EVOB De/Multiplexers

[Pelican9]

When you press play, what do you have in the filter properties window of the Default Directsound Device ?
There are three pages (Audio renderer, Advanced, Audio input pin (rendered)).
Nothing changes when I press the play.

The first shows:
wformat tag: 1
nchannels: 2
nsamplepersec: 48000
navgbytespersec: 192000
nblockalign: 4
rate: 1.00

Advanced shows nothing (I mean all '-' and '0')

The last shows:
Major type: Audio
sub type: PCMaudio
Format: waveformatex: 48.000 kHz 16 bit
stereo

[Pelican9]

Feedback:

Just tried SUPRead v0.25b. Access violation when opening app doesn´t happen anymore but is still present when opening .sup file...

EVOdemux creates wrong .sup files if you uncheck any stream.
I'm working on it.

Edit:
Done.
Download the newest beta of EVOdemux and demux the subtitle again.

[JnZ]

1. Nothing, no sound, no progress bar
(With dump at the end of the graph, the progress bar is working)
2. Yes.
Hmm, strange, something wrong in Your system...that filter must "play" (when connected to DShow audio device).

We are trying to "catch" RAW data, which goes to soundcard redirecting them to the file (dump). So if you can't play this filter on soundcard, not able to "save" any raw data to file too.

[JnZ]

If sombody interested to making 5.1 OGG channel audio from DD+ (which is the best IMHO), here is guide:

1) Make graph filter: DD+ src-> Sonic Audio 4.2 -> Dump
2) Use SoX: sox -r48000 -S -t .raw -c 6 -3 -s xxx.raw -2 y:\xxx.wav
3) Use wavewizard and separate it to six mono WAVes
4) Wrote MUX file with 0-2-1-4-5-3 channel order. (OGG have different ch. order, than 5.1ch WAVe)
5) Use BeSweet and make WAVe file (test.wav).
6) Use aoTuV encoder: venc -q6 test.wav (-q switch means quality, -q6 gives average 460kbps)

Note: It's possible to use oggenc directly to RAW data: oggenc.exe -r -B24 -C6 -R48000 -q6 xxx.raw, but this create bad channel-ordered OGG.

[Darth Pinous]

There are three pages (Audio renderer, Advanced, Audio input pin (rendered)).
Nothing changes when I press the play.

The first shows:
wformat tag: 1
nchannels: 2
nsamplepersec: 48000
navgbytespersec: 192000
nblockalign: 4
rate: 1.00

Advanced shows nothing (I mean all '-' and '0')

The last shows:
Major type: Audio
sub type: PCMaudio
Format: waveformatex: 48.000 kHz 16 bit
stereo


Seems to me that the sonic audio decoder sends crap to your soundcard... Did you install Sonic 4.1 filters ?

[Pelican9]

Changes for v0.623

- Display H.264 video stream info (finished)
- Rebuild change the DSI packet's audio and subpicture information and PTM info too
- Fixed bug with MLP and LPCM stream handling during Rebuild
- Fixed bug with 'Cancelled' message
- Fixed bug with 'Invalid floating point operation' after demux
- Fixed bug with DTS HD info
- Fixed bug with .sup files
- Fixed minor bugs with opening files and closing application

EVOdemux v0.623 (http://pel.hu/down/EVOdemux.exe)


Seems to me that the sonic audio decoder sends crap to your soundcard... Did you install Sonic 4.1 filters ?
No. I've installed 4.2 only.

[Darth Pinous]

No. I've installed 4.2 only.

I've tried to install 4.2 only, but I couldn't connect Sonic with orbitlee filter in Graphedit, and I had repeated errors when closing Sonic configuration panel.

After installing Scenarist 4.1 (which doesn't affect 4.2 Sonic filters), I could connect Sonic Decoder with orbitlee filter and render DD+ files... Maybe that will work for you too ...?

[Jan2001]

I've registered scenarist 4.1 filters (as the application doesn't work with mce) and 4.2.
All i get is a 6-channel output of which 4 channels are crap.

[The1n]

Im also getting that the orbitlee dtsac3source filter cannot connect to sonic audio filter 4.2. Can someone advise me what is wrong, can it be a conflict with filters? , 4.1 connects to orbitlee but not 4.2.

The1n

[zgx]

This might not belong in this thread but here it goes anyway. Since we are discussing "Sonic Cinemaster Audio Decoder 4.2".

When I use it to decode for example a LPCM 5.1 stream I would like to preserve the same dynamic range as the original. There is a setting in Sonic Cinemaster called "environment" that adjusts dymaic range.

My question is if "Quiet" means "Full Dymanic Range = like the original" or if it artificially increases dynamic range even further. Or should I go with "Normal"?

[orbitlee]

Im also getting that the orbitlee dtsac3source filter cannot connect to sonic audio filter 4.2. Can someone advise me what is wrong, can it be a conflict with filters? , 4.1 connects to orbitlee but not 4.2.

The1n




Based on my own test and other people's feedback at here, I must say sonic decoder pack 4.2 is quite unstable. Sometimes it works, sometimes don't. No idea why.

Could you check whether this fiter graph can be created?
EVO->Sonic HD Demux->Sonic Audio decoder 4.2 -> Default DirectSound ?

[orbitlee]

Pelican9, if you got this "Advanced shows nothing (I mean all '-' and '0')", there is definitely no sound.
Are you sure you are using DD+ audio track? And, please give me a sample of your audio track, I'll try to reproduce this and find out the reason.



There are three pages (Audio renderer, Advanced, Audio input pin (rendered)).
Nothing changes when I press the play.

The first shows:
wformat tag: 1
nchannels: 2
nsamplepersec: 48000
navgbytespersec: 192000
nblockalign: 4
rate: 1.00

Advanced shows nothing (I mean all '-' and '0')

The last shows:
Major type: Audio
sub type: PCMaudio
Format: waveformatex: 48.000 kHz 16 bit
stereo

[The1n]

Based on my own test and other people's feedback at here, I must say sonic decoder pack 4.2 is quite unstable. Sometimes it works, sometimes don't. No idea why.

Could you check whether this fiter graph can be created?
EVO->Sonic HD Demux->Sonic Audio decoder 4.2 -> Default DirectSound ?

Hi Orbitlee
With the graph you suggested i got connection error 80040217 with both VC1 to the video decoder and DD+ to the audio decoder both 4.2 version.

I have also just reformatted my pc incase there were trouble in registry land , i have installed directx latest and your modded dtsac3 filter but still no go :(

Can it be something to do with the DS filters for my audigy2 zs card conflict ?

Regards
The1n

[woah!]

i just did a encode of Posidon which seemed to work just fine for sync a/v.

now what is weird is that the video runs a total time of 1:38:13 , but the audio came out at 1:38:15


so a 2 sec difference but when i muxed them together in mp4box i got a perfect sync result?? i am not a master at any of this stuff but does that make sense? heh.. anyways it worked..

i use intervideo audio decoder to process the ddp+ audio in a graph file, and feed it to behappy using a simple avs script.

[xc3ll]

i just did a encode of Posidon which seemed to work just fine for sync a/v.

now what is weird is that the video runs a total time of 1:38:13 , but the audio came out at 1:38:15


so a 2 sec difference but when i muxed them together in mp4box i got a perfect sync result?? i am not a master at any of this stuff but does that make sense? heh.. anyways it worked..

i use intervideo audio decoder to process the ddp+ audio in a graph file, and feed it to behappy using a simple avs script.

woah, what script did u use?

[Wookie Groomer]

EVODemux is an amazing tool. Is there one similar that handles m2ts files? Will EVODemux eventualy integrate that format as well?

[Darth Pinous]

EVODemux is an amazing tool. Is there one similar that handles m2ts files? Will EVODemux eventualy integrate that format as well?

You've got drmpeg's XPORT tool there (http://www.w6rz.net/xport.zip).

It can demux every stream from a m2ts file, except for PCM soundtracks which it can demux only in stereo (for now).

[noclip]

I was finally able to extract the full 5.1 audio. EVO -> Sonic HD Demux -> Sonic Audio Decoder 4.2 -> WAV Dest -> Dump. The trick is, you have to view the properties for the HD demuxer and select the right stream. Probably best to demux and join first so you don't have to deal with picking the right stream in the second file.

[woah!]

woah, what script did u use?

ok i think "i have" a repeatable way now for all my HD DVD needs.

i cant use a vc1 or mpv demuxed video stream as it seems to just jump around randomly throughout the film, as if the frames dont have a 0 1 2 3 type stream. it starts off ok then it can jump to say 25mins in and then back to 10 mins into the film etc..

anyways i solve this by just using the rebuilt EVO file directly, which then works perfect for me. here is my script for Posidon:




DirectShowSource("G:\5\PEVOB_1_PEVOB_2.rebuilt.EVO",framecount=141292,fps=23.976,audio=false)
crop(8,144,-8,-144)
Lanczos4Resize(1280,528)
Nicefps()
ConvertToYv12()



that gives me a end result at 23.976fps and 01:38:13 long.

now the audio which i just figured out what i did from the night prior so there is some sync work to do. i used graphedit and connected the ddp+ file i demuxed out of the joined EVO file to the "Intervideo Audio Decoder" which on my comp here is setup only for 2ch audio, but i dont see why 6ch cant be done if you have 5.1 seup.

PEVOB_1_PEVOB_2.DD+.6ch.stream.00.English 5.1.ddp ---> Intervideo Audio Decoder


i saved that graph out and added it to a simple avs script:

SetMemoryMax(64)
DirectShowSource("G:\5\SOUND.GRF", video=false)


i used behappy to process the script and used the timestretch plugin to change the audio from the "counted frames amount" to the "calculated frames amount"

in Posidon this was 141350 to 141292

now the bit i dont understand, this gives a time of 01:38:15 which is 2 secs slower that the video, but throw these 2 resulting files into mp4box and mux them together and you end up with a mp4 with a runtime of 01:38:15 and synced.

i have watched the result on my computer and over my mce2005 media server onto my plasma and it runs perfect on both.


so there you have it really, long winded but it works ... for me anyway heh..

[Pelican9]

Pelican9, if you got this "Advanced shows nothing (I mean all '-' and '0')", there is definitely no sound.
Are you sure you are using DD+ audio track? And, please give me a sample of your audio track, I'll try to reproduce this and find out the reason.
I've used the demuxed audio stream from UNILOGO.EVO (The Chronicles of Riddick).

[MiniMinimal]

why doesn't my *.avs script work when i'm trying to get video into vdubmod?

It says:

Avisynth open failure:
Script error: DirectShowSource does not have a named argument "framecount"
(c:\lekmedfilm\avisynth.avs, line 1)

[edo1080]

I'm trying the way with last samurai:

mpa ->Sonic HD Demux->Sonic Audio decoder 4.2 -> dump

and then I use sox on the dumped file and I obtain a 5.1 wav, but when wavewizard opens the 5.1 wav the time lenght is 51 min too short for last samurai, and when I play each of the 6 mono wavs I get they are all accelerated.

I think it could depend on thw 1st step because the total dumped file is 2.6 Gb, too small for more than 2 hours movie.

Which are the settings I have to use in graphedit with the blocks? I open the .mpa audio track with graphedit and the chain mpa->Sonic HD Demux->Sonic Audio decoder 4.2 -> default audio device should I try to do it on the evo and not on the mpa file? but I have 2 evo files for the movie while only one mpa demuxed track.
Thank you

[tomos]

why doesn't my *.avs script work when i'm trying to get video into vdubmod?

It says:

Avisynth open failure:
Script error: DirectShowSource does not have a named argument "framecount"
(c:\lekmedfilm\avisynth.avs, line 1)

show the AVS script itself mate :)

mine goes

DirectShowSource("yadda.grf",fps=23.976,framecount=155019)
crop( 4, 132, -4, -140)

and works fine in vdubmod and megui

[MiniMinimal]

Oki, sorry :)

DirectShowSource ("C:\Lekmedfilm\rebuilt.EVO",video=true,audio=false,fps=23.976,framecount=712,seek=false,seekzero=false)ConvertToYV12 ()Crop (0,N,0,-N)

I'm following the guide by: idamien
and using the unilogo.evo file

[edo1080]

I forgot to ask how to set the dump output file at the end of the chain of sonic decoder; I only found the dump filter but I cannot place it if I already placed a sonic filter. Thanks

[Deckard2019]

Did anyone try to process VC1 with last Haali splitter and WMP 11 :

http://deckstuff.free.fr/temp/vid.png

It seems to work great but I would like to be sure before start encoding ...

[Darth Pinous]

why doesn't my *.avs script work when i'm trying to get video into vdubmod?

It says:

Avisynth open failure:
Script error: DirectShowSource does not have a named argument "framecount"
(c:\lekmedfilm\avisynth.avs, line 1)

Because you don't have the correct version of DirectShowSource.dll. You need the one from Avisynth 2.5.7 to have the "framecount" argument.

[SBeaver]

Did anyone try to process VC1 with last Haali splitter and WMP 11 :

http://deckstuff.free.fr/temp/vid.png

It seems to work great but I would like to be sure before start encoding ...

it is certainly a lot smoother that ffdshows decoding, can't say anything about quality though.
common sense might make you think the ones who made the format should have the best decoder, but common sense doesn't always apply to the company in question.

Btw, is there any new decoder coming along for dd+? I'm having a hard time making it work and I rather have something that actually works in a regular player like MPC or similar.

[MiniMinimal]

Because you don't have the correct version of DirectShowSource.dll. You need the one from Avisynth 2.5.7 to have the "framecount" argument.


Ok,I had it, but vdubmod used the old ver. that i did forgot to uninstall :rolleyes:

but now it is another problem:

DirectShowSource: Renderfile, the filter graph manager won't talk to me.

:confused:

[MiniMinimal]

I tryed to demux the *.evo into mpv,
then export it to avisynth and vdub.

vdub open it, but the image is grey..
is it the video file unilogo.evo that don't contain an images?

[lineman]

it is certainly a lot smoother that ffdshows decoding, can't say anything about quality though.
common sense might make you think the ones who made the format should have the best decoder, but common sense doesn't always apply to the company in question.

Btw, is there any new decoder coming along for dd+? I'm having a hard time making it work and I rather have something that actually works in a regular player like MPC or similar.

I have done four encodes using this filter arrangement and the results are perfect. I used ffdshow for 2 encodes and found tearing at the pictures edge and picture breakup at various points through the encode. I would definatley use the DMO filter!

Lineman

[MiniMinimal]

Everything works just fine now...
i got picture! THNX!!

[SBeaver]

I'm running a movie now in MPC using the DMO decoder and Haali splitter.
There is a problem though.
I turned on OSD for haali and it says 1920x1080 frame format, but it is being scaled down to 960x540 and I can't do anything to stop that.
I tried ffdshow again and it downscales too.

Any ideas?

[kornesque]

this might be slightly off topic, but since it applies to my demuxed .evo....

does anybody here have wmvmuxer experience? i've got a perfect .avi containing my vc-1, and have a good .wav (downmixed to stereo) from the dd+, but wmvmuxer throws me errors. i'm assuming i can mux the vc-1 data and my audio into a .wmv without transcoding - does anybody have any ideas how to go about it if it's possible? thank you.

[Neo Fagin]

When running xport on an m2ts with DTSHD audio, it demuxes okay and produces the audio stream on disk, which is fine - but what I want to know is how the 1536kbps DTS core can be extracted from this file? I've searched around and not yet found a way that doesn't involve me playing the file back, saving 6 mono wavs and then building a new DTS file out of them, which involves reencoding which I'd like to avoid if possible.

[MichalHabart]

When running xport on an m2ts with DTSHD audio, it demuxes okay and produces the audio stream on disk, which is fine - but what I want to know is how the 1536kbps DTS core can be extracted from this file? I've searched around and not yet found a way that doesn't involve me playing the file back, saving 6 mono wavs and then building a new DTS file out of them, which involves reencoding which I'd like to avoid if possible.

What movie are talking about?

[Neo Fagin]

What movie are talking about?
Kingdom of Heaven (only has DTSHD audio)

I found out from someone on irc who'd done it - mux in mkvmerge and ignore the errors, mkvmerge just strips the erroneous DTSHD packets and leaves the core intact.

[MichalHabart]

Kingdom of Heaven (only has DTSHD audio)

I found out from someone on irc who'd done it - mux in mkvmerge and ignore the errors, mkvmerge just strips the erroneous DTSHD packets and leaves the core intact.

Well, even when cover says that ice age II, stargate and tomb raider do have DTS-HD tracks, all DTS-HD tracks could be demuxed with xport and are playable in foobar withou any problems. Should be the same with kingdom of heaven (i don not have it yet so i can't tell you exactly)

[xc3ll]

I have tried JuHu's and JnZ's methods for capturing audio. However, my audio will not sync. It always starts going off sync as the movie plays. This may be due to the framerate issue, but I'm not sure. Using JuHu's method, it's off by about 2.5 seconds, which is about the difference between calculated and counted frames. It seems like time stretch doesn't do anything though. I've even tried applying it twice.

Using JnZ's method, the sound is still off at the end, by the same amount. I'm currently using MKVmerge GUI 1.7.0.

EDIT: I've found that the timestretch does work. Going from the bigger number to the smaller number. However, counted and calculated frames arent working for me, so now I'm trying to guess what the 2nd number should be. Audio seems to lag by about a second at the end, so I subtracted 24 from the 2nd number. I'll post my results

[dchard]

Feedback: the problem with sup solved. Nice work.

Now I got the PNGs or BMPs. How can I convert those to .srt ?

Other question: I got a .srt subtitle. Is there any way to convert it to evobdemux compatible .sup ?

Thanks!

Dchard

[MichalHabart]

Feedback: the problem with sup solved. Nice work.

Now I got the PNGs or BMPs. How can I convert those to .srt ?

Other question: I got a .srt subtitle. Is there any way to convert it to evobdemux compatible .sup ?

Thanks!

Dchard

I have the same problem. I have PNGs and srt with line1-xxx instead of text. Is there some program that can convert it to proper srt?

[Deckard2019]

Now I got the PNGs or BMPs. How can I convert those to .srt ?
http://forum.doom9.org/showthread.php?t=121535&page=5

[Pelican9]

I have the same problem. I have PNGs and srt with line1-xxx instead of text. Is there some program that can convert it to proper srt?
OFF: We need a working OCR to convert the bitmaps to text.

[MichalHabart]

OFF: We need a working OCR to convert the bitmaps to text.

Definitely.

[edo1080]

I discovered the dump.raw file obtained with graphedit and Sonic Audio Decoder is accelerated can anyone help me please? Thank you.

[xc3ll]

I discovered the dump.raw file obtained with graphedit and Sonic Audio Decoder is accelerated can anyone help me please? Thank you.

I have the same problem. If you're like me, and its only about 1-2secs too fast at the end(if its a 2hr movie, maybe more for longer movies), then I suggest you try Juhu's method of using BeHappy to timestretch it. I'm using it right now, and although the sync isn't perfect, I'm pretty close. Right now I'm adding 12 to the second number.
http://img103.imageshack.us/img103/9057/fearlessaudioha8.jpg (http://imageshack.us)
EDIT: Image should say 1/2sec too fast.

[SBeaver]

It was the haali renderer that did the scaling, not sure why.

[Neo Fagin]

Well, even when cover says that ice age II, stargate and tomb raider do have DTS-HD tracks, all DTS-HD tracks could be demuxed with xport and are playable in foobar withou any problems. Should be the same with kingdom of heaven (i don not have it yet so i can't tell you exactly)
It's playable in graphedit if I use orbitlee's DTS/AC3 source plugin, but it's a ludicrous 5GB in size and I can't remux it into an mkv or anything else

[enantiomer]

Kingdom of Heaven (only has DTSHD audio)

I found out from someone on irc who'd done it - mux in mkvmerge and ignore the errors, mkvmerge just strips the erroneous DTSHD packets and leaves the core intact.

Another method: use tranzcode. Here's an example of converting the DTS core audio to AC3, lifted from one of my automagic BD_2_TS scripts:

e:
cd e:\bd_2_ts
rename PID_0x1100.bin PID_0x1100.dts
c:\bin\tranzcode\tranzcode PID_0x1100.dts c:\tmp\bd_pt1.wav /mch
c:\bin\aften -b 640 -dnorm 27 c:\tmp\bd_pt1.wav e:\bd_2_ts\bd_pt1.ac3
del c:\tmp\bd_pt1.wav


It does produce a rather large intermediate wave file, but conversion is pretty fast when sloshing data between two hard drives.

[orbitlee]

About the DTS-HD audio track. Actually evodemux/xport works perfectly, they demuxed DTS-HD audio track, which could be as high as 24.5Mbps . But your software cannot tell the difference between DTS-HD and DTS. It assumes this is DTS 1536kbps, and calcuate the runtime based on file length of audio track. So you will get both hugh file and terribile runtime.

Fortunately there is very simple solution to extract DTS 1536kbps from DTS-HD, no re-encoding or any professional software. Since DTS Labs wants to keep compatibility with existing DTS audio components, in every DTS-HD audio track, they encoded a DTS Digital Surround? Core (which is CBR 5.1ch, 1536kbps , with DTS frame header) and extensions(which could be CBR or VBR, 7.1ch, up to 24.5Mbps, with DTS-HD frame header). So legacy DTS device will recognize the DTS core only and ignore the extensions, DTS-HD device will recognize both DTS core and the extension to archive full benefit of DTS-HD audio track.
http://www.dts.com/dts-hd/better-sound-today-perfect-sound-tomorrow.php

Unfortunately, I found DTS labs does not prevent the start code emulation. What I'm talking about is, you may find 7FFE8001(which is sync word for DTS frame) in DTS-HD extension bitstream. If the software don't know the details of DTS-HD, it may pick the wrong sync word from DTS-HD extension bitstream, which will generate a very small glitch(you may never hear that probably). A real case is
the DTS-HD MA audio in "Ice Age: The meltdown". I tried delaycut(1.2.1.2) to process the DTS-HD audio, it will strip out the extension bitstream and generate smaller audio track. But if you process the output again, delaycut will warn you that some parameters are changed. This is because delaycut 1.2.1.2 pick the wrong sync word and then wrong DTS frame. I don't know whether mkvtoolnix or other mux/demux tools have similar issues.

So I wrote a small tool to extract DTS core from DTS-HD bitstream. Since I don't have any specs on DTS-HD extension bitstream(If I had it, I will know the exact length of extension bitstream), to avoid the sync word issue, I assume all frames in DTS core bitstream will have same frame length, sample rate and bitrate. I believe this assumption makes sense.

Command line is
dtscore input.dts output.dts

If you give it traditional DTS audio track, it will output the original untouched.

You should be able to feed the output to any legacy software/hardware(which does not know DTS-HD) without any problem.

BIG thanks to drmpeg for informations on DTS spec :-)