MusicBrainz Picard/Documentation

From MusicBrainz Wiki
< MusicBrainz Picard
Revision as of 07:25, 29 March 2006 by Shrike (talk | contribs) (Linkin to Linux install page for picard (Imported from MoinMoin))
Jump to navigationJump to search

Documentation for Picard

This page discusses the next generation PicardTagger.

Quick start for the impatient

We have a quick start guide that shows you HowToTagFilesWithPicard. If you've never used Picard before, please read the guide or the remainder of the page for it to make sense to you!

Basic Picard Documentation

This tagger takes an album approach to tagging, whereas the old tagger required the user to work with one track at a time and check to see if the tagger correctly identified the track. The amount of time spent reviewing the old tagger's suggestions will probably equal the amount of work the user has to do to find the right albums in the new tagger. More details on this thought later -- first let's explain the basic new operations.

Start with opening individual music files or directories by dragging them into the "New files (drag files to tag here)" folder. The tagger will read the metadata from each of the files and unless they have been tagged before, the files will be deposited into the "Unclustered files" folder. Files that have been tagged before and contain the MusicBrainz track identifier will be opened up as albums in the album view.

Once the tagger finishes processing the files, press the cluster button (white square with 'clusters' on them). This will cause the tagger to attempt to group the files into album clusters by examining the metadata read from the files and clustering files that appear to belong to the same album. Files that are not matched into album clusters will remain in the "Unclustered files" folder. In the future the tagger will automatically cluster the files when it has finished reading the metadata from the files.

This is where the magic currently ends -- from now on you will need to look up album clusters or individual files by selecting the cluster or file and clicking on the lookup button. This will open a web browser window where you can browse/search for the appropriate album using the MusicBrainz web-site. Once you've found the right album, click on the 'tagger' icon in the album title: mblookup-tagger.png . Once you click on this link , the tagger pops up and loads that album into the tagger.

Now you can drag tracks from the "Unclustered files" folder or from the album clusters onto tracks in albums. Dragging a file to an album 'tags' it to that album and you should see an icon in front of the track:

  • a green check mark indicates the track is up to date and saved.
  • a small rectangle ranging from red to green indicates the quality of the match. red -> bad match, green -> good match
  • a blue ? indicates an unmatched file.
  • a red error triangle means the file has an error. See the status bar in the bottom of the picard window to see the error.
  • no icon means that no file is associated with that track

You can also drag whole directories, multiple files or album clusters onto albums and the tagger will attempt to match the dragged files to the album. Any track that doesn't match up well enough, will be added to an "unmatched files" sub-folder specific to that album. You can drag files out of this folder and into the right slots in the album to fix up the files the tagger couldn't get right.

Once you've matched up your tracks to albums you can select a track or an album and click on the save button to save that track/album. Depending on your settings this may move the track to a new directory and/or rename the track according to its metadata. Take a look at the options dialog to fine tune your settings.

Once you've saved a file, or otherwise want to get rid of a track or album, select the track and click on the delete button in the toolbar.

Libraries Required (Linux Only)

Latest versions recommended for all libraries.

  • python-wxGTK
  • libtunepimp python bindings

Likely more as well.

Read more details about PicardLinuxInstall.

Isn't there more to Picard?

Not yet -- but soon!

Currently the cluster identification process is completely manual -- in the future Picard will automatically download albums for all the artists it finds during the cluster phase. Picard will then go through the unmatched files and attempt to match them against album information that was downloaded.

But for right now, Picard is a little more 'do it yourself' -- however, I think that Picard will let you tag files faster than the old tagger -- you just need to get used to the new way of doing things.

Unfortunately, the current version of Picard is also not very user-interface friendly. The tagging/matching/association works well if you have just a few albums, but as soon as you have several hundred clustered albums, dragging them to associate with the TRM identified real album data gets too unwieldly. Hopefully this will be something the developers of Picard might address in a future version. Besides that, though, Picard works great!

What works?

  • Open files into [new files] by double clicking files
  • Drag files, and directories into [ Drag tacks to identify here ]
  • Drag files into specific album slot to associate file with the track.
  • Drag files/dirs onto an album line and picard will auto match tracks to the album. Tracks that don't match the minimal matching threshold, will go into a special[ Unclustered files ] folder that gets created under the album.
  • Saving files with move and rename support, with support for UNICODE filenames and tags.
  • Remove files/albums
  • Cluster unmatched files into album clusters and automatically matching clusters to albums when albums get opened.
  • Preference dialog features
  • Cover art is shown for albums that match CDs on Amazon.com. The user can click on info and buy links right from the tagger.
  • Looking up Audio CDs
  • Searching for artists/albums/tracks from the toolbar
  • Localized to Norwegian, German, French, Russian and Slovak. See PicardTranslation for more details.
  • Some multi-select support in the album panel.

What is planned to work in the future?

The following is a non-exhaustive list of all the things that don't work yet:

General

  • The edit menu isn't hooked up yet.
  • No keyboard shortcuts for drag and drop operations
  • All files that have no MB ids in their tags will always end up in the unclustered tracks folder when dropped ontonew files -- use the Cluster button to cluster these unmatched files into album clusters.
  • Add Helix Community support by using PyPlayer for music preview and for mp3 decode support.

Album Panel

  • Multiple select in the album view works only in a limited fashion

File Browser

  • The file details should show up in Local mdata when file clicked
  • Files should have appropriate icons

Metadata Information Panel

  • The artist and album combo box drop downs do not populate with correct data
  • Even though the metadata for a track can be edited, it still can't be saved.

Menus

  • The menu is still sparse and needs more items
  • The edit menu cut, copy and paste options are not functional yet.
  • The application needs (more) context menus.

Browser launching

  • For each lookup a new browser window is opened -- the user should have more control over this process.

Platform Specific Issues

  • The application has not been tuned for Mac OS X -- wxPython still has some serious bugs that prevent Picard form running well. We'll have to wait for further releases before we can attempt to release a version for Mac OS X.

This WikiDocsPage is MaintainedByNobody (Robert, Luks?)