Hi (sorry this became long, but replies to many),

I have before spoken warmly in favour of Musicbrainz, and I shall do
so again since it without doubt is *the* best and most ambitious free
(and probably also non-free) meta-data system for music and we should
all use it (and contribute)!

Apologies if I'm repeating stuff or stating the obvious in the
following.  

First a word of warning. Picard does not presently afaik support all
MB data and SC does not support all tags used by Picard. There is no
inherent reason afaik this could not change as to allows SC to take
more advantage of the data in MB.

Also some SC-users have very specialised taggins-schemes used in
combination with various SC-plugins to obtain a highly customised
setup. MB might not work well for those.

First an example of tags that I got automatically using Picard (with 3
plugins).

http://img11.imageshack.us/img11/1306/tagsu.png 

Moonbase;421140 Wrote: 
> 
> '-MusicBrainz-' (http://musicbrainz.org/) is a big effort
> to build a comprehensive database of existing music, based on
> human-edited entries and much much more information, ...  

Indeed true. As an example, such information can include composer
(down to the details of lyricists etc), band members, performer
vs. composer, the roles of group members (guitar,drums,vocals etc). On
top of that there are cross relations so that one can extract e.g. a
band-members other gigs (bands, solo, composer etc.). Additionally
there are relevant links to eg amazon, wikipedia, myspace or
whathever. And there's more; as said, it is *very* ambitious.

Mnyb;421242 Wrote: 
> Have you found any taggin application for Linux
> that uses musicbrains and has an human freindly interface 

What??? Picard has very few menu-items and probably less than 10
menu-buttons. Its about at clean as it can get. Additionally it does
some things automatic in a very clever way.

Mnyb;421242 Wrote: 
>  i tried picard but that was just horrible
> cumbersome and slow and no cover art either.  

Cumbersome meaning exactly what? I postulate that if its slow, then
you are using it wrongly. Eg. one is not supposed to analyse all
tracks by default? Cover-art is available using a semi-official
plugin. By that I meat that AFAIK its not in the core for the sole
purpose of keeping the core minimal.

Mnyb;421242 Wrote: 
>  And picard added a bunch of non standard
> "musicbrainz" tags ?  

Non-standard? What do you mean? Flac has no standard set of tags. I'm
pretty sure Picard does not add stuff not coming from MB as it is
*the* official MB tagging utility.

Mnyb;421242 Wrote: 
>  is any other application using this, does
> SqueezeCenter use any of this info ?  

SC uses at least composer, performer, sort-order and such (but still
only a small subset of the MB info). This in addition to the standard
tags such as artist/track/album of course.

NFLnut;421794 Wrote: 
> But I find their tagger, Picard, to be a little
> too automated, cumbersome, slow, and IMO, erroneous for my tastes.
> 

In what sense exactly (apart from the issue you mention below if any)?

NFLnut;421794 Wrote: 
> 
> For example, for older music (much of my catalog is from the 60s and
> 70s and earlier) it would often pull up some greatest hits compilation
> from the 80s or 90s.  When *I* want to tag with the original artists'
> album including cover art from that album, and the actual year of the
> song's first release. 
> 

OK, these cases can be difficult as such tracks often appear on many
releases and in many different versions. Moreover releases that old do
not have an original CD release, hence can not be provided with a
discid. Finally that type of release is probably not that well covered
in the database at this point in time. However it is up to us to
improve that.

Even so, provided the album is in the database it should work
flawlessly with the magic wand. If not one might attempt an
analysis. Now this step can be rather inexact for the reasons I
mentioned above. Therefore it might be better to manually invoke a
search from Picard which will launch a query at the MB-site. If the
release is found, just click "tagger" in the browser and the info is
transfered to Picard. If the release is not in the MB database you
have to enter it. This is about as easy as it can get with Picard/MB
though. The most common ways are either by importing from freedb or by
submitting the tags already in your files. Its then readily
available. If you do this the release will forever??? be available to
you and all others that have the release. And if someone improves (by
correcting or adding information) some information, you will get those
improvements for free. I don't think this is worse than other tools
given the premises. The only difference is that you edit a central
database as opposed to your local files and therefore have to click to
get that information into the tagger. (Sorry if you knew all this).

NFLnut;421794 Wrote: 
> 
> It allows more individual manipulation of track
> IDs without a lot of fuss

This is the single weakest point of Picard. If you do not collect
by release, but by tracks, tagging will work fine, but the procedure is
sligtly ugly. 

snarlydwarf;421807 Wrote: 
>  It's annoying when MB doesn't have the
> actual disc (especially with Classical, since they are very strict on
> how to enter data), but that's the case of freedb type things too.
> 

In what sense is it annoying? It takes a lot of work to tag classical
music properly. But the goal is to have good tags isn't it?  So with a
bit of persistance it is not difficult to add a new release at MB. So
maybe its annoying that someone else doesn't, either for free or by
charging some amount, provide this information. However if one wants
good tags the work has to be done by someone, so I think its a best
effort scenario.

Its based on sort of a peer-reviewing process, and they are strict. In
essence you should follow the official scheme (unless you can provide
sound arguments for not doing so), and changes should be documented by
"witnesses". Eg if you add an album, provide a link to perhaps amazon
to show that it exists and has the title/tracks/etc you claim. This is
part of the reasons (most of) the tags are so good. I know it can be
discouraging to get ones edits rejected because of some seemingly
benign deviation from the standard. But remember that *all* other
users, including the very meticolous ones, will also get hit by ones
personal "customisation". Freedb is a nightmare as one can often find
a release, but just as often it will be full of errors and/or
non-standardised formatting.

(In practise the rules are not always enforced)

MrSinatra;425594 Wrote: 
>  maybe SC should support that in its core
> first, b4 they try other multiple property recognition methods.
> 

I've suggested tighter integration between SC and MB before and do not
mind repeating that request here.

MrSinatra;425594 Wrote: 
>  question though; does the MB ID tags
> differentiate between say remasters of the same song? the doors for
> instance have at least three different releases, and who knows how
> many differing pressings, of the same songs. would MB ID tags know,
> and tell apart, one from the other?  

You already heard from others that the answer is yes. In the case of
albums sometimes a remastered or reissued album will have bonus
tracks, or other track-list differences compared to the original. This
means several separate albums in most meta-data system. But usually we
actually want the release to be "the same". MB has in its newest
(beta?) release facilities that enables grouping such albums while at
the same time maintaining their separate ids. This is afaik part of
the solution to a more difficult problem as albums we think of as
being the same can sometimes even have differences in artists and/or
title.

Btw. I'm not affiliated with MB, just enthusiastic.


-- 
bhaagensen
------------------------------------------------------------------------
bhaagensen's Profile: http://forums.slimdevices.com/member.php?userid=7418
View this thread: http://forums.slimdevices.com/showthread.php?t=62556

_______________________________________________
ripping mailing list
[email protected]
http://lists.slimdevices.com/mailman/listinfo/ripping

Reply via email to