Clean Music Metadata with a GUI





Clean Music Metadata with a GUI

Use the Picard GUI tool to access the MusicBrainz functionality for cleaning up your music files.

This hack focuses on the GUI version of pimpmytunes. If you prefer GUI tools over command-line tools, this hack presents the best way for you to clean up your music metadata. [Hack #31] explains pimpmytunes and why you should clean up your music metadata. For an introduction to MusicBrainz and its approach to metadata cleanup see [Hack #30].

To install Picard, go to the Picard Tagger homepage (http://musicbrainz.org/wd/PicardTagger) and follow the download and software prerequisite instructions listed there. Once you have Picard running, take a look at the Settings dialog box and tweak the settings to fit your tastes. In particular, notice and adjust the Directories and Naming tabs in the Settings dialog. The pimpmytunes hack talks a bit about file masksplease refer to it for a quick introduction to file masks.

Now that you've set up the basics of Picard, its time to take a closer look at the user interface (see Figure).

Picard has three main user interface areas:


Directory/file browser

This is on the left half of the screen. The top half of the panel allows you to browse your file system and find directories that contain your music collection. Click on a directory and the lower panel will show all the files Picard can work with. Note that Picard shows unsupported files in gray and does not let you drag those files.


Album browser

This panel on the right side of the window shows folders that organize your new files that are pending analysis by Picard, files that Picard hasn't matched, clustered albums, and albums loaded from MusicBrainz. I'll cover this panel in a lot more detail in just a bit.


Metadata information panels

This panel along the bottom of the window contains two tabs that show local metadata (for a selected file), MusicBrainz metadata (the metadata that MusicBrainz has for this file), match album information, and album cover art.

To tag your files with Picard, first use the directory browser to locate your music collection, and then use the file browser to select a handful of files to start. Next, drag the selected files from the file browser to the New Tracks (drag files to identify here) folder in the Album browser panel, which tells Picard to read the metadata from these files. If Picard identifies a file, it will open the matching album in the Albums folder and move the file to that album. Picard will add all other files to the Unmatched Files folder.

Picard user interface


Once Picard completes reading the metadata from the files, click on the Cluster toolbar button to have Picard try to group the files into albums. If Picard finds tracks from the same artist and same album, it considers those files a cluster and adds a cluster entry under the Album Clusters folder. It is easy for Picard to create clusters when you have almost clean metadata that needs complete tagging or small corrections.

If Picard finds some clusters, the next step is to select a cluster and click the Lookup button to launch a web browser and find the right album using MusicBrainz. Once you've found the right album (be careful to select the album that has the same number of tracks that you have), click on the green tagger icon in the album web page to open this album in Picard. Switch back to Picard and drag the recognized album cluster onto the newly opened album (if Picard didn't already do it for you). This should match the files from the cluster to the newly opened album.

If Picard didn't find any album clusters, you can click on a track in the Unmatched Files folder and click the Lookup button to look up the album for this track. You could also enter the artist/album name and use the search function in the toolbar to look for the right album. Either way you do it, click on the tagger icon to open the selected album in Picard.

Once you have the correct album open in Picard, you can drag files from the Unmatched Files folder or an album cluster and drop it onto the appropriate track in an album. You can even drag a file onto the album to have Picard guess which track this file should be associated with. If Picard guesses incorrectly, simply drag the file from the wrong track onto the right track. If Picard doesn't find a good match, it will open the subfolder Unmatched files for this album and deposit unmatched files there. Figure shows a list of Picard's drag-and-drop operations.

Each file that has been associated with a track on an album has a small, colored icon in front of it. A green box indicates a good match between the file's metadata and the track with which it's associated. An orange box indicates a medium-level match, and a red box indicates a poor match. The background color of an associated track changes from the normal background color for good matches to a reddish background for bad matches. These visual cues make it easier for you to quickly grasp how well Picard matched your files to tracks and where you need to spend your time verifying Picard's work.

Once you're happy with a match, select the track or the whole album and click on the Save toolbar button to save the new metadata to the file(s). After Picard saves the file, the icon in front of the file changes to a green check-mark, showing that the file is done and checked off.

Picard's drag-and-drop operations

Action

Function

Drag album cluster onto album

Match files from album cluster to this album.

Drag file onto album

Match file to this album.

Drag file onto track

Associate file to this track.

Drag album onto another album

Match the files from source album to target album.

Drag Unmatched files for this album onto Unmatched Files

Move all the files that Picard could not match to an album back to the Unmatched Files folder.

Drag a file from the file browser onto a track/album

Load the file into Picard and then associate the file with the track/album onto which it was dropped.

Drag a directory from the directory browser into the New Files folder

Load all the files (and files in all its subdirectories) from the directory into Picard.


Robert Kaye



 Python   SQL   Java   php   Perl 
 game development   web development   internet   *nix   graphics   hardware 
 telecommunications   C++ 
 Flash   Active Directory   Windows