You said "With the exception of where to put a picture file, VorbisComments in a Vorbis stream are the same as VorbisComments in a FLAC stream."
In METADATA_BLOCK_PICTURE case they would not be the same. Up to now we could exchange the complete block of VorbisComments. But with the addiction of METADATA_BLOCK_PICTURE doesn't happen this way. In Ogg files the METADATA_BLOCK_PICTURE data are split in numerous pieces, each one of them with a Oggs header (44 up to 282 bytes of non Base64 data among each piece). Just as they are written would be impossible to put them in the block 4 of Flac Metadata. But the specifications are only for Ogg files. For Flac is not specified if the data, after METADATA_BLOCK_PICTURE=, would be a long string of Base64 format (without split) or otherwise. It seems that the decision is of Ogg team and Flac team has not said anything on this matter. I thought I could choose among both forms of inserting a picture. But thinking it well, the decision is made because of the silence of Flac. Flac will continue using the block 6. And in METADATA_BLOCK_PICTURE case, VorbisComments are not compatible. I continue thinking that to have two different blocks for this is very bad design. Regards. Santiago ----- Original Message ----- From: "Martin Leese" <[email protected]> To: "Santiago Jimeno" <[email protected]> Sent: Friday, January 21, 2011 7:01 PM Subject: Re: [Flac] Flac] Where Cover Art? > On 1/21/11, Santiago Jimeno <[email protected]> wrote: >> Thank you Martin for your answer. >> >> Regarding the COVERARTMIME name, I found this entry in ExifTool: >> http://www.sno.phy.queensu.ca/~phil/exiftool/TagNames/Vorbis.html#Comments >> I suppose that it would be useful to enlarge the information of which >> COVERART lacked. > > Hi Santiago, > > As the webpage you cite states, "The tags [on > the page] are only some common tags found > in the Vorbis comments of Ogg Vorbis and > Ogg FLAC audio files". This list is not official. > I have never seen COVERARTMIME used > anywhere else and, as COVERART is going > away, I would not recommend its use. > > The official list of tag names is in the Vorbis > documentation, and consists of only 15 tags. > There are also several lists of suggestions for > other tag names here: > https://wiki.xiph.org/VorbisComment#Proposed_field_names > >> To change the code to use METADATA_BLOCK_PICTURE is simple, since >> COVERART >> also wrote in Base64. The result continues being the same handling >> easiness >> of Vorbis Comments. >> >> But in the FLAC case the problem continues existing. When I spoke of >> error, >> I referred to try editing (adding or removing comments or pictures, >> writing >> besides reading) in two different blocks at the same time. We can call it >> error, ambiguous or otherwise. But, have you tried to write code to >> maintain >> this double localization in a simultaneous way? > > As I explained, you don't put the same picture > file in two places. You put it in only one place. > You get to choose where you put it, either in > a VorbisComment or in a FLAC > METADATA_BLOCK. What is needed is some > official advice on which of these two locations > is the preferred place to put a picture. > > ... >> In LYRICS case, also "needs to be some official guidance." It's possibly >> not >> very utilized, but it's used. But I cannot invent a Comment name, >> although >> this you suggest could be the most appropriate. If somebody that works in >> Vorbis Proposal reads this that takes note of the suggestion. > > Many, many tags are used. People invent new > tag names all the time. The Xiph community > will never produce an exhaustive list. The > official list will always be a short list of > commonly used tag names. It will always be > permitted for others to invent new tag names. > >> In any event, if we want Vorbis comments have the same success in FLAC >> like >> in OGG, we need clear, simple and mainly effective official norms. > > With the exception of where to put a picture > file, VorbisComments in a Vorbis stream are > the same as VorbisComments in a FLAC > stream. They are not as widely supported by > software applications, but there isn't much the > Xiph community can do about that. > > Regards, > Martin > -- > Martin J Leese > E-mail: martin.leese stanfordalumni.org > Web: http://members.tripod.com/martin_leese/ > _______________________________________________ Flac mailing list [email protected] http://lists.xiph.org/mailman/listinfo/flac
