Skip to main content ↑ ↓

Notice

Please note that most of the software linked on this forum is likely to be safe to use. If you are unsure, feel free to ask in the relevant topics, or send a private message to an administrator or moderator. To help curb the problems of false positives, or in the event that you do find actual malware, you can contribute through the article linked here.
Topic: Request For Comments (Read 3952 times) previous topic - next topic
0 Members and 1 Guest are viewing this topic.

Request For Comments

Hi,

We are currently in the process of freezing the current matroska specs (well, there's still a lot of room for additions, changes but it starts now). And therefore I'd like to make sure that we haven't ommited any information that people often store in their preferred container/tagging system.

After discussing between ChristianHJW and Case (who made the great Tag tool), we decided to post our request for comments here and collect the results to be applied to matroska.

So you are welcome to comment what you think you be in (the tags you use).

If you want to have a look at what is already there, have a look at the current specs. The "meta" informations usually stored in an audio file are spanned in the Information element and Track element.

Mmmm, please no container war or flame here. 

Request For Comments

Reply #1 –
I can already think of a BPM value for a specific part of an audio track (handled as a Chapter in matroska) and a quality associated (as used in Native Instruments Traktor DJ).

Request For Comments

Reply #2 –
Is there any way to define your own metadata entries? Or would it have to go through you then, putting it in the spec and making it official before you can use it?  Or is it just because I didn't read the proposal properly?

Request For Comments

Reply #3 –
Well, the format is extensible as much as you wish. But the idea is to have standard tags that everyone would know and support (or not).

Otherwise everyone can create their own set of tags and nothing is ever compatible. That's why we want to create a list as large as possible (but not add unused things). 

We also hope to be able to freeze the format someday and make it a stable standard. So everything needs to be handled before it's freezed.  B)

Request For Comments

Reply #4 –
Ok... So then where is the name of the artist / performer supposed to go?

Request For Comments

Reply #5 –
In the case of a simple file (1 segment and 1 audio "track" in this segment) :
Code: [Select]
Title    2    [7B][A9]    -    -    -    UTF-8    General name of the segment.
Author    2    [60][93]    *    -    -    UTF-8    Author name of the segment (company, team or individual).


In the case where you have multiple audio track in one segment (a whole album in one file/segment for example) :
Code: [Select]
ChapString    5+    [85]    -    -    -    UTF-8    Contains the string to use as the chapter atom.
ChapAuthor    5+    [43][61]    *    -    -    UTF-8    Author name of the chapter (company, team or individual).


I'm considering merging the 2 cases into one (the more general one).

Request For Comments

Reply #6 –
Interesting.. When saying that performer = author then I wonder what the composer would be... My suggestion is to have just one ChapterMetaData of type UTF-8 and then the users can be free to repeat that filling it up with pairs of the same type as let's say ape tags or vorbiscomments: "KEY=Value". Then of course also have (Album)MetaData for the whole file as well...

Alternatively: <ChapterMetaData><Key>Performer</Key><Value>Name here</Value></ChapterMetaData> since EBML apparantly is some sort of binary equivelent to xml...

By the way... Your spec look a bit similar to asf imo. Did you look at that while writing it?

Request For Comments

Reply #7 –
Hi ErikS,

Thanks for the comments !

Actually the key=value system is exactly what we use. The only difference is that we just store a binary ID as key. If you don't have the "dictionary" to read/understand these IDs, you just discard the information. It's exactly like in vorbis comments.

You are completely right, there is no difference between the author and the performer. I'll add that right now.

Also you can already put as many authors as you wish in the file (notice the * next to the Author elements which means that the element can be found multiple times).

The ChapterMetaData you want to add is already there and called ChapterAtom. It just contains more than just text elements (the start and end timecodes for example). I don't think one more level of concatenation is needed at this level.

Finally no, I've never read the ASF specs at all. I think ASF is just AVI made for streaming or something like that.

Request For Comments

Reply #8 –
BTW, I just figured out that if you have many authors and authorURL you have no idea of which URL goes with what author... I'll change that.