Linux.com

Kid3

Like most of these tag editors, Kid3 uses a directory tree browser to select local audio files for editing. It uses external archives, such as Discogs, MusicBrainz, GnuDB, and TrackType, to search for tag information.

Kid3 will import tags using any of the available external archives, and you can easily experiment with archives to find the best match. Most of the searches are fast, with the expected exception of MusicBrainz fingerprinting, which requires reading the files to create the fingerprint instead of using simple term-based searches.

Kid3 is the best tool for finding exact matches quickly. Some albums, such as Boston's self-titled debut release, required MusicBrainz fingerprinting, but these searches were able to identify every track correctly. Unfortunately, MusicBrainz does not supply genre information, so you must add it manually before you apply the tags to the files.

I found other albums, such as Billy Joel's Storm Front, much faster and just as accurately using a term-based search with Discogs, though this still required some manual browsing of multiple potential matches to find the exact match.

Kid3 does not include external plugin support to extend the feature set. It does download cover art automatically when it's available and attaches it to files. You can tag and rename multiple files at one time, though only for a single album directory at a time. The tool is not preconfigured to rename files with underscores to files with spaces as most other editors are, but you can set this manually in the preferences dialog.

Kid3 supports renaming files using multiple tag fields. A large set of preset formats is included and can be edited to suit your needs on a per-file basis. Tags supported include track, album, title, artist, and year, with most of the presets focused on variations on the use of spaces, dashes, and other types of punctuation marks. The same format specification can be used to retrieve tags from filenames.

The biggest problem with Kid3 is that it's not clear how you should use it. Apparently you have to select Tools -> Apply File Name Format and Tools -> Apply Tag Format before you save the updated tags or the files are not renamed and the tags are not applied. However, selecting these two menu options causes no change to the UI, so it appears as if nothing happens when you select them. If these settings can be enabled or disabled, then there should be an indication of which are enabled and which are not, either in the menu, the preferences dialog, or in the toolbar. Either that or the program's Save option should handle this automatically.

Kid3 is the best tool I tested for finding exact matches to individual tracks, but its UI is confusing and the online documentation is not complete enough for typical desktop users.

Picard

Picard is a successor to most of the other tag editors in that it tries to do more than simple term-based searches. Instead, it relies on a process of fingerprinting audio files, in a manner similar to a checksum but not quite as simple, and matching the fingerprint against a database of audio collections. Theoretically the fingerprinting should offer better match results when you're doing automated tagging. In practice the number of potential matches is large and exact fingerprint matches are uncommon. Still, despite the lack of exact fingerprint matches, the metadata associated with the matches (track names, ordering, album titles, artists, and so forth) does match. This makes tagging files fairly easy by associating the returned matches with any set of audio files through a drag and drop process.

Picard offers a traditional directory browser to select audio files for editing. You drag album directories from the browser into the middle pane of the main window, where untagged files are added to an Unmatched Files folder and tagged files that Picard can recognize are added to the Clusters folder. Since none of the files I was working on for this article had any tags, all the files end up in Unmatched Files.

Picard works with your default Web browser. When you search for matches, the potential matches show up in the browser. You select the correct match in the browser by clicking on the Tagger icon to open the selected album in the right pane of the Picard window. Then you can drag the files in the Unmatched Files folder onto their matches in Picard's right pane. As you drop them in this pane the entries are given color-coded icons to give you some idea of how well they match up with the data from MusicBrainz. Icons include a green check mark for an up-to-date track or a musical note icon for files in the database that are not matched to one of your files. Colored triangles ranging from red (poor fingerprint match) to green (good fingerprint match) give the quality of the match between your files and the file information from MusicBrainz.

One of the nicest features of Picard is its ability to work with multiple albums at a time. You're still better off acquiring matches one album at a time, but you can accumulate multiple album matches and edit the tag information for each before applying all the tags at once. This makes working on a large collection of untagged audio files easier in Picard than with most of the other tag editors.

Picard is highly configurable. It is tightly integrated with MusicBrainz (the two projects are built by the same teams) but also includes support for plugins to access other Internet archives. A large number of plugins are included in the base program and integrate with a variety of external resources, including Last.fm, Google, Amazon, CastAlbums.org, and the Lortel Archives. Cover art can be embedded in music files or saved in the album directory. Filenamess can be set from or used to gather tag data through the use of a built-in scripting language.

With all the good, there is some bad for Picard. If you use a workspace manager and have Picard in one workspace but your browser in another (as I do) then every time you switch workspaces Picard's directory browser tree returns to the root directory. This means you have to navigate back to the folder where your music is every time you switch workspaces, which is extremely annoying.

I couldn't get Picard to add genre tags. And while Picard is designed around searching and clustering multiple albums at a time, the fingerprinting process and matching across the Internet is very slow.

Picard's drag and drop features give you a little more direct control than the other tag editors to apply matches and tags while also allowing multiple album editing. But in general I found Picard to be slower and less accurate with its matching than some of the other taggers. Its best use is as a initial tag editor for large collections that have no tags at all. Further editing on individual files will probably be faster, more complete, and possibly more accurate using one of the other editors.

Ex Falso

Ex Falso is a standalone tag editor based on the QuodLibet audio player. Both projects are hosted by Google. Unfortunately, they simply didn't perform in these tests.

Ex Falso supports file renaming based on tag fields but there is little information on what tags can be used. A separate tab in the main window allows editing and previewing the filename format. A similarly formatted tab can be used to map from the filename to tag information.

Ex Falso provides a typical directory tree browser. Once you select an album directory, all songs in that directory are displayed in the Songs pane of the main window. Online documentation says that external plugins are supported, though information on finding and using plugins is scarce on the Web site. Automated searches are integrated with Google, Wikipedia, and CDDB via plugins, but you have to manually enable their use. Right-clicking on selected songs opens a menu that performs the searches -- but none of these searches worked with the version of Ex Falso I was using. For this reason, I cannot recommend this tool for tag editing.

Summary

In each of these editors, audio file tagging took quite some time to complete, likely because the tag information has to be prepended to each file, and that typically requires reading the entire audio file.

For tagging large numbers of files in a mostly automated process, start with Picard, based on its ability to queue up multiple albums for processing in an intuitive manner.

Kid3 offers multiple options for automated matching, and makes the process easy, once you learn to navigate the UI and to experiment with different Internet resources to acquire tag information.

But the best of breed designation has to go to EasyTag. Its only real drawback is that it doesn't queue up multiple albums for tag processing in batch while you wander off to eat dinner. But it was faster and better at finding exact matches for non-mainstream albums such as classical and holiday music and soundtracks than most other tools. It also automatically provided genres for albums and offers users the ability to manually edit a larger variety of tags than most of the other editors. Its UI is intuitive and easy to navigate. It also happens to be one of the older and more stable editors around.

Michael J. Hammel is a principal software engineer for Colorado Engineering, Inc. (CEI) in Colorado Springs, Colo. He has more than 20 years of software development and management experience, and has written more than 100 articles for numerous online and print magazines. He is the author of multiple open source related texts, including three books on the GIMP.

Share    Print    Comments