Commercial MP3s and other digital music files provide a wealth of information about the songs in addition to the audio-playback itself. This metadata makes it easy to display the track name, artist, album, and other facts about the songs in your playback device or program.
When you use an application such as the free Audacity audio-editing utility to convert music from LPs, cassettes, or another analog source, the only metadata accompanying the tracks is whatever information you provide when you create the digital file. There's the rub.
In July 2011 I described how to. A follow-up post from last September explained how Audacity helps .
Over the years I've used Audacity to convert about 2,000 songs from analog LPs and cassettes to MP3s and other digital formats. Most of the cassettes were recorded manually (as opposed to being store-bought) and date back to the 1970s and 1980s.
Alas, my friends and I weren't always thorough when we made the tapes. Many lacked any information about the tracks or artists other than a generic label such as "Blues" or "Cajun." Of the 5,000 or so files in my iTunes library, about 175 are named something like "Unknown," "Instrumental," or "Irish," usually without an artist name.
Other nameless cuts provide some hints as to their identity, such as "Grisman 5" or "De Dannan Ballroom." I combined three different programs -- one free and two fee-based -- with some Web searching to reduce the number of mystery tracks in my 5,000-track iTunes library to fewer than two dozen.
TuneUp tops SongGenie as a music identifier
For songs you imported from iTunes, you can find song titles by right-clicking the track and choosing Get Track Names. You can often find more information about any track in your library by right-clicking it and selecting Show in iTunes Store. This option rarely works for entries lacking a title, artist, or album name, however.
I had limited success using TuneUp Media's $50 TuneUp add-on for iTunes to ID the unknown tracks in my library. (Note that through March 2, 2013, the company is offering a 30-percent discount when you enter the promo code "LOVE30.") I used the Mac version of TuneUp; a Windows version is also available.
After reading the user ratings for the program I feel fortunate TuneUp didn't do more harm to my iTunes library than good. I didn't experience the problems with the program that other people report, but I also didn't let TuneUp make any changes automatically.
The program opens like an iTunes sidebar. Its four tabs are Clean, Cover Art, DeDuper, and "Tuniverse," which shows concert and other information related to the current track, and offers links for sharing your recent activity via social media.
TuneUp's DeDuper was particularly error-prone when I tested it. For example, two very different versions of the same song were identified as duplicates more than a dozen times. The program also recommended several times that I retain the lower-bit-rate version of a song that had been converted at both 128 kbps and 256 kbps.
Since I'm not a big fan of cover art generally, I wasn't particularly enthusiastic about testing this feature of the app. While I didn't bother conducting a before-and-after count, TuneUp found about 15 percent of the album covers missing from my library of just under 1,000 albums.
TuneUp's automatic library analyzer indicated that my iTunes "collection" was 61 percent clean: more than a third of the files in my library had "missing or inaccurate information." Nearly all the errors the analysis offered to correct were in the "Genre," "Year," and "Track Number" categories.
Since the "Track Artist," "Track Name," and "Album Name" categories came up pretty clean after TuneUp's scan of my library, I decided not to allow the program to do any automatic correcting. Instead, I focused on "cleaning" those 175-or-so mystery tracks.
TuneUp identified the name and artist of only about 25 percent of those tracks. While the program misidentified about 10 percent of the music files I scanned, it often came up with the correct album but the wrong track, so I was ultimately able to ID many of these tracks as well.
The album information TuneUp displays after it scans your songs proved particularly useful. For example, when I scanned 16 nameless tracks I had converted from an LP of Renaissance dances, TuneUp identified only three of the tracks, but I was able to add the names of the other tracks from the complete album song list shown in the scan results.
Some TuneUp users complain that the program made a complete mess of their iTunes library that required an effort to repair. The utility has an Undo tab that promises to revert any files it cleaned to their precleaned state. I didn't test this feature, but considering the number of misidentifications I encountered, an undo option could come in handy.
Equinox's $30 SongGenie, which is available only for the Mac, promises to identify a song's title, artist, album, year, and track number. It also offers to retrieve the song's lyrics and search the Web for the lyrics it can't supply from its own database.
While SongGenie's interface beats that of TuneUp hands down, the program identified fewer of my mystery tracks. Whenever I scanned songs from the same album in both programs, TuneUp consistently recorded more hits. In SongGenie's favor, the app was able to retrieve lyrics to more songs, and it seemed to make fewer errors than TuneUp.
To search for information about a song, select it in the list a the bottom of the SongGenie window and click the Identify button under the box of information for the selected song at the top of the window. SongGenie highlights its proposed changes in red. To accept the suggestions, click the Apply button.
You can also apply only some of the changes SongGenie suggests and add or change song information manually. The program lets you select and scan several files at once. (Note that not all features are available on the version of SongGenie that's sold on the Mac App Store.)
SongGenie works with any music library, not just iTunes. The program rates the completeness of your library by displaying a "genie" with bar charts indicating song information, cover art, and lyrics, as well as an overall rating, such as "Not bad" or "Nearly there."
In my testing SongGenie appeared to be more accurate overall than TuneUp, although the program frequently proposed changing the artist of covers. For example, when I checked the Modern Mandolin Quartet's version of Aaron Copeland's "Hoedown," SongGenie suggested replacing the quartet's name with the composer's.
Shazam helps fill in the music-track blanks
After scanning my iTunes library with both TuneUp and SongGenie I was still left with several dozen unidentified tracks. These songs I played for the free Shazam app for iPads and iPhones.
Shazam has a reputation for being great at identifying pop songs but not so good at ID'ing classic tracks and other genres. Most of my remaining unknown tunes were from compilations of blues, jazz, Cajun, and Irish folk music, with a couple of obscure classical pieces thrown in for good measure.
I knew these tracks would challenge Shazam's music-signature database, particularly since several of the recordings were live performances, which Shazam states up front it can't identify. Still, I was pleasantly surprised at the program's ability to retrieve song names, albeit with more than a few erroneous matches.
To use Shazam to identify a song, simply play the track, click the stylized "S" icon in the top-right corner of the program's window, and position the device's microphone near the speakers. After a few seconds Shazam either displays the name, artist, and other information about the song, or pops up a window stating it was unable to make the identification and offering to try again.
As expected, Shazam wasn't able to identify four unknown classical works or about a half-dozen Irish instrumental tracks, but the program made quick work of some pop, rock, and blues numbers.
Work the clues to determine the title that goes with the track
Among the remaining untitled songs in my iTunes library were several with hints as to their identity, such as "Norman Blake 4" and "Grisman cut 9." The music-ID programs provided clues about the albums the songs were recorded from, so I searched for those albums in the iTunes Store and on Amazon.com.
It's no surprise that the tunes I had the toughest time tracking down were instrumentals. You can't always surmise the title of a track based on its lyrics, but searching lyrics databases for the words repeated prominently in the song's chorus will usually lead you to the correct name pretty quickly.
Two other clues to a song's identity are the track length and the songs that were recorded immediately before and after it, since most people record songs in the same order as the tracks on the original medium. Neither method is foolproof, but they help you make educated guesses about a song's title.
I hesitate to draw conclusions based on the small number of tracks I looked for, but I found more information about the missing songs on Amazon.com than in the iTunes Store. After playing the samples on Amazon.com, comparing track lengths, and doing a little research on Wikipedia and other sites, I deduced the names, artists, and albums of all but 23 of the mystery songs.
And I'm not done yet. Much of the information I gleaned about the tracks when I converted them from analog to digital turns out to be inaccurate: wrong artist names, album names, and track numbers.
Which reminds me of what a friend said when I told him about this project: Will knowing the name of a song or artist make me enjoy it more? Maybe not, but it will definitely make it easier for me to recommend the tune to somebody I think might like it.