The release lists both Dolby and DTS as just one tracklist with the usual (5.1 mix) suffix.
Is it acceptable to extend the tracklist so that it shows both versions of the audio?
This is for tagging’s sake. I had lots of trouble extracting the audio using mplayer’s confusing syntax and I finally did it.
I also learned that a DVD-Video disc’s AC3 audio can be dumped raw by mplayer and then contained in an .m4a container using ffmpeg (no transcoding, just the container), and DTS in .mp4 container (likewise), both of which beets and Picard can tag, and MPD and other players can index into the library and play, surprisingly without issues.
So, it can be done. I don’t think many people do do it though.
Still, the questions stand: Should we index both available versions of the audio? Should the recordings be separate for both surround versions? And for video tracks, given that video containers can contain both audio streams simultaneously, what do we do?
There’s cases where this is not relevant, such as where the DVD only contains Dolby audio:
However this release contains stereo as well (I checked). Should we list both versions of the audio in the tracklist for tagging?
1 Like
It is a bit of an overkill to list both separately. Video media is more complex to fit MB due to the way video and audio can be paired. What happens when you get multi-lingual versions? You could have a documentary on the disk with five different languages available. That would not want to be listed five times.
If this is just a tagging thing, it is easier to get the tagging software to apply the titles multiple times.
Just think of the mess a boxset could look like if stereo\surround\DTS editions were all being added as separate listings to multiple discs.
DTS and Surround versions are still the same recording, just a different type of compression.
And example 7 disc box that would look a complex mess if this happened: Release group “(____surrounded):” by Björk - MusicBrainz Seven disks. One side is stereo CD, the second side is DVD with 5.1 Dolby and DTS versions. MB lists the unique recordings but not all the audio variations.
I am sure this discussion is somewhere else in the history of this forum.
SACDs are already a mess so I don’t care.
Seriously though, yes it’s for tagging. Having the codec on the filename would be helpful. beets can’t tell them apart though.
MusicBrainz really needs to implement alternate tracklists…
Yeah same example I gave at the start 
Haha - whoops. Sorry. I thought I saw it, but when scrolled back up I saw the other one.
But that is a good reason why it doesn’t make sense to do it. The recordings only exist the once. Different encoding options are not really relevant to MB. Look at Digital Media files - no way to say low res MP3 or high res FLAC.
There are people who have knocked up scripts to do this kinda thing. I don’t know if they delved as deep as number of channels \ DTS vs Dolby though. But I assume you can’t have hundreds of albums like this, so isn’t it quick enough to do them by hand? That’s the beauty of Picard. It can take the MB data and then twist it to your needs.
Don’t forget that MB is about documenting the release, not about tagging. That is just secondary.
Personally I’d like to see even a simple field added for noting the number of channels. MB has no way to show a mono apart from a quadraphonic. So the differences for the types of encoding of surround sound are currently well outside of MB.
With SACDs there is a solution there, each layer gets a separate medium. But then they have different physical layers that are being documented.
Well I already spent the time extracting each track one by one, I suppose a couple more minutes wouldn’t hurt 
why don't you use bash
I don’t know bash. I’m meaning to learn though.
I just ripped a DVD that came with an album. Then remembered Picard still can’t do MKV tags yet.
Whipped over to MP3TAG and it is pretty flexible for hacking bulk listings around. For something like this it is ideal. You can get MP3TAG to export a set of tags, then edit them in notepad, before importing them back in. Makes it a simple task to bang through complex albums like that.
In this example I let Picard rename the MKV filenames, then MP3TAG took over from there to push those filenames across the tags. And its a simple job to export that to notepad add the [DTS] bit on the end of the tags before re-importing those tags. Bit of copy\pasta work and you’d have that track list duplicated down with [dobly] on the other half of the files.
I just keep the audio. I’m collecting these for the day I finally get a surround receiver setup, so I can just fire up whatever audio player I have.
If the DVD contains just stills I don’t care about the video. If it contains video I just pop the disc in.
Anyways, MPlayer can dump the DCA audio, and FFMPEG can copy it without transcoding to an MP4 file. Same with AC3 audio and m4a files. Picard can tag all of those like regular audio files!
I’m still waiting for .mka to catch on.
I use MakeMKV on my DVDs and BluRays as they are being dropped onto a media sever. Too lazy to go get the discs out. I already have that Surround System so want to have nice easy access to the best audio on those videos. Same with the surround audio rips.
With the media player it is pretty important to have the [DTS] or [quadraphonic] text in the album and track titles to I pick the right one to play.
Picard is great for looking up in the database, but MP3Tag is where it gets much more flexible for me.
1 Like
I have a good Windows based cover DVD chapter some audio rip method.
If someone could help me find a good Linux method… 
It’s audio rip because I just want to submit release to MB. 
1 Like
Are you one of those people adding AcoustIDs to videos? Can I assume that you don’t do that from your “split by chapter” method as I assume your AcoustIDs would be unique to your method of splitting.
I’m experimenting with CUE sheets on the big concerts instead of cutting up that single video track. Just me - I like keeping things in the format I rip. 
Why would it be unique? For one the chapters are marked on the DVD, so it’ll be split at the same position if two people do this. And AcoustID when comparing fingerprints allows for some seconds shift
1 Like
Replace any values in [these] as needed
First find the audio id (dumped as aid) of the stream you want to rip:
mplayer dvd://[title] -dvd-device -vo null -ao null -frames 0 -v
Now rip:
mplayer dvd://[title] -dvd-device [dvd mount point] -benchmark -vc null -vo null -ao null -aid [audio id] -dumpaudio -chapter [start chapter]-[end chapter] -dumpfile [name of output file]
This would rip all chapters selected as one file. You’d have to pick each one manually, or use some bash magic I guess.
I tried converting on the fly using -ao pcm:waveheader:fast:file=audio.wav but that mixed down the multichannel audio to stereo.
Be careful on conversion to flac, ffmpeg switches the channels around sometimes. (I don’t convert to FLAC because neither DTS/DCA nor AC3 are lossless.)
1 Like
I will try to see if I can get player, and test:
for i in {01..20}; do mplayer … -chapter $i-$i …; done
I hope it works. I’m pulling out my hair every time I have to figure all this out again. But the basic idea is there!
You can even rip Blu-ray discs with MPlayer. Well, provided they’re not encrypted. I don’t know how that one would work. Just get the audio ID with ffprobe bluray:[blu-ray mount point] I think it was? It’d be in the form 0x00 something. Different from a DVD where the audio ID is an integer and where you can get it with mplayer -v
1 Like
Thanks for Blu-ray command but I have no Blu-ray drive on PC.
For Blu-ray tracklists with times, I do it more ma-nual-ly with the TV Blu-ray deck.
1 Like