = 0x05, 0x04, 0x41, 0x43, 0x2d, 0x33
ES descriptor for stream type 0x81 = 0x81, 0x04, 0x08, 0x3c, 0x0e, 0x00
ES descriptor for stream type 0x80 = 0x05, 0x08, 0x48, 0x44, 0x4d, 0x56, 0xff, 0
x80, 0x31, 0x7f
ES descriptor for stream type 0x81 = 0x05, 0x04, 0x41, 0x43, 0x2d, 0x33
ES descriptor for stream type 0x81 = 0x81, 0x04, 0x08, 0x3e, 0x04, 0x00
Audio PID = 4355 <0x1103>, type = 0x80
ES descriptor for stream type 0x80 = 0x05, 0x08, 0x48, 0x44, 0x4d, 0x56, 0xff, 0
x80, 0x61, 0x7f
LPCM Audio Mode = 3/2+lfe
LPCM Audio Bits/sample = 16
LPCM Audio Sample Rate = 48000
ts rate = unspecified
============================
With the xport demuxed audio file everything seems to work for me. Just use "eac3to xportdemuxed.pcm xportdemuxed.flac". I think the audio track you demuxed with TsRemux was corrupt.

[madshi]

F:\>eac3to\eac3to.exe Innocence.LPCM16.8ch.pcm innocence.eac3to.ac3 -8 -16 -down6 -640
RAW, 7.1 channels, 0:01:00, 16 bits, 48khz
Converting the raw file to wav. Please wait...
Converting the wav file to ac3. Please wait...

Aften: A/52 audio encoder
Version 0.0.8
(c) 2006-2007 Justin Ruggles, Prakash Punnoor, et al.

input format: RAW Signed 16-bit little-endian 48000 Hz stereo
output format: 48000 Hz stereo (2/0)

SIMD usage: MMX SSE SSE2 SSE3
Threads: 2

progress: 100% | q: 306.6 | bw: 60.0 | bitrate: 640.0 kbps

Done.

innocence.eac3to.ac3 is just noise and 4 times bigger than the correct 5.1ch file.


tried converting the eac3to-generated wav file to ac3 using aften.exe.
F:\>eac3to\aften\aften.exe Innocence.LPCM16.8ch.wav innocence.aften.ac3 -b 640

Aften: A/52 audio encoder
Version 0.0.8
(c) 2006-2007 Justin Ruggles, Prakash Punnoor, et al.

input format: WAVE Signed 16-bit little-endian 48000 Hz 5.1-channel
output format: 48000 Hz 3/2 + LFE

SIMD usage: MMX SSE SSE2 SSE3
Threads: 2

progress: 100% | q: 255.5 | bw: 57.9 | bitrate: 640.0 kbps

innocence.aften.ac3 seems OK

converting to the other formats than ac3 works correctly.
That was a bug in eac3to. Thanks for reporting it. I'll upload a new build in a moment. Btw, "-down6" and "-640" are default when converting to ac3.

[madshi]

whats that:stupid:
You need to download GraphEdit. The file "dump.ax" is shipping with GraphEdit. Then run "regsvr32 dump.ax" in the command line to register the "dump" filter.

[madshi]

eac3to v1.15 released

http://madshi.net/eac3to.zip

This time there are only a few bugfixes.

[tebasuna51]

I think there are different kinds of corruption. The usual kind is probably that the length of the audio file is correct but that some of the data is corrupt. In that case "Silence" should maintain the sync. However, if the source container is corrupt (or if the demuxing routine has a bug) there could be additional invalid data in the source file. As a result the final repaired audio file could be too long. In that case "Skip" might maintain sync while "Silence" could lose sync.

Well, you are the author of last version I can't discuss with you :rolleyes: (I'm busy with ffmpeg to analyze your version)

But with previous DelayCut versions, see the post (http://forum.doom9.org/showthread.php?p=810264#post810264) all short frames are skipped and only full erroneus frames are replaced by silence then never the output can be long than input. But with 'Skip' option always the output is short than input because short and full frames are ignored.

[madshi]

Well, you are the author of last version I can't discuss with you :rolleyes:
Heh! No, actually I think you know more about delaycut than I do. I just copied every AC3 function and changed it to E-AC3 without really diving too deep into how it all really works. I definitely didn't change anything about how "skip", "silence" and "fix" work.

I'm busy with ffmpeg to analyze your version
That's good!

But with previous DelayCut versions, see the post (http://forum.doom9.org/showthread.php?p=810264#post810264) all short frames are skipped and only full erroneus frames are replaced by silence then never the output can be long than input.
The output can't be longer than the input. But it can be longer than the correct result would be (and that is what I meant). Let me give you an example:

Blu-Ray TrueHD tracks have AC3 frames + TrueHD blocks. delaycut in "Silence" mode will extract all correct frames just fine. But it might find the AC3 sync word in some TrueHD blocks. The CRC will not fit, but "silence" mode will then add silent frames instead. As a result you'll get an audio track which is too long. Agreed? So there are situations where "skip" is the better choice compared to "silence".

[tebasuna51]

Blu-Ray TrueHD tracks have AC3 frames + TrueHD blocks.

Then you are right one more time. :)

In standard ac3 streams only ac3 frames (goods or corrupted) can be present.

[Schotenhüter]

You need to download GraphEdit. The file "dump.ax" is shipping with GraphEdit. Then run "regsvr32 dump.ax" in the command line to register the "dump" filter.

i`am test it:thanks:

[madshi]

In standard ac3 streams only ac3 frames (goods or corrupted) can be present.
One thing I'm not sure about: Can a recorded TS stream be so corrupt that a video packet is mistakingly interpreted as an audio packet? If that's possible and if you then demux the AC3 stream, it might also contain too much data, I think.

[tebasuna51]

One thing I'm not sure about: Can a recorded TS stream be so corrupt that a video packet is mistakingly interpreted as an audio packet?

Well, is possible but improbable, not only the header must match but also the CRC frame.

[madshi]

Well, is possible but improbable, not only the header must match but also the CRC frame.
Ah, ok. Don't know much about TS yet. One of the things I still need to look into... :)

[Schotenhüter]

next prob :o

the audio decoder didn`t accept the input stream

:confused:

[madshi]

next prob :o
I don't have a telepathical connection to your computer. So if you don't give me any more details I can not help you. I need the full eac3to output and information about the source audio file you fed to eac3to.

[Icemaan]

Hallo

By me your great Tool works fine. But only with 1536kps Dts.When I want downsample to 768kps it not work.The Sound looks horrible with 1536 all ok
Any Idea because I am new may be I made somethinh not correct
Thanks for help

Icemaan

[VempX]

Sorry,I have to go to school Orz
I demux the audio track from m2ts by tsremux then I get a .pcm file.
I try to use eac3to to convert it to flac,but I get a strange .flac file which I gave you~
Does eac3to can't convert .wav file?
I demux the audio by gdsmux so I get a .wav file.But eac3to can't convert it =A=

[VempX]

Sorry,I have to go to school Orz
I demux the audio track from m2ts by tsremux then I get a .pcm file.
I try to use eac3to to convert it to flac,but I get a strange .flac file which I gave you~
Does eac3to can't convert .wav file?
I demux the audio by gdsmux so I get a .wav file.But eac3to can't convert it =A=

[VempX]

sorry
net hava a wrong
please delete this post

[markrb]

Strange problem and sorry if it's been discussed before. I did try and find it.

After converting an EVO to ac3 the file is small (165mb) and plays way too fast. I checked the wav file and it is 1.74gb, but it also plays way too fast.

There is no indication of anything wrong. The screen looks exactly like the successful screen shots I have seen here.
It even says 100% done at the end when aften is finished.

I did check the EVO with the single dd+ stream and that plays fine. I have tried it both command line and using the Gui with the same results.

2 Discs with the same issue. So I don't think it's the discs.

EDIT: I think my Nero was hosed. I reinstalled and I am testing it now.

Yep that was it. Un-install Nero. Use clean tool. Re-install. Seems to work fine.

Mark

[nautilus7]

You need to download GraphEdit. The file "dump.ax" is shipping with GraphEdit. Then run "regsvr32 dump.ax" in the command line to register the "dump" filter.

Where can i download graphedit?

[markrb]

Where can i download graphedit?

Try Google. Then look for the register.bat file and run that.

Mark

[Icemaan]

Hi all together

I have following Problem
When i merge the Video and Audio with MkvMerge i get this Message

audio/video go out of sync, but this stream is damaged.
Warning: dts_packetizer: skipping 3 bytes (no valid DTS header found). This might make audio/video go out of sync, but this stream is damaged

This only come bei Dts not by Ac3. Can anywhere help me.The Video File is 264 the Audio dts

Thanks a lot
Icemaan

[TheSof]

If I have just the TrueHD extracted via evodemux (.mpa), is there anyway I can get eac3to to accept it?

[honai]

@madshi

I've found a parsing problem with Pan's Labyrinth HD-DVD (EU release). For the English DTS-HD track EVOdemux reports:


DTS HD (DTS) audio stream 1 found!
First PTS = 0000D6D8
Substream id = 89
PCM core samples = 32
PCM sample blocks = 16
Primary frame byte size = 1023
Channel arrangement = LT + RT (left and right total)
Sampling frequency = 48 kHz
Transmission bitrate = 768 kbit/s
LFE channel = not present
Encoder software rev. = 7, Current
Source PCM resolution = 16 bits
PCM core samples = 32
PCM sample blocks = 16
Primary frame byte size = 1023
Channel arrangement = LT + RT (left and right total)
Sampling frequency = 48 kHz
Transmission bitrate = 768 kbit/s
LFE channel = not present
Encoder software rev. = 7, Current
Source PCM resolution = 16 bits


It's actually DTS stereo (2.0). eac3to complains:

The source file format is unknown.

[Roscoe62]

I've followed this thread as best as I can, but I still haven't figured out whether it will do this or not. I have a couple of questions and any help is appreciated!

1. Is it possible to demux a Dolby TruHD audio track from a BR disc? From what I've seen earlier in the thread it might be possible with xport...but where can I find xport?

2. Will EAC3to convert a Dolby TruHD audio track demuxed from a BR Disc into a multi-channel wav file? (or flac). If so, how?

3. If all the above IS possible how can I mux the newly created 5.1 wav file into a .ts container with the AVC video, while still staying in sync? If .ts isn't the way to go, could it be done with mkv? (I thought mkv had problems with AVC video)

I know there are a lot of questions here, but even getting answers for the first 1 or 2 would be great.

Spiderman 1 & 2 are being released very soon and they both have TruHD soundtrack which I'd like to backup without losing any quality. This is all pretty standard from HD-DVD, but I've not done anything like this with BR.

Again, any help is appreciated!

[madshi]

By me your great Tool works fine. But only with 1536kps Dts.When I want downsample to 768kps it not work.The Sound looks horrible with 1536 all ok
Any Idea because I am new may be I made somethinh not correct
Will need to check if I can reproduce the problem. Can you please try to use "eac3to source.whatever dest.wavs"? This will give you 6 wav files. Then please convert these 6 wavs to dts with Surcode manually. Is the resulting file ok?

[madshi]

Sorry,I have to go to school Orz
I demux the audio track from m2ts by tsremux then I get a .pcm file.
I try to use eac3to to convert it to flac,but I get a strange .flac file which I gave you~
Yeah, but you also gave me the PCM track demuxed by xport - and that works fine! So it seems that the PCM track taken from tsremux is corrupt. Use xport and it should work fine.

[madshi]

I have following Problem
When i merge the Video and Audio with MkvMerge i get this Message

audio/video go out of sync, but this stream is damaged.
Warning: dts_packetizer: skipping 3 bytes (no valid DTS header found). This might make audio/video go out of sync, but this stream is damaged

This only come bei Dts not by Ac3. Can anywhere help me.The Video File is 264 the Audio dts
Where is that dts file coming from? Was it created by eac3to?

[madshi]

If I have just the TrueHD extracted via evodemux (.mpa), is there anyway I can get eac3to to accept it?
Please read the first post of this thread. Short answer: No.

[madshi]

I've found a parsing problem with Pan's Labyrinth HD-DVD (EU release). For the English DTS-HD track EVOdemux reports:

It's actually DTS stereo (2.0). eac3to complains:
Can you please rebuild the DTS-HD track into a separate EVO file (with only that DTS-HD track in and nothing else) and then send me the first 5MB of that rebuilt EVO? Thanks!

[madshi]

1. Is it possible to demux a Dolby TruHD audio track from a BR disc? From what I've seen earlier in the thread it might be possible with xport...but where can I find xport?
Yes, it's possible with xport. You can download it here:

http://www.w6rz.net/

TsRemux should also be able to demux the TrueHD track.

2. Will EAC3to convert a Dolby TruHD audio track demuxed from a BR Disc into a multi-channel wav file? (or flac). If so, how?
Blu-Ray TrueHD tracks consist of an AC3 core with some additional TrueHD packets on top of that. eac3to can extract the AC3 core. Do "eac3to source.thd dest.ac3". This will give you the AC3 core of that TrueHD track. Obviously you'll only get 640kbps AC3 quality this way and not the full TrueHD quality. Currently to my best knowledge only PowerDVD can decode Blu-Ray TrueHD tracks and PowerDVD's Audio Decoder cannot be used outside of PowerDVD.

Spiderman 1 & 2 are being released very soon and they both have TruHD soundtrack which I'd like to backup without losing any quality. This is all pretty standard from HD-DVD, but I've not done anything like this with BR.
I'd suggest that you use TsRemux to rebuild the TrueHD tracks into a separate m2ts file. I think this way you should be able to back them up without wasting much space and with good hopes that we might find a way later to fully decode them. Of course you could also backup the whole movie m2ts. That would be the easiest way, but consume more space. Ideally, why backing them up right now? You should buy the Blu-Ray discs. So you have the data on your shelf, anyway... :)

[Roscoe62]

Ideally, why backing them up right now? You should buy the Blu-Ray discs. So you have the data on your shelf, anyway... :)

Thanks for the info Madshi. I will be placing my order for the discs tomorrow but I will not use PowerDVD for playback. I only playback via .ts or .mkv files so I can use my fav player - Theatertek. So I guess I will stick with the AC3 core for the time being, until someone smarter than me can figure out how to fully decode Blu-Ray TruHD tracks.

Here's hoping anyway! :)

[TheSof]

I only playback via .ts or .mkv files so I can use my fav player - Theatertek.

TT will happily play m2ts files.

[Icemaan]

Where is that dts file coming from? Was it created by eac3to?

The Dts File was created with eac3to. The File comes from the rebuild evo with only one audio stream inside

I have split the Evo with Evosplit may be this is the problem what you think

icemaan

[madshi]

The Dts File was created with eac3to. The File comes from the rebuild evo with only one audio stream inside

I have split the Evo with Evosplit may be this is the problem what you think

icemaan
I'm not sure. Does that warning come very often or just once? Is there a timecode for that warning? Could it be the timecode where you split the evo?

[Roscoe62]

TT will happily play m2ts files.

Perhaps,

but because of the TruHD soundtrack Theatertek won't play it unless I demux the BR M2TS, re-code the audio and then re-mux into a new container. And if I'm going to all that effort I'd rather re-mux to a container I'm familiar with.

[ACrowley]

Perhaps,

but because of the TruHD soundtrack Theatertek won't play it unless I demux the BR M2TS, re-code the audio and then re-mux into a new container. And if I'm going to all that effort I'd rather re-mux to a container I'm familiar with.


recode the Audio !? Just extract the DD core, no need to renencode it. Also maybe TT decodes only the DDcore from, Bluray TrueHD ?! i

[pimpMyHD]

I am interesting in using your software, but i have a question :

do i need nero hddvd/bluray plugin, if i already have sonic scenarist filters ?

I currently do not own nero 's plugin, but i own the sonic hd filters.

[TheSof]

I am interesting in using your software, but i have a question :

do i need nero hddvd/bluray plugin, if i already have sonic scenarist filters ?

I currently do not own nero 's plugin, but i own the sonic hd filters.

You need them if you are after the best sound quality. The sonic filters add nasty DRC.

[pimpMyHD]

You need them if you are after the best sound quality. The sonic filters add nasty DRC.
So, i should be able to use eac3to without the nero filters ?

Because it does not work on my system.

[TheSof]

So, i should be able to use eac3to without the nero filters ?

Because it does not work on my system.

If you just have the sonic decoders yes it should still work. What error message are you getting?

[Thunderbolt8]

I have a disc with 5.1 DD+ audio codec, what should I do now when I want to create a remux and this should include the best possible audio I can make out of it, which is currently possible to playback with a normal setup? I also guess that keeping the track as it is is also not a good solution (apart from the question if it was playable at all then?), because theres lot of additional junk in it like with dts-hs & dtscore ?
I think the solution would be something like converting to normal ac3 or DTS. but im not sure if dts conversion is possible then from a DD+ source, or if it does make sense at all?
another question I have is the removal of dialog normalisation, is such dialog normalisation included in every audio track? is the dialog normalisation somehow intended by the film makers, otherwise one wouldnt understand what people say in certain scenes? does it sound more natural & better without dialog normalisation?

[madshi]

I have a disc with 5.1 DD+ audio codec, what should I do now when I want to create a remux and this should include the best possible audio I can make out of it, which is currently possible to playback with a normal setup?
Define "normal setup".

If you have analog connection to your receiver or if you can transport multi channel PCM audio over HDMI to your receiver then the best way to handle DD+ audio is to keep it as it is. However, if you have SPDIF/coax connection to your receiver, you can only transport AC3 or DTS. So in that case you'd have to reencode the DD+ track to either AC3 or DTS.

I also guess that keeping the track as it is is also not a good solution (apart from the question if it was playable at all then?), because theres lot of additional junk in it like with dts-hs & dtscore ?
There's no junk in DD+ at all. And it's a good idea to keep it as it is. But see above.

another question I have is the removal of dialog normalisation, is such dialog normalisation included in every audio track? is the dialog normalisation somehow intended by the film makers, otherwise one wouldnt understand what people say in certain scenes? does it sound more natural & better without dialog normalisation?
The one and only purpose of dialog normalization is that it tries to save you from having to readjust volume during advertising breaks and when you switch from one movie to another. Basically dialog normalization turns the volume knob on your receiver so that the volume of speaking voices is the same in advertising and in every movie you're watching. Without dialog normalization advertising can sometimes be much louder than the movie audio track. Or it could happen that one movie is much louder mixed than another movie. So without dialog normalization you'll have to adjust the volume everytime there's an advertising break or everytime you switch to another channel/movie to keep volume at the same level.

Unfortunately, in real life most movies leave the "dialog normalization" information in their audio tracks at the default value. That means dialog normalization often simply doesn't work correctly. Furthermore many people argue that doing dialog normalization harms audio quality. Because practically it's not really the physical volume knob of your receiver that is changed by dialog normalization, but instead the digital audio track is processed to make audio volume lower. This effectively reduces bitdepth of the audio track (a very little bit).

Sony has decided to not use dialog normalization, anymore. This is based on feedback they received from home cinema consumers. But most other studios still leave dialog normalization activated (and most of the time on the default value).

[Thunderbolt8]

well, standart setup (should have said basic setup) means I just use my pc via ac3filter (or maybe ffdshow audio) to playback the movie/audio and currently just have 2 hi-fi boxes (non computer boxes), which are addressed via the rear green output. so basically nothing speacial, no external receiver or such :D
would the original, non converted track be able to played back via ac3filter? would I be able to listen to any improvements compared to a converted ac3/dts track? I guess the removal of the dialog normalization of the DD+ track then wouldnt change the fact that its still (un)playable with ac3filter?
in case I decide to convert to dts or ac3 and also would like to use dialog normalization, is there a fixed order I should apply to, e.g. first dialog normalization removal and then converting (or vice versa)?
when remuxing with mkvmerge, should the ending for the DD+ track kept .mpa or should/must it be changed into something else ?

[madshi]

well, standart setup (should have said basic setup) means I just use my pc via ac3filter (or maybe ffdshow audio) to playback the movie/audio and currently just have 2 hi-fi boxes (non computer boxes), which are addressed via the rear green output. so basically nothing speacial, no external receiver or such :D
Not even multi channel! :(

would the original, non converted track be able to played back via ac3filter?
Nope, at least not until ac3filter adds E-AC3 decoding support.

in case I decide to convert to dts or ac3 and also would like to use dialog normalization, is there a fixed order I should apply to, e.g. first dialog normalization removal and then converting (or vice versa)?
As long as you use eac3to for the conversion, you don't have to care.

when remuxing with mkvmerge, should the ending for the DD+ track kept .mpa or should/must it be changed into something else ?
Don't know how mkvmerge behaves. But I'd suggest renaming it to "ddp" or "eac3". Just try it out. If mkvmerge detects that it's E-AC3, you're done.

[Thunderbolt8]

well, since the track should be quite small compared to the video size of the remux I guess I'll do both, convert it, but also try to mux the original track into the .mkv to be on the save side for the future. when time comes, I only need to remux without the converted track then, instead of working with the HDDVD source again :S

edit: no changing of the ending of the demuxed audio at all needed for DD+, mkvmerge detects is as eac3 at once :)

[Thunderbolt8]

hm just read the first line of the OT again, wouldnt it make more sense for me just to convert the DD+ to flac? would it be completely lossless then, compared as when taking flac from a LPCM track? I guess that might be the best option then, because I can use it already now and im sure with better hardware it should still be possible to use it still in the future (wasnt there something about analog plug that allows flac use with external receivers?). or am I mistake here?

btw. what the difference between that flac121a, flac121b and the win.zip file from sourceforge, which one should I download?

[Taktaal]

Has anyone kept an old version of eac3to? It used to work very well but the versions in the last 2 months added only bloat and more dependancies. Unfortunately I overwrote my download from earlier this year...

[juzzy999]

I cant get eac3to encode a DTS-HD track.

I wont recognise the file.. How can i re-encode it to an AC3??????

[madshi]

hm just read the first line of the OT again, wouldnt it make more sense for me just to convert the DD+ to flac? would it be completely lossless then, compared as when taking flac from a LPCM track? I guess that might be the best option then, because I can use it already now and im sure with better hardware it should still be possible to use it still in the future (wasnt there something about analog plug that allows flac use with external receivers?). or am I mistake here?
Sure you can do that. But DD+ decodes to 24bit. And reencoding that to FLAC will consume a lot of space. Lossless compression works best for 16bit. The higher the bitdepth of the audio track is the lower is the compression efficiency of lossless compression.

btw. what the difference between that flac121a, flac121b and the win.zip file from sourceforge, which one should I download?
The FLAC which comes with eac3to is the only one which can handle big files without any problems.

[madshi]

Has anyone kept an old version of eac3to? It used to work very well but the versions in the last 2 months added only bloat and more dependancies. Unfortunately I overwrote my download from earlier this year...
Which dependancies do you mean? If you have a suggestion how to improve eac3to, just let me know.

[madshi]

I cant get eac3to encode a DTS-HD track.

I wont recognise the file.. How can i re-encode it to an AC3??????
I can't say anything without more details. Where did you get the DTS-HD track from? Which file extension does it have? Try renaming it to "something.dtshd". And please post the full eac3to text output here.

[onlym3]

How do i convert a TrueHD track from a Bluray? The info on the first page says use evodemux to create an evo but that only works for HD DVDs since you first need to load an evo to be able to rebuild it.

[Thunderbolt8]

regarding the blue ray thing, how do I manage to use something similar like offsetpts for evo files for m2ts files?
afaik a common way is, when having multile m2ts files to add them together like with copy /b etc. but what if there are also slightly audio delay problems then, like in some hddvd cases? for that reason madshi created offsetpts, but how can I treat the blue ray files to prevent desyncing?

[madshi]

How do i convert a TrueHD track from a Bluray? The info on the first page says use evodemux to create an evo but that only works for HD DVDs since you first need to load an evo to be able to rebuild it.
Blu-Ray TrueHD tracks are not fully supported yet. You can demux them with "TsRemux" or "xport" and then eac3to can extract the AC3 track which is stored inside the TrueHD track. But currently decoding of the full TrueHD information is not possible yet. However, I hope to be able to make that work soon.

[madshi]

regarding the blue ray thing, how do I manage to use something similar like offsetpts for evo files for m2ts files?
afaik a common way is, when having multile m2ts files to add them together like with copy /b etc. but what if there are also slightly audio delay problems then, like in some hddvd cases? for that reason madshi created offsetpts, but how can I treat the blue ray files to prevent desyncing?
Currently I don't know any way to "delay" m2ts files. Maybe I'll enhance offsetpts sooner or later to also handle m2ts files.

[Thunderbolt8]

i wonder about one thing, I converted the 5.1 eac3 from mad max: the road warrior to ac3 and the size remained the same for the ac3 track. I did the same with the 2.0 eac3 commentary track, but the ac3 size doubled in that case. is this a normal procedure? afaik the eac3 track had 192kbps, but the ac3 file now has 384kbps :S (made with eac3to)

[madshi]

i wonder about one thing, I converted the 5.1 eac3 from mad max: the road warrior to ac3 and the size remained the same for the ac3 track. I did the same with the 2.0 eac3 commentary track, but the ac3 size doubled in that case. is this a normal procedure? afaik the eac3 track had 192kbps, but the ac3 file now has 384kbps :S (made with eac3to)
File size depends only on bitrate and movie duration. You can tell eac3to which bitrates to use. If you don't do that eac3to decides for its own.

[Thunderbolt8]

hm since the program made it right with the 640kbps for the 5.1 track I assume it could somehow autodetect the right bitrate. guess I have to convert again then and enter 192kbps manually

[madshi]

hm since the program made it right with the 640kbps for the 5.1 track I assume it could somehow autodetect the right bitrate. guess I have to convert again then and enter 192kbps manually
Well, what is "right" and what is "wrong"? Reencoding with a lossy encoder always hurts audio quality. eac3to has *intentionally* chosen a higher bitrate than the original bitrate in order to keep the audio quality loss as small as possible. For 5.1 tracks eac3to is generally using 640kbps. For stereo tracks eac3to is generally using 384kbps, IIRC. These are default values I've chosen. If you don't like them, you have the freedom to tell eac3to to use other values. But I'm not going to lower the defaults. I don't agree that matching the original bitrates is "right".

[Taktaal]

The problem is that both Nero and Sonic are really bugged. Nero is a 1.4GB program just to burn some CDs and play movies, which would be funny by itself if it wasn't so pathetic, is just about impossible to configure properly for HD-DVD playback and hoses your Windows install after a few tries of getting it to work.
And Sonic just dumps a few uncaught exceptions right onto the command line in a stunning display of programmer incompetence if you want to use it to decode E-AC3.

[honai]

The problem is that both Nero and Sonic are really bugged.

If you have found bugs you can always report them to Nero or Sonic.

Nero is a 1.4GB program

What? My installation folder weighs less than 200 MB.

which would be funny by itself if it wasn't so pathetic

What is so pathetic about it?

is just about impossible to configure properly for HD-DVD playback

Impossible as in not possible? How come that it works for others?

and hoses your Windows install after a few tries of getting it to work.

No, it doesn't. The problem most likely exists between keyboard and chair.

And Sonic just dumps a few uncaught exceptions right onto the command line in a stunning display of programmer incompetence if you want to use it to decode E-AC3.

The Nero filters are working fine.

And looking back at your post from 24th September 2007, 23:56:

Has anyone kept an old version of eac3to? It used to work very well but the versions in the last 2 months added only bloat and more dependancies.

and considering that you don't offer any specific bug reports at all it seems that your major intention in this thread is to badmouth software that - gasp! - seems to work just fine for many other users.

I find it especially rich that you claim that the latest versions of madshi's work "added only bloat". He's working on this for free, and you get even the full source code. I don't think you're in a position to discard his work in such a way.

[Taktaal]

It was perhaps not said in a very nice way, I'm sorry. But I did specify that the issues aren't with the eac3to code itself, but with the dependencies.

[honai]

So what are the issues exactly? Perhaps readers can help you out.

[madshi]

I did specify that the issues aren't with the eac3to code itself, but with the dependencies.
I'm still not sure which dependencies you're talking about. There have been no dependencies added at all in recent versions. Actually I've been *reducing* dependencies over time.

[madshi]

The problem is that both Nero and Sonic are really bugged. Nero is a 1.4GB program just to burn some CDs and play movies, which would be funny by itself if it wasn't so pathetic, is just about impossible to configure properly for HD-DVD playback
Fortunately all we need from the Nero package in this thread is the Nero Audio Decoder and that one is relatively small and works quite well.

[Thunderbolt8]

I read somewhere that with sonic audio decoder can I can actually use the DD+ track, it is able to decode it. is this right and can I benefit from this track with this decoder already fully, meaning hes able to decode & output 'everything there is' additionally compared to an ac3 only track?

[madshi]

I read somewhere that with sonic audio decoder can I can actually use the DD+ track, it is able to decode it. is this right and can I benefit from this track with this decoder already fully, meaning hes able to decode & output 'everything there is' additionally compared to an ac3 only track?
Read the first post of this thread to learn about disadvantages of the Sonic decoder.

[madshi]

eac3to v1.16 released

http://madshi.net/eac3to.zip

This version can now fully decode demuxed TrueHD files from both HD DVD and Blu-Ray!!

That means: Sony TrueHD tracks can now be decoded bit perfectly because Sony doesn't use Dialog Normalization. Sadly, for other studios TrueHD decoding still isn't bit perfect because the Nero Audio Decoder does post processing (Dialog Normalization).

Here are test results with "The Fifth Element (Remastered)":

(1) Original 20bit TrueHD track in m2ts container: 4.12GB.
(2) Demuxed TrueHD track: 3.36GB.
(3) TrueHD after removal of interweaved AC3 frames: 2.81GB.
(4) 20bit TrueHD reencoded to FLAC: 2.54GB.
(5) 16bit LPCM track reencoded to FLAC: 1.55GB.

From the numbers it seems that TrueHD decoding is perfect for Sony Blu-Ray tracks. Now if only I knew how to remove Dialog Normalization for Blu-Ray tracks of other studios... :(

[HD-DVD±RW]

@madshi just wanted to thank you greatly for providing a super useful application. I've used it on many occasions with great success. Keep up the good work.

[Thunderbolt8]

^^completely agree. well done.

[nmeli15]

When trying to convert ddp to ac3 with eac3to with Sonic-orbitlee error message.............gettin dts/ac3/dd+ source instances failed............what does that mean........

[madshi]

@madshi just wanted to thank you greatly for providing a super useful application. I've used it on many occasions with great success. Keep up the good work.
Thanks for the feedback!

[madshi]

When trying to convert ddp to ac3 with eac3to with Sonic-orbitlee error message.............gettin dts/ac3/dd+ source instances failed............what does that mean........
It probably means that the orbitlee filter is not properly registered.

[Thunderbolt8]

whats the difference between orbitlees's DD+ source filters and the ones that are applied otherwise, when those are not applied?

[homerpez]

I've hit a serious snag while trying to back up the Cantonese audio from "Kung Fu Hustle" on Blu-Ray. The goal is a standard, ordinary 5.1 AC3 file, since it's just for backup.

The "Lossless PCM" track is 5.1 audio, and Power DVD info tells me it's LPCM.

I Demux this track with TSremux, and call it "Kung Fu Hustle.pcm".

Using Eac3to (the latest as of this date), this is what it gives me:

C:\temp\TEST>"C:\VIDEO\Rippak\EAC3TO\eac3to.exe" "C:\temp\TEST\Kung Fu Hustle.pc
m" "C:\temp\TEST\Kung Fu Hustle.ac3" -448 -orbitlee -16 -big
RAW, 5.1 channels, 1:39:04, 16 bits, 48khz
Converting the raw file to wav. Please wait...
Converting the wav file to ac3. Please wait...

Aften: A/52 audio encoder
Version 0.0.8
(c) 2006-2007 Justin Ruggles, Prakash Punnoor, et al.

Warning, unsupported file size.
input format: WAVE Signed 16-bit little-endian 48000 Hz 5.1-channel
output format: 48000 Hz 3/2 + LFE

SIMD usage: MMX SSE SSE2 SSE3
Threads: 2

progress: 0% | q: 188.4 | bw: 34.0 | bitrate: 448.0 kbps

Done.

I know the LPCM stream works, when I play the movie off the hard drive in PowerDVD,it sounds fine.

When I play back the RAW PCM file, and it's garbled digital sound (I think it's supposed to at this stage). I play back the resulting AC3 file from running Eac3to, and it's still garbled, though I can hear the sound some in the background (but it is still digital noise). On the AC3, it turns into an audible BUZZZZZZ sound.

I am fairly sure that it is 16-bit instead of 24 bit, because when I select that, the timecode becomes correct in Eac3to's status, and the resulting AC3 matches the movie (on 24-bit, it doesn't). This does no good, when the sound is digital mish-mash. :P

I have tried both 24-bit and 16-bit, big endian, little endian, as well as forcing either the nero filter or the orbitlee filter. The result is the same: the AC3 has an audible BUZZZZZZ and is digitally garbled.

Is there something I'm doing wrong? Hopefully the copied info above may be able to help? Thanks in advance...

[madshi]

@homerpez, try using xport instead of TsRemux. I've had trouble with TsRemuxed LPCM tracks before while xport has never failed me yet.

[Sephiroth0000]

Can anyone help at all??
I know most of you are far ahead with regards to HD and making it WMV and what not with the 5.1 audio but im finding myself having serious problems here. I am following a persons tutorial on converting HD-DVDs to WMV HD with 5.1 audio channels http://forum.doom9.org/showthread.php?t=123010&page=5 thread no. 87. Now I have got to the very last stage of the tutorial but when I attempt to transcode I get this error...

C:\HD>eac3to.exe "C:\hd\PEVOB_1_MERGED.DD+.stream.00.ddp" "C:\hd\output.wav"
E-AC3, 5.1 channels, 2:16:18, 640kbit/s, 48khz, dialnorm: -27dB
Remove Dialog Normalization information. Please wait...
Muxing eac3 file to raw. Please wait...
Getting "Nero Audio Decoder 2" instance failed.

C:\HD>tranzcode.exe "C:\hd\output.wav"
Error: Unable to open file: C:\hd\output.wav

C:\HD>wavavimux -o audio.avi -iwav 6 output-FL.wav output-FR.wav output-C.wav ou
tput-LFE.wav output-SL.wav output-SR.wav -mask 63
'wavavimux' is not recognized as an internal or external command,
operable program or batch file.

C:\HD>cscript.exe "C:\Program Files\Windows Media Components\Encoder\wmcmd.vbs"
-input "C:\hd\Video_Only.avs" -a_input "C:\hd\audio.avi" -output "C:\hd\output.w
mv" -v_bframedist 1 -v_bdeltaqp 1 -v_mslevel 1 -v_codec wvc1 -v_keydist 3 -v_mod
e 2 -v_quality 96 -v_mmatch 0 -v_msrange 0 -v_performance 20 -v_loopfilter 1 -a_
codec wmapro -a_setting 384_48_6_16

I have tried looking up these errors but with no success. If anyone could give me any input with this I would really appreciate it because to be honest im all out of idea's.

[Taktaal]

Your first command already failed so there's no use in continuing with the others.
The error code seems to indicate that you don't have Nero installed on your system. Once you install it you should get to a point where it says something like "expected 4GB file, but got 0 byte". And that's about as far as you're gonna get!

[Sephiroth0000]

Your first command already failed so there's no use in continuing with the others.
The error code seems to indicate that you don't have Nero installed on your system. Once you install it you should get to a point where it says something like "expected 4GB file, but got 0 byte". And that's about as far as you're gonna get!

But I do have Nero installed and I also have the Blu-Ray HD DVD plugin also. You seem to know what you are talking about so do you have any suggestions at all please? Where is that I have gone wrong? I understand that it is trying to load up Nero to continue the process of the audio but what I do not understand is why the error?

[TheSof]

I just used eac3to on a demuxed truehd .mpa (superman returns) and it worked fine. However, it produced a 24bit flac. If the thd is 16 bit, should I be using the -16 switch? Or do the nero decoders only output 24bit?

[Thunderbolt8]

good question, I was wondering in a similar way, whether I should use 16 or 24 bit for FLAC for a 24-bit LPCM lossless track. madshi told somewhere above that the compression works better for 16-bit, but I cant judge how much the loss would be compared to 'real lossless' then with 24-bit or something like that. at least most people I know restrict flac to 16-bit.

[tebasuna51]

I've hit a serious snag while trying to back up the Cantonese audio from "Kung Fu Hustle" on Blu-Ray. The goal is a standard, ordinary 5.1 AC3 file, since it's just for backup.

The "Lossless PCM" track is 5.1 audio, and Power DVD info tells me it's LPCM.
...

If your file is really LPCM and you are a BeHappy user maybe you can try open this avs file in BeHappy:
NicLPCMSource("C:\temp\TEST\Kung Fu Hustle.pcm", 48000, 16, 6)

Maybe you need the last NicAudio_20070821 (http://avisynth2.sourceforge.net/NicAudio_20070821.zip)

[homerpez]

@homerpez, try using xport instead of TsRemux. I've had trouble with TsRemuxed LPCM tracks before while xport has never failed me yet.

Thanks! That worked!

It was a little effort getting xport to do what I wanted (I hate command line programs...) but it did eventually work for me. You saved the day.

[madshi]

But I do have Nero installed and I also have the Blu-Ray HD DVD plugin also. You seem to know what you are talking about so do you have any suggestions at all please? Where is that I have gone wrong? I understand that it is trying to load up Nero to continue the process of the audio but what I do not understand is why the error?
Do you have Nero 7 or Nero 8 installed? Try the latest Nero 7 version. The error eac3to is reporting simply means that the Nero Audio Decoder could not be loaded successfully. I don't know why that happened. You could try (just as a check) renaming GraphEdit.exe to Recode.exe and check if GraphEdit can successfully load the Nero Audio Decoder.

[madshi]

I just used eac3to on a demuxed truehd .mpa (superman returns) and it worked fine. However, it produced a 24bit flac. If the thd is 16 bit, should I be using the -16 switch? Or do the nero decoders only output 24bit?
The Nero Audio Decoder always outputs 24bit for TrueHD regardless of which bitdepth the TrueHD file really has. And it gets even worse: Because of dialog normalization post processing all 24bits do have valid audio data in them. The "-16" switch will not help. This switch is only there for situations where the raw audio data is really 16bit. But in your case the audio data is output by the Nero Audio Decoder is 24bit. So using the "-16" switch would result in garbled audio. You could use the "-down16" switch to downconvert the 24bit data to 16bit, but you'd lose a bit of audio quality this way. What we really need is a way to turn dialog normalization post processing off. If we do that, the Nero Audio Decoder will still output 24bit, but the lower 8 bits will be all zero. FLAC will then encode that as a 24bit FLAC file, but the final FLAC file size will be similar to a 16bit FLAC.

[Sephiroth0000]

Does anyone have any suggestions as to what I should do with regards to the problems im having? (problem on page 44). I know that the command prompt has failed because it cannot find Nero Audio Decoder 2 but I do have Nero installed and I even installed the BluRay HD DVD plugin but still with no success. I have tried looking for this Nero Audio Decoder 2 on my system but cannot find it but my Nero install is complete and it is uptodate aswell.

[SvT]

Does anyone have any suggestions as to what I should do with regards to the problems im having? (problem on page 44). I know that the command prompt has failed because it cannot find Nero Audio Decoder 2 but I do have Nero installed and I even installed the BluRay HD DVD plugin but still with no success. I have tried looking for this Nero Audio Decoder 2 on my system but cannot find it but my Nero install is complete and it is uptodate aswell.

Check your product setup in NERO. The serial for the plugin should show in this list ! If not ... your plugin is not fully registered. (there must be at least 2 serials 1 for Nero and 1 for the plugin)
If the serial is there, I can see no reason why eac3to can't find it.
Did you try Madshi's suggestion ? (renaming GraphEdit.exe......)

Greets.

[Sephiroth0000]

Check your product setup in NERO. The serial for the plugin should show in this list ! If not ... your plugin is not fully registered. (there must be at least 2 serials 1 for Nero and 1 for the plugin)
If the serial is there, I can see no reason why eac3to can't find it.
Did you try Madshi's suggestion ? (renaming GraphEdit.exe......)

Greets.

At last :) Thankyou! I do have the serials for both Nero and the BluRay HD plugin. Renaming Graphedit!? I know its a pain but could you go more into detail please?

[SvT]

@Sephiroth0000 See Madshi's reply to your question (post #884)

[Sephiroth0000]

Do you have Nero 7 or Nero 8 installed? Try the latest Nero 7 version. The error eac3to is reporting simply means that the Nero Audio Decoder could not be loaded successfully. I don't know why that happened. You could try (just as a check) renaming GraphEdit.exe to Recode.exe and check if GraphEdit can successfully load the Nero Audio Decoder.

I have Nero 8 and the Blu-Ray HD plugin.

[Sephiroth0000]

SVT Graphedit can load up Nero Audio Decoder 2. Im sorry everyone but this is very new to me.

[Rectal Prolapse]

Nero 8 will not work - if I read what a Nero deveoper said to someone correctly, Nero 8 blocks access to their filters outside of Nero Showtime and other Nero software.

You are out of luck - you will have to downgrade to Nero 7 in the above case.

[Rectal Prolapse]

Despite the bad news - it could be possible to defeat Nero's protection by monitoring the calls - perhaps the techniques used to make a codecave could be used...

[Sephiroth0000]

I thankyou everyone for the input they have given me and im very thankful! I have downgraded to Nero 7 and now the audio decoder is picked up and it moves on to the next step which is where I encounter another error... (you were right by the way Taktaal)

Microsoft Windows XP [Version 5.1.2600]
(C) Copyright 1985-2001 Microsoft Corp.

C:\HD>"C:\HD\eac3to.exe" "C:\HD\PEVOB_1_MERGED.DD.stream.00.ddp" "C:\HD\output.
wav" -nero
E-AC3, 5.1 channels, 2:16:18, 640kbit/s, 48khz, dialnorm: -27dB
Remove Dialog Normalization information. Please wait...
Muxing eac3 file to raw. Please wait...
Disabling DRC for Nero E-AC3 decoding...
The file size of the raw file doesn't seem to fit.
The expected file size for 16 bit is 4.38 GB.
The expected file size for 24 bit is 6.58 GB.
The real file size is 0 Bytes.

Any suggestions at all? I do appreciate this everyone and at the end I will be putting all my notes together to make my own tutorial on converting HD DVD to EVO to WMV HD with 5.1

[madshi]

The real file size is 0 Bytes.
This usually happens when the audio decoder refuses to work. I'd suggest that you try uninstalling Nero and reinstalling it. Also make sure that you have really bought the HD DVD plugin. Pirated serial numbers are known to make problems with the HD DVD plugin.

[madshi]

eac3to v1.17 released

http://madshi.net/eac3to.zip

BIG NEWS.

The TrueHD decoding is now perfect for all HD DVD and Blu-Ray 5.1 TrueHD tracks when using the latest eac3to version and Nero 7 !!

Here's a quick test:

(1) TrueHD test file: 22.8 MB.
(2) Converted to FLAC with old eac3to version: 52.6 MB.
(3) Converted to FLAC with new eac3to version: 20.3 MB.

The size problem in (2) was caused by dialog normalization post processing which I've finally gotten rid of now. Please note that TrueHD files are always 24bit, even if the source was 16bit. But as you can see in the quick test above, this doesn't harm the file size, as long as no post processing is applied by the audio decoder.

TrueHD tracks with 2.0 and 7.1 channels are not supported yet. If you have such tracks, please send me a small sample (5MB should be enough) of the demuxed TrueHD audio track.

[Taktaal]

By the way make sure you take an Acronis image before trying to get Nero to work. After the second or third try I had to reinstall Windows from scratch.

[Thunderbolt8]

Great Job, Madshi, Well Done :)


edit: I understand it right, that the less space consumed by the flac track now is only restricted to 'real' 16-bit trueHD tracks? (because I made a test with a 24-bit file (at least evodemux says its 24-bit) and the size stayed the same with the new version, without using any additional option). so what basically happened to that track, size still the same, but dialog normalization removed now?

[Sephiroth0000]

I really appreciate the answers and suggestions I am getting from people.Thankyou :) and I will let you know what happens when I do as you have all suggested.

[hristoff2]

Nice one! :)

[madshi]

edit: I understand it right, that the less space consumed by the flac track now is only restricted to 'real' 16-bit trueHD tracks?
The size advantage is only there for anything < 24bit. TrueHD tracks can be 16bit or 24bit or anything in between. Sony uses 20bit for some Blu-Rays. In theory a TrueHD track can even temporarily switch to different bitdepths throughout the movie. But for a full 24bit TrueHD track the new eac3to version doesn't bring any size advantages. However, having dialog normalization turned off is nevertheless an advantage IMHO cause combining "lossless" with "post processing" doesn't really sound like a good idea to me.

[madshi]

A note to all people interested in TrueHD -> FLAC reencoding:

The new eac3to version is brand new and not well tested yet. So I'd suggest that you do some checks to make sure that the results are as expected before you delete the TrueHD file. E.g. a good test would be to check if audio is in sync both in the beginning and end of the movie.

If you've done some testing I'd be thankful for your feedback. A simple note like "worked fine for [x] movies" would be great. Thank you!

[Sephiroth0000]

Madshi I downloaded and used that new EAC3TO you offered for download and now it has worked the file properly and created the WAV. file and then even took it to the 5.1 multi channel WAV. outputs files but it crashes yet again afterwards saying that WavAviMu is not reconised as an eternal command but that's what im supposed to use next. If you look about 2 pages back you should see the error within my bat. command file when it saids it cannot reconise it. Any suggestions please!? (Im so close :))

[Thunderbolt8]

so far I'd say it works fine for fear & loathing HDDVD. although when observing correctly the cpu usage has increased, compared to the -down16 flac file I had used with 1.16. but I guess this is normal :P I havent observed a sync difference between the 1.16 -down16 flac and the 1.17 one.

[madshi]

Madshi I downloaded and used that new EAC3TO you offered for download and now it has worked the file properly and created the WAV. file and then even took it to the 5.1 multi channel WAV. outputs files but it crashes yet again afterwards saying that WavAviMu is not reconised as an eternal command but that's what im supposed to use next. If you look about 2 pages back you should see the error within my bat. command file when it saids it cannot reconise it. Any suggestions please!? (Im so close :))
Can I please see the full output? Is it eac3to which is crashing or is it a program in your batch after eac3to?

[madshi]

so far I'd say it works fine for fear & loathing HDDVD. although when observing correctly the cpu usage has increased, compared to the -down16 flac file I had used with 1.16. but I guess this is normal :P I havent observed a sync difference between the 1.16 -down16 flac and the 1.17 one.
Thanks for the feedback. I've no idea why the CPU usage should increase, but as long as the FLAC sounds correct (and the CPU usage isn't *too* high) I don't think we need to be worried.

P.S: When using "-down16" the resulting FLAC file is really a 16bit FLAC file. The FLAC file created by 1.17 is a 24bit FLAC file. This might explain the difference in CPU usage. But the file sizes should be roughly similar. How big is your 1.16 -down16 FLAC file in comparison to the 1.17 one?

[Sephiroth0000]

C:\HD>wavavimux -o audio.avi -iwav 6 output-FL.wav output-FR.wav output-C.wav ou
tput-LFE.wav output-SL.wav output-SR.wav -mask 63
'wavavimux' is not recognized as an internal or external command,
operable program or batch file.


That is the error I get after the audio process has completed. I did check that the WavAviMux is on my system and it is(c:\programs folder\windows media encoder\components\WAVAVIMUX\WavAviMux.exe) but for some reason it saids that it is not a reconised command or something. Im so close to sealing this. Thankyou for the input I really do appreciate it.

[madshi]

C:\HD>wavavimux -o audio.avi -iwav 6 output-FL.wav output-FR.wav output-C.wav ou
tput-LFE.wav output-SL.wav output-SR.wav -mask 63
'wavavimux' is not recognized as an internal or external command,
operable program or batch file.

That is the error I get after the audio process has completed. I did check that the WavAviMux is on my system and it is(c:\programs folder\windows media encoder\components\WAVAVIMUX\WavAviMux.exe) but for some reason it saids that it is not a reconised command or something. Im so close to sealing this. Thankyou for the input I really do appreciate it.
Well, it seems that you've now successfully passed the eac3to step. That's really all I know about. I've never used WavAviMux myself yet. But the reason why it fails for you is probably that WavAviMux.exe is in a folder the batch file doesn't know about. In your batch file try adding the full path to WavAviMux.exe. You might have to put the whole path+exe in "".

[Thunderbolt8]

How big is your 1.16 -down16 FLAC file in comparison to the 1.17 one?
0.98GB: 1.16 -down16 file
2.8GB: normal 1.16
2.8GB: normal 1.17 file (both exact same size)
3.29GB: rebuilt .evo (contains only the 5.1 trueHD track)

[madshi]

both have exactly the same size
Ah, ok, that's nice.

[Thunderbolt8]

sorry, had to edit, mixed things up :P
seems like fear and loathing is a "true" 24-bit file, so the filesize didnt change, but only dialog normalization removed?

[madshi]

sorry, had to edit, mixed things up :P
seems like fear and loathing is a "true" 24-bit file, so the filesize didnt change, but only dialog normalization removed?
Yep...

[Rectal Prolapse]

madshi - wow, amazing stuff. I think you did the impossible by defeating dialnorm. Congratulations and I can't wait to try it out (when I find more titles with TrueHD!)

Thanks for such a great tool!

[madshi]

I think you did the impossible by defeating dialnorm.
Well, it was not impossible, but really difficult. Just finding and changing the dialnorm field in the TrueHD headers was not even the problem. The main problem was that the TrueHD headers are protected by a checksum (CRC). So I had to find out how this CRC is calculated. It's quite hard to figure such stuff out without having an official documentation...

[TheSof]

I'm having some trouble with the latest version. With the previous version, I went .thd => flac, no problems except for the dial norm.

Now, the same .thd file throws "The source file format is unknown".

[markrb]

Is there any benefit to using the TrueHD source over the DDP if my final mux source will be as an ac3 file?

Thanks,
Mark

[madshi]

I'm having some trouble with the latest version. With the previous version, I went .thd => flac, no problems except for the dial norm.

Now, the same .thd file throws "The source file format is unknown".
Yeah, I noticed this problem, too. Sometimes demuxed TrueHD files have some "garbage" bytes in front of the real data. Don't know why. eac3to doesn't like this. The next version will fix that.

[madshi]

Is there any benefit to using the TrueHD source over the DDP if my final mux source will be as an ac3 file?
When encoding and especially when reencoding it's the best solution to always use the highest quality source you can get. So the question is: Do TrueHD files have a better audio quality compared to DDP files? Well, that question may be harder than you think. Most people would say yes. However, DDP files are 24bit while most HD DVD TrueHD tracks are only 16bit. It's not totally clear whether DDP files with 1.5Mbps sound better or worse than 16bit TrueHD tracks. The insiders don't really agree there, either.

What is clear is that if the TrueHD track has more than 16bit, it's definitely the way to go. And if the DDP track has less than 1.5Mbps then the TrueHD track is also the way to go. If the TrueHD track is 16bit and the DDP track is 1.5Mbps then I don't know what to recommend...

[madshi]

eac3to v1.18 released

http://madshi.net/eac3to.zip

* bugfix: some TrueHD files were not accepted ("The source file format is unknown")
* change: EVO files are not accepted as source files, anymore
* added: detection and repacking of 16 bit TrueHD tracks
* added: proper detection of "DTS-HD Master Audio" and "DTS-HD High Resolution" tracks
* added: runtime information for "DTS-HD High Resolution" tracks
* bugfix: bitrate information for "DTS-HD High Resolution" tracks
* added: decoding of "DTS-HD Master Audio" tracks (Sonic)
* added: decoding of "DTS-HD High Resolution" tracks (Sonic)
* added: decoding of conventional DTS tracks (Sonic/Nero)
*** full DTS-HD decoding support ***

I think this is another very big release because of full "DTS-HD Master Audio" and "DTS-HD High Resolution" support! But actually I've not done much to make that work. It's just that I found out that the Sonic Audio Decoder can already decode DTS-HD just fine! So I just added automation for that.

Please reread the first post of this thread. I've added/corrected a lot of information there.

Basically a quick summary would be: "DTS-HD Master Audio" decoding should work bit perfectly (I cannot check if it's really bit perfect, though, I'm just guessing from what I can see). So reencoding to FLAC would make a lot of sense. "DTS-HD High Resolution" decoding works mostly fine, too (details see first post). Of course the big question with "DTS-HD High Resolution" is: Into which format should we reencode this to (especially with 24bit tracks)? Reencoding to FLAC would greatly increase file size. Reencoding to anything else doesn't make much sense. Instead we could just extract the DTS core and be done with it. Fortunately most DTS-HD tracks from HD DVD and all DTS-HD tracks from Blu-Ray are lossless (DTS-HD Master Audio). And for those FLAC reencoding should be perfect.

There's one change for TrueHD decoding: Basically eac3to now checks (after decoding) whether the decoded data contains only 16 bits of information or more than that. If there are only 16 bits of information in the raw 24 bit file, eac3to strips the zero bytes and reduces the raw file to 16 bit. This was not really all that necessary. I've compared and the space saving when encoding in FLAC is only about 0.1 percent. However, I just found it cleaner to have 16bit TrueHD tracks encoded in 16bit FLAC instead of 24bit FLAC. Anyway... Interestingly, I also compared DTS and AC3 encoding and the resulting DTS and AC3 files were bit identical, regardless of whether I stripped the zero bytes or not! Huh, I didn't expect that... :) There's a big saving if you reencode a 16bit TrueHD file to WAV, of course. But who does that? :)

Here are some quick checks with DTS-HD decoding/reencoding:

Eragon Blu-Ray, DTS-HD Master Audio, 24 bit:
- runtime 00:09:15
- 1.5Mbps dts core: 100.0 MB
- dtshd file: 267.5 MB
- reencoded to 24bit FLAC: 230.2 MB

Pan's Labyrinth HD DVD, DTS-HD Master Audio, 16 bit:
- runtime 00:09:15
- 1.5Mbps dts core: 100.0 MB
- dtshd file: 128.5 MB
- reencoded to 16bit FLAC: 77.5 MB

Perfume HD DVD, DTS-HD High Resolution, 16 bit:
- runtime 00:09:15
- 1.5Mbps dts core: 100.0 MB
- dtshd file: 134.9 MB
- reencoded to 16bit FLAC: 110.2 MB

I think these are very interesting results! It's clear to see that 24bit lossless compression (Eragon) comes at a big price. DTS-HD Master Audio compression compares "ok" to FLAC here. With 16bit lossless compression FLAC even beats the 1.5Mbps DTS core in file size! DTS-HD Master Audio compression again works "ok" compared to the DTS core. But it's quite clear that for 16bit movie tracks lossless compression is actually a great idea! It's just too bad that lossless compression works noticably worse as soon as you go above 16bit.

It should be added that lossless compression efficiency depends a lot on the audio data. If there's not much happening in the audio track, lossless compression can work very well. You can compare that to how zip behaves: With text files zip reaches extreme compression ratios. But try to compress an EXE file and the compression ratio goes noticably down. The same thing is true with lossless audio compression (FLAC, TrueHD, DTS-HD Master Audio). In comparison the lossy codecs (AC3, E-AC3, DTS and DTS-HD High Resolution) always consume the same space. The audio track file size only depends on the constant bitrate and on the movie runtime. So depending on the movie lossless codecs can sometimes beat lossy codecs in file size. That's especially true for 16bit tracks. But as soon as you go above 16bit, the situation changes dramatically because lossless compression is just much less efficient for more than 16bit (because there's more noise in the additional bit and noise is difficult to compress losslessly).

[hristoff2]

Oh snap! :devil: :)

btw. do we need to use Sonic 4.2 Audio Decoder or 4.3 for DTS-HD decoding? (I remember sth about 4.3 decoding TrueHD badly - but at least it can decode it, so they updated sth)

[madshi]

I'm not sure, personally I'm using 4.3.

[Thunderbolt8]

so I guess apart from that dts-hd high resolution, which is inferior to dts-hd master anyway and also only rarely used, eac3to can convert anything else from blue-rays/hddvds bit per bit perfectly now (to flac)?

:thanks:

[Taktaal]

Madshi I downloaded and used that new EAC3TO you offered for download and now it has worked the file properly and created the WAV. file and then even took it to the 5.1 multi channel WAV. outputs files but it crashes yet again afterwards saying that WavAviMu is not reconised as an eternal command but that's what im supposed to use next. If you look about 2 pages back you should see the error within my bat. command file when it saids it cannot reconise it. Any suggestions please!? (Im so close :))

Which codec did you use to convert the DDP? Nero or Sonic?
How did you fix the 0 byte expected error?

[madshi]

so I guess apart from that dts-hd high resolution, which is inferior to dts-hd master anyway and also only rarely used, eac3to can convert anything else from blue-rays/hddvds bit per bit perfectly now (to flac)?
Yes, I think so... :)

As I mentioned before, I've no way to check if the Sonic DTS-HD Master Audio decoding is bit perfect, but it seems like that. At least there's definitely no dialnorm and no DRC applied. And the Sonic decoder definitely gives out a different result for the DTS-HD Master Audio track compared to if you only feed it the DTS core of the same track. So it looks like being bit perfect.

[Thunderbolt8]

tested the trueHD decoding once again with fear & loathing.

3.10 GB = demuxed trueHD track
5.71 GB = raw file size
2.80 GB = FLAC file size (cant tell though if size was byte
identical with the one produced yesterday with v1.17)

"this track contains more than 16-bit of information" so we have a true 24-bit (or at least 20-bit or something like that) track here on the fear & loathing HDDVD :)

[madshi]

tested the trueHD decoding once again with fear & loathing.

3.10 GB = demuxed trueHD track
5.71 GB = raw file size
2.80 GB = FLAC file size (cant tell though if size was byte
identical with the one produced yesterday with v1.17)

"this track contains more than 16-bit of information" so we have a true 24-bit (or at least 20-bit or something like that) track here on the fear & loathing HDDVD :)
Good. Looks like things are working as expected... :)

[Thunderbolt8]

hm is it possible that the 24-bit FLAC from truehd 5.1 is very slightly out of sync? the one I created for fear & loathing now seems to be a tiny bit, I also believe it was a little better with the former versions. I muxed the flac into .mka and then muxed this together with the video with a 23.9760239 fps timecode .txt file into .mkv
I tested it with ffdshow and also the WMVideo DMO decoder (which uses less cpu performance, because its distributed to both cores; cpu usage for each core is <50 % according to task manager), same result for both (using coreflac btw.)

[madshi]

I've also noticed that I needed to add 200ms delay to my main test FLAC. I'm not sure why that is the case. I'm very sure that this is not a bug in eac3to. I guess that demuxing the TrueHD track loses timecode information which results in this slight audio delay. I don't think it's a big issue, though. Of course we'll need to check and eventually correct audio sync for all TrueHD -> FLAC files and probably also all DTS-HD -> FLAC files now. Oh well, probably that means that I need to write a tool which can apply delays to FLAC tracks! Shouldn't be difficult, fortunately. Maybe sooner or later we'll find a way to calculate the necessary delay? That would be more comfortable, of course.

So everybody who's using the new TrueHD/DTS-HD -> FLAC reencoding: Please check which delay you need to apply and try to find out whether this delay seems to relate somehow to the EVO timestamps (which you can read out with EvoDemux). If we can find out how to calculate the delay, that would be great!

[Thunderbolt8]

im not that familiar with all that technical stuff, how can I manage to find out the evo timestamps with evodemux?
and since you only mentioned truehd and dts-hd now, could this delay thing also happen when remuxing eac3 tracks to flac?

as it seems the delay needed for fear & loathing also needs to be between 175 and 225 ms

[madshi]

Check out the "PTM" of the first video frame and compare that to the "First PTS" of the TrueHD/DTS-HD/E-AC3 audio track in EvoDemux. If these are identical, there should be no audio delay necessary. If they differ, we might need an audio delay. This should be valid for any demuxed audio track, I think. EvoDemux should show the necessary delay. However, for negative delays EvoDemux doesn't show the correct value but a very big positive number instead.

P.S: Or maybe the "First PTS" of the audio track is the delay we need - independently of what the "PTM" of the first video frame sais?

These PTM and PTS values are hexadecimal values. That means: Start calc.exe, switch it to hexa mode, then enter that hex number. Then convert to decimal. Now divide the decimal number by 90 and you get milliseconds. E.g. hex "000034C8" is decimal "13512", divided by 90 is about "+150ms".

[Thunderbolt8]

values are identical for the video stream and all audio tracks on that disc. I ran offsetpts before rebuilding the video evo, but it told me that offset was fine and didnt need to be processed

edit: wait, lol, this does only apply for the 1st evo file. for the 2nd evo file theres a difference, mom


Opening file FEATURE_1.EVO

PTM of first video frame = 00000D8E
Dolby TrueHD audio stream 0 found!
First PTS = 00000D8E



Opening file FEATURE_2.EVO

PTM of first video frame = 125C7D0A
Dolby TrueHD audio stream 0 found!
First PTS = 125C438F (+3422589ms)

(3 of the 4 audio tracks of the 2nd evo have a different delay btw, only the delay of the 2 DD+ 2.0 tracks is the same)

125C7D0A - 125C438F = 397B = 14715 : 90 = 163,5 <-- the delay needed ? (if yes, for the 2nd .evo file only though)


btw. I still have a remux with the v1.17 created flac file from that trueHD track and the sync issue there in the 2nd half of the movie is NOT present. so I guess it must be some little mistake which got in when you made the latest changes maybe.

watched the same scene over and over again, still, the 1.17 version is either complete identical with the evo or has a delay of <20ms or such, which I cant really spot any more. at least I believe they are identical and other thoughts might as well just come from having this watched already too often now. the 1.18 version however, although I set the delay to the above calculated 163ms in mkvmerge now, is STILL A LITTLE out of sync. it has gotten better compared to a remux without any delay before or compared to my attempts to sync it manually, but its still a bit noticeable, compared to the original evo and the 1.17 remux.

[Roscoe62]

This is probably a tiny bit off topic, but I haven't been able to figure out how to find the "right" audio track to extract when I use xport. I know in the command line you have to specify which program number, video track & audio track you wish to extract, but how do I determine which ones are correct?

I'm guessing there's a method to it, but after searching around the forum I still haven't managed to find the answers.

Sorry for the noob question, but with eac3to doing such a fantastic job of the audio (thanks again Madshi! - eac3to is turning into a regular swiss army knife for audio!) my lack of understanding of how to correctly use xport is the only thing slowing me down.

Can anyone point me to a guide, or help me to find what I'm looking for?

Thanks!

[Thunderbolt8]

well yes, it would be nice to have a counterpart to evodemux just for blue ray :P

[Sephiroth0000]

Which codec did you use to convert the DDP? Nero or Sonic?
How did you fix the 0 byte expected error?


Taktall all I done was loaded the new EAC320 up to the folder that contained all the video, audio, grf and avi files and then loaded up the EAC320GUI (must have EAC320 file there to use properly) It actually does the convert for you. I have downgraded my Nero 8 to Nero 7.10.0.1 with the HD BluRay plugin aswell so I do not know if that makes a difference. Oh and also EAC320 GUI lets you force the encoder use so you can use Nero directly which is good to do if you ask me. Oh and WAVAVIMUX you must install it to the folder you are working from NOT WINDOWS MEDIA ENCODER.

[Sephiroth0000]

Im really coming close to getting this done properly everyone (EVO to WMV HD with 5.1) And when I do I shall be adding my own tutorial on how to do it and it will be dummy proof :)
(YES even I got stuck on doing this ha, ha, ha

[Thunderbolt8]

another thing, I tried to install the illiminable_flac decoder to see if hes better/faster than the coreflac decoder, using this download link here:
http://rapidshare.com/files/9974707/illiminable_flac_0.73.1936.exe.html

but after the installation & restart of windows I still dont seem to be able to use it, at least I cant find it as external filter in mpc. did I do something wrong or am I just blind?

[madshi]

edit: wait, lol, this does only apply for the 1st evo file. for the 2nd evo file theres a difference, mom
That shouldn't matter.

btw. I still have a remux with the v1.17 created flac file from that trueHD track and the sync issue there in the 2nd half of the movie is NOT present.
Wait a moment! So your sync problem is only in one half of the movie? I think you didn't say that before, or did I miss that? For me, after using a delay the sync works throughout the full movie.

watched the same scene over and over again, still, the 1.17 version is either complete identical with the evo or has a delay of <20ms or such, which I cant really spot any more.
That's really very strange... I can't see why 1.17 and 1.18 should behave differently!! You did use 1.17 with the *demuxed* TrueHD track, didn't you? Or did you feed the TrueHD evo file to 1.17? That would explain the difference...

the 1.18 version however, although I set the delay to the above calculated 163ms in mkvmerge now, is STILL A LITTLE out of sync. it has gotten better compared to a remux without any delay before or compared to my attempts to sync it manually, but its still a bit noticeable, compared to the original evo and the 1.17 remux.
The calculated delay is all wrong. The timecode values of the 2nd EVO file have no meaning.

[madshi]

This is probably a tiny bit off topic, but I haven't been able to figure out how to find the "right" audio track to extract when I use xport. I know in the command line you have to specify which program number, video track & audio track you wish to extract, but how do I determine which ones are correct?
Personally, I'm using TsRemux to see which audio tracks are in the m2ts files in which order. The first audio track is usually number 1, the 2nd audio track number 2 etc. The I use xport for demuxing, using the numbers I've taken from TsRemux.

[Thunderbolt8]

after watching again, I'd say the delay problem with the 1.18 file also happens in the 1st half of the movie. before I just didnt bother to check it properly enough, because I though it would only apply to the 2nd (evo) half, but as it seems there is also a slight delay present in the 1st half with no delay entered at the muxing stage.

and yes, I used a rebuilt evo (that audio track only) for the 1.17 muxing, because since remuxing and rebuilding both takes ~same time and when demuxing he would have to remux it temporarely into an evo file anyway again for the flac conversion, I thought I just save myself that time. just not possible any more in 1.18 :P

dont have further time to test things throughout the day today as it seems. maybe a bit more time in the evening but cant say that for sure.

[ACrowley]

another thing, I tried to install the illiminable_flac decoder to see if hes better/faster than the coreflac decoder, using this download link here:
http://rapidshare.com/files/9974707/illiminable_flac_0.73.1936.exe.html

but after the installation & restart of windows I still dont seem to be able to use it, at least I cant find it as external filter in mpc. did I do something wrong or am I just blind?

Thats correct..the Decoder is not listed in Directshow. Only the native FLAC Source Filter.
But any Dshow Player should select the Decoder without Problmes.

Note :
illiminable FLAC Decoder cat handle FLAC in mkv container!
So use ffdshow for FLAC in Ciontainer and illiminable for single FLAC Files

[Thunderbolt8]

ffdshow is too slow as I have noticed. though I can still play any 1080p, with ffdshow as it seems theres a little delay sometimes, which is not present when using coreflac. and because of that conversion problem here with the flac delay I thought maybe coreflac is too slow either and tried to find a faster decoder for .mkv files with mpc.

[ACrowley]

ffdshow is too slow as I have noticed. though I can still play any 1080p, with ffdshow as it seems theres a little delay sometimes, which is not present when using coreflac. and because of that conversion problem here with the flac delay I thought maybe coreflac is too slow either and tried to find a faster decoder for .mkv files with mpc.


You should use only the ffdshow AudioDecoder for FLAC in mkv...i wasnt talking about the VideoDecoder.
Thats another Thing...

[Thunderbolt8]

thats what I mean, the audiodecoder :P
im using ffdshow video anyway, but when also using ffdshow audio to decode flac sometimes a little audio delay can occur, which is not present with coreflac.

[ACrowley]

thats what I mean, the audiodecoder :P
im using ffdshow video anyway, but when also using ffdshow audio to decode flac sometimes a little audio delay can occur, which is not present with coreflac.


I never had any Delay with ffdshow Audio decoder.
Try to use 16bit output...dont think it use lot of CPU usage and a DualCore can handle it
I dont like Coreflac cause it wont connect to AC3 Filter. I use AC3Filter in the decoding chain fro flac to get AC3 640 Output ( which sound s still great)

@Madshi
Great Work!!
TrueHD decoding on demuxed Files and without DialogNorm !! Wow!
And full dts hd decoding...wonderfull!

But ive Problems with the dtshd Tracks Perfume HDDVD. eac3to shows the Information and stops...the dtshd must be little bit corrupted .
I demxued both dtshd via Graphedit beacause there where Problmes with evodemux. The extracted core was cracking, you know...
So eac3to 1.18 stops decoding bad dtshd Tracks too ?

[Sephiroth0000]

Do not know if anyone can lend a helping hand to this but I am getting the following error with my Avisynth video only file when I try to play it

DIRECTSHOWSOURCE: GRF file does not have a compatiable open video pin. Graph must have 1 output pin that will bid RGB24, RGB32, ARGB, YUY2 or YV12

Now apart from this error I have mastered everything with regards to making a HD DVD evo image into WMV HD with 5.1 (Yes the audio has been mastered) Someone help !

[Taktaal]

Has anyone actually been able to get the Sonic Decoder to work with an eac3to version after 1.12? After reading this thread and the older EVOdemux one, I think that's when the support broke.

[Sephiroth0000]

Has anyone actually been able to get the Sonic Decoder to work with an eac3to version after 1.12? After reading this thread and the older EVOdemux one, I think that's when the support broke.


Taktall what is it you are trying to do? I assume it is with Audio maybe I can help.

[ACrowley]

Do not know if anyone can lend a helping hand to this but I am getting the following error with my Avisynth video only file when I try to play it

DIRECTSHOWSOURCE: GRF file does not have a compatiable open video pin. Graph must have 1 output pin that will bid RGB24, RGB32, ARGB, YUY2 or YV12

Now apart from this error I have mastered everything with regards to making a HD DVD evo image into WMV HD with 5.1 (Yes the audio has been mastered) Someone help !

Set :

HKEY_LOCAL_MACHINE\SOFTWARE\Sonic\CommonMPEGDecoders\4.2\VideoDecoder

AllowAllRenderers to 1

Then you can use the VideoDecoder 4.3 as usual via Directshowsource in Avisynth :)

[madshi]

after watching again, I'd say the delay problem with the 1.18 file also happens in the 1st half of the movie. before I just didnt bother to check it properly enough, because I though it would only apply to the 2nd (evo) half, but as it seems there is also a slight delay present in the 1st half with no delay entered at the muxing stage.
Ok, that's actually good because that means that probably the audio is in sync throughout the whole movie, as soon as you apply the correct delay to the FLAC file. It's the same with some E-AC3 files. Not all of them are automatically perfectly in sync, either. If audio sync was correct in one half of the movie, but out of sync in the other half, that would be a much tougher problem to crack!

and yes, I used a rebuilt evo (that audio track only) for the 1.17 muxing, because since remuxing and rebuilding both takes ~same time and when demuxing he would have to remux it temporarely into an evo file anyway again for the flac conversion, I thought I just save myself that time. just not possible any more in 1.18 :P
The main reason why I removed EVO input support in 1.18 was because the dialnorm removal only works with demuxed TrueHD files. So the TrueHD track you converted with 1.17 still has dialnorm applied. When creating v1.18 I noticed that I didn't clearly say that EVO input should no longer be used. And even if I documented that in 1.18 I feared that some people wouldn't read the documentation properly. So I just removed EVO input to force all people to use the proper way to have dialnorm defeated.

Sorry, my fault, should have clearly said that EVO input is not recommended, anymore (when releasing v1.17).

[madshi]

But ive Problems with the dtshd Tracks Perfume HDDVD. eac3to shows the Information and stops...the dtshd must be little bit corrupted .
I demxued both dtshd via Graphedit beacause there where Problmes with evodemux. The extracted core was cracking, you know...
So eac3to 1.18 stops decoding bad dtshd Tracks too ?
I'm not sure. I think 1.18 should just complain about the dirty track, but try to continue. Maybe the audio decoder stopped? How does the full text output look like?

Anyway, you can avoid to have a corrupted audio file in the first place. EvoDemux has a bug with Perfume. I fear if you have rebuilt the DTS-HD tracks into a separate EVO file you might already have corrupted the audio files. Try demuxing the original EVO file by using drmpeg's EVO demuxer. Here's the download link:

http://www.w6rz.net/evob_demux.zip

Using that instead of EvoDemux for Perfume results in perfectly clean DTS-HD tracks. Unfortunately Perfume only has DTS-HD High Resolution tracks encoded from a 16bit master. So the Sonic audio decoder forcefully downconverts to 16bit. I'm not sure which is better: DTS-HD High Resolution downconverted to 16bit (Sonic) or just the core, but with full 24bit (Nero).

[Thunderbolt8]

The main reason why I removed EVO input support in 1.18 was because the dialnorm removal only works with demuxed TrueHD files. So the TrueHD track you converted with 1.17 still has dialnorm applied. When creating v1.18 I noticed that I didn't clearly say that EVO input should no longer be used. And even if I documented that in 1.18 I feared that some people wouldn't read the documentation properly. So I just removed EVO input to force all people to use the proper way to have dialnorm defeated.

Sorry, my fault, should have clearly said that EVO input is not recommended, anymore (when releasing v1.17).
just to sum up the status quo: this means we can only use demuxed trueHD tracks, because otherwise that normalization removal cant work. but then we have that flac delay issue. so when seeing it right, the way to get it all right now is to wait with any conversion until you figure out how it will be right? ;)

[Sephiroth0000]

Set :

HKEY_LOCAL_MACHINE\SOFTWARE\Sonic\CommonMPEGDecoders\4.2\VideoDecoder

AllowAllRenderers to 1

Then you can use the VideoDecoder 4.3 as usual via Directshowsource in Avisynth :)

ACrowley I think I may have my AVIsynth wrong. This is what the tutorial told me to put in my Video Only.avs file

Directshowsource ("c:\HD\Video_Only.grf", fps=23.976, audio=false, seekzero=false, seek=true, framecount=196142)

I take it I am supposed to have something there in place of DIRECTSHOWSOURCE thing right? What do I put there mate or is it right? I am using Sonic 4.3

[Sephiroth0000]

Oh and also how do I set it please? Like I said im quite new to this. What do I do withh this HKEY thing? Oh and im using Sonic 4.3 not Sonic 4.2

[Thunderbolt8]

go to the command line and type 'regedit'
Oh and also how do I set it please? Like I said im quite new to this. What do I do withh this HKEY thing? Oh and im using Sonic 4.3 not Sonic 4.2
doenst matter

[madshi]

just to sum up the status quo: this means we can only use demuxed trueHD tracks, because otherwise that normalization removal cant work. but then we have that flac delay issue.
Correct.

so when seeing it right, the way to get it all right now is to wait with any conversion until you figure out how it will be right? ;)
NO.

I don't know why you're so obsessed with the "delay issue"? Having to add a small positive or negative delay to an audio track is "everyday business" for any serious reencoder. It's a very usual thing and not really an issue. The same problem sometimes (not always) also occurs with E-AC3 tracks. And you'll have the same problem everytime you try to mux a DVD audio track to a HDTV broadcast.

The only real issue right now is that delaycut doesn't support FLAC and it probably never will. However, the very next eac3to version will be able to "delay" FLAC files. You'll still have to figure the correct delay value out yourself manually, though.

Maybe sometime in the future we'll find a way to automatically determine which delay value is needed (for TrueHD, DTS-HD and also for E-AC3 tracks). But right now I don't know how to do that. But as explained above, I don't consider this as an important issue. We can handle the situation just fine today. It just needs a very little bit of additional manual work.

[madshi]

eac3to v1.19 released

http://madshi.net/eac3to.zip

* bugfix: still some TrueHD files were not accepted ("The source file format is unknown")
* added: FLAC supported as source/input file format now
* added: full delay functionality
If you want to delay a FLAC audio track by 200ms, you can now do this:

"eac3to source.flac dest.flac 200ms"

The FLAC track will then be decoded, the delay will be applied on the raw decoded audio data, and then the final raw audio data will be reencoded with FLAC again. Since FLAC is a lossless decoder this is like unzipping, changing a text file and zipping it again. There's no loss in audio quality doing it this way because FLAC is lossless. The one little disadvantages of this delay technique is that a full decode and reencode is necessary which of course costs time.

Delay also works for any other audio format, as long as decoding or encoding is involved. E.g. you can apply a delay when converting TrueHD to FLAC. However, eac3to's new delay functionality doesn't work without reencoding. If you want to delay AC3, E-AC3 or DTS files, your obvious choice is still the "delaycut" tool, of course.

[Thunderbolt8]

I don't know why you're so obsessed with the "delay issue"? Having to add a small positive or negative delay to an audio track is "everyday business" for any serious reencoder. It's a very usual thing and not really an issue. The same problem sometimes (not always) also occurs with E-AC3 tracks. And you'll have the same problem everytime you try to mux a DVD audio track to a HDTV broadcast.
its not that im obsessed with delay, I just want to be able to have it as perfectly as possible. the problem with delay is just once you know its not 100% accurate you'll notice it throughout the film and this is really anyoing. and especially with movies where mimic when speaking and quickly numbled stuff occurs throughout the whole movie in extreme ways syncing is a real nightmare, because one scene it looks fine and in the other one it doesnt any more. manually applying a delay when muxing is not a problem, but knowing a delay for such movies is almost impossible. I already synced a king kong broadcast with something like 23.9755 fps or something like that, because a static delay didnt help me,so I had to find out how to alter the video fps to come closer to the right delay and it all took hours to find out. its quite good now, but still not perfect towards the end when watching real closely.
therefore, if a way to calculate the exact delay exist, it would be an enormous help.

[madshi]

its not that im obsessed with delay, I just want to be able to have it as perfectly as possible. the problem with delay is just once you know its not 100% accurate you'll notice it throughout the film and this is really anyoing.
I see no reason why it shouldn't be possible to get it 100% accurate. Ok, maybe only 99%. But it should be possible to get the sync so near that it looks perfect to our eyes/brain.

I already synced a king kong broadcast with something like 23.9755 fps or something like that, because a static delay didnt help me,so I had to find out how to alter the video fps to come closer to the right delay and it all took hours to find out. its quite good now, but still not perfect towards the end when watching real closely.
Well, yes, trying to sync an audio track which doesn't want to fit even after a lot of work can be a royal pain in the a**. But this should happen with TrueHD/DTS-HD/E-AC3. These tracks should only need one specific static delay. And with the right delay sync should be perfect throughout the whole movie. So I do not consider this as a problem.

therefore, if a way to calculate the exact delay exist, it would be an enormous help.
It would be more comfortable. But IMO it's not that important. As I've already mentioned several times, with some movies E-AC3 tracks are not in sync, either, and need manual fixing. And nobody has even complained about that yet.

[Thunderbolt8]

well its REALLY more comfortable, when that attempt takes up several hours already. depending on how big the delay is and how much you think that video&audio is trying to fool you right now you need to remux quite often, because the mpc audio delay tool only works up to a certain limit and if thats not enough you need to remux the whole file with a delay again and try again with that mpc setting. and when doing this with disc remuxes that take up >15gb of space it begins to take a lot of time. I wouldnt say it when it only took like 15 mins altogether
I just muxed the stupid fear & loathing thing for the Xth time now, from like 50 to 500 ms and delay is almost the same, when watching (cpu usage of both cores <50%) and it seems complete oddly :(

I just added -200ms delay right now just to test and the delay hasnt really changed a bit compared to the 500ms. theres something wrong :S
added 1000ms. same delay as with 500 and -200 :/ whats wrong? could there be a problem, because I muxed the flac into .mka before muxing it all to .mkv?

[madshi]

well its REALLY more comfortable, when that attempt takes up several hours already.
It usually takes me only some minutes to figure the correct delay value out. I think you need to improve your workflow... :)

If you call MPC with the "dub" parameter, you can check delay with the external FLAC file!!! No need to do *any* muxing for delay checking.

I just muxed the stupid fear & loathing thing for the Xth time now, from like 50 to 500 ms and delay is almost the same, when watching (cpu usage of both cores <50%) and it seems complete oddly :(

I just added -200ms delay right now just to test and the delay hasnt really changed a bit compared to the 500ms. theres something wrong :S
added 1000ms. same delay as with 500 and -200 :/
Sounds strange. Does eac3to claim to apply the delay? Try adding 10000ms and check if the FLAC runtime really gets 10 seconds longer.

[Thunderbolt8]

If you call MPC with the "dub" parameter, you can check delay with the external FLAC file!!! No need to do *any* muxing for delay checking.
I know, but as I said that works only for sure for quite small delay values. can get tricky already >500 or sometimes >200 ms

[madshi]

eac3to v1.20 released

http://madshi.net/eac3to.zip

* bugfix: some Blu-Ray TrueHD tracks were not accepted
* change: eac3to output text slightly improved

[Thunderbolt8]

eac3to v1.20 released
*canceling current remuxing*

trying to do a "fresh" demuxing and conversion to flac again. maybe it helps

[madshi]

I know, but as I said that works only for sure for quite small delay values. can get tricky already >500 or sometimes >200 ms
I've no trouble with even e.g. 20 seconds of delay. With bigger delays it helps to write the delay into the "Audio Switcher" -> "Audio Time Shift" edit box and then to restart MPC. From there you can still do smaller delay changes with the keypad "+" and "-" keys. I never need more than a few minutes to find the right delay value - as long as only one static delay value is needed. Of course it takes much MUCH longer if sync keeps drifting away in the middle of the movie.

[Thunderbolt8]

hm Ive always only done it via the audio switcher -> audio time shift box and it only works for shorter delays for me. otherwise everything just get sloooooouuwww. thats why I need remuxes sometimes.
but you mean you can actually change the delay with + and - keys WHILE watching?

edit: LMAO didnt know that! omg all the hours I spent with remuxing and manually tpying etc.:S

hm regarding the external flac, I have the flac file given the same filename as the .mkv file has (apart from the .flac ending), but I cant choose it as external audio tracks. does this only work when I put the flac into .mka container?

[madshi]

hm Ive always only done it via the audio switcher -> audio time shift box and it only works for shorter delays for me. otherwise everything just get sloooooouuwww. thats why I need remuxes sometimes.
but you mean you can actually change the delay with + and - keys WHILE watching?

edit: LMAO didnt know that! omg all the hours I spent with remuxing and manually tpying etc.:S

hm regarding the external flac, I have the flac file given the same filename as the .mkv file has (apart from the .flac ending), but I cant choose it as external audio tracks. does this only work when I put the flac into .mka container?
You need to call MPC with the "dub" command line parameter. Otherwise MPC doesn't pick the "flac" file extension up as an external audio track. Using that "dub" parameter you can even feed MPC TrueHD and DTS-HD EVO and M2TS files as external audio tracks!

[ACrowley]

I'm not sure. I think 1.18 should just complain about the dirty track, but try to continue. Maybe the audio decoder stopped? How does the full text output look like?

Anyway, you can avoid to have a corrupted audio file in the first place. EvoDemux has a bug with Perfume. I fear if you have rebuilt the DTS-HD tracks into a separate EVO file you might already have corrupted the audio files. Try demuxing the original EVO file by using drmpeg's EVO demuxer. Here's the download link:

http://www.w6rz.net/evob_demux.zip

Using that instead of EvoDemux for Perfume results in perfectly clean DTS-HD tracks. Unfortunately Perfume only has DTS-HD High Resolution tracks encoded from a 16bit master. So the Sonic audio decoder forcefully downconverts to 16bit. I'm not sure which is better: DTS-HD High Resolution downconverted to 16bit (Sonic) or just the core, but with full 24bit (Nero).


Ok....Problem is i dont have the evos anymore from Perfume
Both DTSHD was demuxed via SonicHDDemuxer in graphedit.

eac3to Text is simple :
DTS Hi-Res, 5.1 channels, 2:28:00, 16 bits, 2082kbit/s, 48khz, dialnorm: -4dB
g:\

It stops without processing...all other dtshd tracks are working perfect

DTS HD 6.1 Discrete isnt supported at the Moment right ? On Xmen3 BluRay DTSHD 6.1 is get a unsupported Message from eac3to

Ah, and the dtscore from Perfum is 16 bit , not 24bit....

[Zelos]

Hi all,

i have something strange.
i tried to encode dtshd source ( riddick ) and get this message.


J:\Test Riddick\eac3to119>eac3to feature.dtshd test.dts -768
DTS, 5.1 channels, 2:14:31, 24 bits, 1536kbit/s, 48khz
This is already a normal DTS file.

[madshi]

Ok....Problem is i dont have the evos anymore from Perfume
Both DTSHD was demuxed via SonicHDDemuxer in graphedit.

eac3to Text is simple :
DTS Hi-Res, 5.1 channels, 2:28:00, 16 bits, 2082kbit/s, 48khz, dialnorm: -4dB
g:\

It stops without processing...all other dtshd tracks are working perfect
v1.21 will at least try to decode. Well, it does on my PC at least. However, the Sonic Audio Decoder crashes due to the corrupt file. So it doesn't really help. Most DTS tracks have no CRC, so the decoder can't check if a frame is valid or not.

DTS HD 6.1 Discrete isnt supported at the Moment right ? On Xmen3 BluRay DTSHD 6.1 is get a unsupported Message from eac3to
Try v1.21.

[madshi]

i have something strange.
i tried to encode dtshd source ( riddick ) and get this message.

J:\Test Riddick\eac3to119>eac3to feature.dtshd test.dts -768
DTS, 5.1 channels, 2:14:31, 24 bits, 1536kbit/s, 48khz
This is already a normal DTS file.
That's not really all that strange. Some DTS tracks taken from HD DVD and Blu-Ray are simple conventional DTS tracks and not DTS-HD tracks. So there's nothing you need to do. The track you have is already a normal DTS track. No conversion necessary for this one.

[Zelos]

ok thanks madshi i understand now.
but how to encode to 768k ?

[madshi]

This must be something like the twentiest release in the last few days. I will slow down soon, though. Just doing the necessary bugfixes and then I'll take a little break.

eac3to v1.21 released

http://madshi.net/eac3to.zip

* bugfix: 2 channel DTS files were not accepted
* added: DTS-ES 6.1 support
* added: DTS-HD High Resolution Matrix 5.1 support
* added: DTS-HD Master Audio 6.1 support
The discrete 6.1 DTS formats only have one additional channel, opposed to LPCM tracks who have two channels (which are identical, though). To keep everything more or less similar, eac3to is doubling the 6th channel, so that 6.1 DTS tracks end up being 7.1. Of course you can use the "-down6" parameter to limit output to 5.1.

If you find any further DTS or DTS-HD tracks which eac3to is still not accepting, please send me a small sample.

[madshi]

ok thanks madshi i understand now.
but how to encode to 768k ?
Oh, I see! Currently eac3to doesn't understand what you want. You want to reencode DTS to DTS. eac3to didn't expect that because it always things that you want the max audio quality. However, you can work around that by doing "eac3to source.dts dst.wavs". This will give you 6 mono channels. You can then manually feed them to Surcode for encoding with 768kbps. It's a bit more work to do it this way, but the end result should be fine. You should run the final DTS track through delaycut, though, to remove the zero byte padding Surcode usually applies to DTS encodes.

[Zelos]

Perfect madshi !
thanks for the help.

[Thunderbolt8]

You need to call MPC with the "dub" command line parameter. Otherwise MPC doesn't pick the "flac" file extension up as an external audio track. Using that "dub" parameter you can even feed MPC TrueHD and DTS-HD EVO and M2TS files as external audio tracks!
hm have a lot of trouble with that. tried it now with additional commands "pathname:\filename.mkv /dub pathname:\filename.flac"
but coreflac kept crashing all the time right at the start of mpc. I set it to block, it still kept crashing. I removed it as external filter and set ffdshow audio decoder to active instead, but coreflacdecoder still was active and kept crashing. only when I unregistred the coreflacdecoder.ax file via regdrop it would accept ffdshow as audio renderer. but then it made those funny clicking noises all the time, no normal sound came at all and eventually also crashed.
guess I have to stick to the remuxed .mka file when I want to find out the delay :S

[nautilus7]

Hi, i tried to convert a 5.1 dts track to mono wavs using tranzcode and eac3to, but each program gave me different size files:

1. eac3to wav (center channel) 1.116.851.756 bytes
2. tranzcode wav (center channel) 1.489.135.660 bytes

Eac3to reports no DiagNorm in the dts and i used the disable DRC option in tranzcode (don't really know if there's any DRC though). Both wavs look identical in Audacity.

[tebasuna51]

Hi, i tried to convert a 5.1 dts track to mono wavs using tranzcode and eac3to, but each program gave me different size files:

1. eac3to wav (center channel) 1.116.851.756 bytes
2. tranzcode wav (center channel) 1.489.135.660 bytes

Eac3to reports no DiagNorm in the dts and i used the disable DRC option in tranzcode (don't really know if there's any DRC though). Both wavs look identical in Audacity.

The default output for Tranzcode is 32 bits float per sample, but eac3to seems output 24 bit int. This is the difference in size.

With Tranzcode you can use the parameter /24 to obtain a equivalent output.

[madshi]

hm have a lot of trouble with that. tried it now with additional commands "pathname:\filename.mkv /dub pathname:\filename.flac"
but coreflac kept crashing all the time right at the start of mpc. I set it to block, it still kept crashing. I removed it as external filter and set ffdshow audio decoder to active instead, but coreflacdecoder still was active and kept crashing. only when I unregistred the coreflacdecoder.ax file via regdrop it would accept ffdshow as audio renderer. but then it made those funny clicking noises all the time, no normal sound came at all and eventually also crashed.
guess I have to stick to the remuxed .mka file when I want to find out the delay :S
The CoreFlac filter never worked well for me. Try this one:

http://www.free-codecs.com/download/DC-Bass_Source_Filter.htm

It works very well. However, it has two problems:

(1) Output is always only 16bit.
(2) Seeking only works in the first 2GB of the FLAC file.

It's still good enough to do syncing, though. When you're done with syncing, you can still mux the final FLAC file into some container to work around the 2 bugs of this filter.

[nautilus7]

The default output for Tranzcode is 32 bits float per sample, but eac3to seems output 24 bit int. This is the difference in size.

With Tranzcode you can use the parameter /24 to obtain a equivalent output.
:thanks:

[Thunderbolt8]

The CoreFlac filter never worked well for me. Try this one:

http://www.free-codecs.com/download/DC-Bass_Source_Filter.htm

It works very well. However, it has two problems:

(1) Output is always only 16bit.
(2) Seeking only works in the first 2GB of the FLAC file.

It's still good enough to do syncing, though. When you're done with syncing, you can still mux the final FLAC file into some container to work around the 2 bugs of this filter.
hm I need some of the last sections for fear & loathing for syncing, so it wont work for me at least for this movie. But I could just mux it into .mka additionally and then sync it from there and then apply the delay to the .flac file and delete the .mka. flac muxed in mka has still exactly the same delay as before, has it? or could there be some differences because of that putting into the container?

[madshi]

hm I need some of the last sections for fear & loathing for syncing, so it wont work for me at least for this movie. But I could just mux it into .mka additionally and then sync it from there and then apply the delay to the .flac file and delete the .mka. flac muxed in mka has still exactly the same delay as before, has it? or could there be some differences because of that putting into the container?
Yeah, that should work. I think the delay should be the same inside mka as it is outside. Well, if not, you'll find out soon enough... :)

[ACrowley]

This must be something like the twentiest release in the last few days. I will slow down soon, though. Just doing the necessary bugfixes and then I'll take a little break.

eac3to v1.21 released

http://madshi.net/eac3to.zip

* bugfix: 2 channel DTS files were not accepted
* added: DTS-ES 6.1 support
* added: DTS-HD High Resolution Matrix 5.1 support
* added: DTS-HD Master Audio 6.1 support
The discrete 6.1 DTS formats only have one additional channel, opposed to LPCM tracks who have two channels (which are identical, though). To keep everything more or less similar, eac3to is doubling the 6th channel, so that 6.1 DTS tracks end up being 7.1. Of course you can use the "-down6" parameter to limit output to 5.1.

If you find any further DTS or DTS-HD tracks which eac3to is still not accepting, please send me a small sample.

WOW ! Perfect

Until eac3to there was no To0l which decodes 6.1 dts to 7 mono waves :)

Madhsi ,i think Sonic Decoder is Refertnce Decoder ,right ?
So ,for the max Quality its perfect for all DTS decodes i think so ?
I mean it should be similar or better in Quality compared with Tranzcode or NicDTSSource or Foobar ?

Mh..i will take it for all my dts in the future :)

@zelos
Riddick HDDVD has a standard dts File...you can take without any change

@nautilus
Tranzcode use no DRC on dts decoding by default.
But for AC3 tranzcode applies DRC ,but no DialNorm.

[madshi]

i think Sonic Decoder is Refertnce Decoder ,right ?
Yes, I think so.

So ,for the max Quality its perfect for all DTS decodes i think so ?
As far as I can say: Yes.

I mean it should be similar or better in Quality compared with Tranzcode or NicDTSSource or Foobar ?
Yep. FWIW, I've compared DTS decoding with Nero, Sonic and Ac3Filter. Nero and Sonic decodes were identical to the last bit! Probably both are using the DTS reference code. Ac3Filter was too loud. I reencoded the Ac3Filter with Surcode and decoded it again with Ac3Filter. The peaks got bigger and bigger! So definitely too loud. With Sonic the volume stayed the same even with an additional Sonic -> Surcode -> Sonic step.

One thing to note: For DTS Discrete 6.1 decoding you may need to manually set OS settings to 7.1 speakers. Otherwise Sonic might only output 5.1. I'll work around that with the next eac3to build. The OS settings will then not matter, anymore.

[Thunderbolt8]

have a bit of a problem with an eac3 -> flac commentary track in media player classic. the track is DD+ 2.0 192kbps 24-bit and plays fine in an outside .mka file along the movie .mkv but when muxing it into the .mkv file and selecting it via filters -> pathnamefilename -> commentary track, the playback of both, video and audio is suddenly accelerated, the video is running with ~40fps and sound has mickey mouse voices (muxed video with timecodes to 23.9760239). again everything is fine with the normal 5.1 eac3 -> flac track (wasnt even delay needed), but as soon as I switch to the commentary track inside the .mkv this happens. tried both, coreflac and also ffdshow audio decoder, but same result.

[madshi]

have a bit of a problem with an eac3 -> flac commentary track in media player classic. the track is DD+ 2.0 192kbps 24-bit and plays fine in an outside .mka file along the movie .mkv but when muxing it into the .mkv file and selecting it via filters -> pathnamefilename -> commentary track, the playback of both, video and audio is suddenly accelerated, the video is running with ~40fps and sound has mickey mouse voices (muxed video with timecodes to 23.9760239). again everything is fine with the normal 5.1 eac3 -> flac track (wasnt even delay needed), but as soon as I switch to the commentary track inside the .mkv this happens. tried both, coreflac and also ffdshow audio decoder, but same result.
That's weird. I mean if it works inside the .mka file why doesn't it work in the .mkv file? mka and mkv are really the same. Can't explain it... :(

[Thunderbolt8]

found a little workaround for my truehd syncing problem with fear & loathing. I also demuxed the english 5.1dd+ stream and compared the length of both tracks. I guess the difference is the delay I need then for my trueHD track. at least I guess its now the same as to be observed with the original evo. still, even that seems to bit off sometimes there are lot of scenes in that movie where the delay could be +100ms more as well. if it wasnt for that eac3 track I would have had to guess forever.

[nautilus7]

@ madshi

May i suggest to add .ac3 decoding to .wav(s)? I think this would be good as eac3to can all other conversions, except this one. Then, eac3to will become the most complete audio tool.

[ACrowley]

@ madshi

May i suggest to add .ac3 decoding to .wav(s)? I think this would be good as eac3to can all other conversions, except this one. Then, eac3to will become the most complete audio tool.

yeah, but imho there are enough 100% perfect working Methods/Tools

Behappy-NicAsC3Source / Azid etc..

@Madhsi
Yep. i noticed too that AC3Filter DTS decodes are to loud! Peaks are louder compared with any other decoder.
This is not DialNorm related ..its simply to loud

[madshi]

found a little workaround for my truehd syncing problem with fear & loathing. I also demuxed the english 5.1dd+ stream and compared the length of both tracks. I guess the difference is the delay I need then for my trueHD track. at least I guess its now the same as to be observed with the original evo. still, even that seems to bit off sometimes there are lot of scenes in that movie where the delay could be +100ms more as well. if it wasnt for that eac3 track I would have had to guess forever.
With "length" you mean the runtime, I guess? I'm not sure if that is a reliable way to find out delay. Especially if the E-AC3 track is not in sync, either (which happens on some HD DVDs). But it might be a good starting point, so that afterwords only fine tuning is needed?

[nautilus7]

yeah, but imho there are enough 100% perfect working Methods/Tools

Behappy-NicAsC3Source / Azid etc..
Yes, but it would be perfect if there is an all-in-one tool. I 'm not complaining though, eac3to is very good.

[madshi]

May i suggest to add .ac3 decoding to .wav(s)? I think this would be good as eac3to can all other conversions, except this one. Then, eac3to will become the most complete audio tool.
The main problem with that is that all reference AC3 decoders are usually applying DRC when being used outside of their native player software. So if I added AC3 decoding support through Sonic's or Nero's AC3 decoder, we'd end up with DRC. Of course I could use AC3Filter or a similar open source decoder. That way I could probably get around DRC, but then that wouldn't be the reference decoder and there are lots of other tools which do it that way.

[Thunderbolt8]

With "length" you mean the runtime, I guess? I'm not sure if that is a reliable way to find out delay. Especially if the E-AC3 track is not in sync, either (which happens on some HD DVDs). But it might be a good starting point, so that afterwords only fine tuning is needed?
at least the ac3 track seemed to be in sync, cant say if he is 100%, its just too difficult for that movie.
another question, when just opening one of the 2 single .evo files and checking the sound from there, could it also be possible that theres a little delay then, so that its basically only 100% accurate when starting the complete movie 'normally'. or are the delays, when just opening the evo files seperately, also always accurate?

[madshi]

at least the ac3 track seemed to be in sync, cant say if he is 100%, its just too difficult for that movie.
another question, when just opening one of the 2 single .evo files and checking the sound from there, could it also be possible that theres a little delay then, so that its basically only 100% accurate when starting the complete movie 'normally'. or are the delays, when just opening the evo files seperately, also always accurate?
I'm not sure about that. I'm always first joining the movie and muxing it to MKV before I sync the audio tracks.

[TheSof]

Using the latest eac3to, to go from supermanreturns.thd to flac, it gave an flac with a different runtime:

thd 2:34
flac 2:47

Same with V for Vendetta, an extra 10mins.

With v1.17 (i think thats the version, the one where it didn't to 16bit, had dial norm) the runtimes were the same. I can't test it at the moment, but why would this be?

[Thunderbolt8]

I'm not sure about that. I'm always first joining the movie and muxing it to MKV before I sync the audio tracks.
I just asked because I always used that single original .evo file as syncing reference. but in case this could out of sync too, I would always try to sync my remuxes wrong then -.-

[honai]

@madshi

Well, if you added "native" AC3 decoding I might be tempted to create a GUI that incorporates all the latest features of your tool.

[madshi]

Using the latest eac3to, to go from supermanreturns.thd to flac, it gave an flac with a different runtime:

thd 2:34
flac 2:47

Same with V for Vendetta, an extra 10mins.

With v1.17 (i think thats the version, the one where it didn't to 16bit, had dial norm) the runtimes were the same. I can't test it at the moment, but why would this be?
Can't imagine that there'd be a difference between v1.17 and v1.18. Except if you fed the EVO file into v1.17 instead of the demuxed file? Please check if the 2:47 FLAC stays in sync throughout the movie (after you applied the eventually necessary static delay). Maybe the 2:47 has some extra seconds of silence at the end or beginning of the movie? Don't know...

[madshi]

I just asked because I always used that single original .evo file as syncing reference. but in case this could out of sync too, I would always try to sync my remuxes wrong then -.-
The first EVO part should be in sync. But I'm not sure about the 2nd part.

[madshi]

Well, if you added "native" AC3 decoding I might be tempted to create a GUI that incorporates all the latest features of your tool.
Are there important features missing in The_Keymaker's GUI? I've no idea, I'm always using the command line, only...

Maybe I'll add AC3 decoding if I find a way to disable DRC in a reference decoder. That would be worthwhile. But it will be a while before I invest time into that. I've spent too much time on eac3to lately.

[honai]

Are there important features missing in The_Keymaker's GUI? I've no idea, I'm always using the command line, only...

His latest version is missing some command-line parameters.

Maybe I'll add AC3 decoding if I find a way to disable DRC in a reference decoder.

Yes, that's the idea.

By the way, do you have plans to implement STDIN streaming for the source?