Apple Music launched just over a month ago, promising the most comprehensive and fluid user streaming experience available. Users, however, are experiencing problems matching up songs from their library using the service, which is providing the wrong songs or incorrect versions of songs.
Apple Music appears to be successful so far for the company, having attracted over 11 million subscribers since its June 30 launch. The real test will come, however, when subscribers' free three-month trial periods are over and they have to decide whether to continue paying for the service on a monthly basis.
One of the selling points of the service was the promise that iTunes users would be able to easily synchronize songs from their iTunes library into the streaming service, allowing for a seamless transition between the two. However, it appears as if many subscribers are having problems with the promised song matching, as they are receiving the wrong versions of songs, and, in some cases, a completely different song than they are looking for.
The problem seems to be a result of Apple Music using a more simple and less sophisticated song matching system than iTunes Match, the older and more complex system available to iTunes users. iTunes Match uses acoustic fingerprinting to analyze a song and properly identify it, whereas it appears Apple Music only uses words and tag lines, often based on user input, to identify a tune.
Kirk McElhearn, a senior contributor at Macworld, tested the Apple Music tagging system by changing just the title of a classical Bach tune in his library to a hit pop song, "I Can't Feel My Face" by The Weekend. Apple Music, apparently only utilizing the metadata contained in McElhearn's written tag, downloaded the pop tune from the cloud instead of the Bach composition. The matching system seemingly ignored not only the clear musical differences, but also the huge variation in length of the two tracks.
"This is a very big problem with Apple Music," says McElhearn. "Since Apple already has the technology to match tracks using acoustic fingerprinting, they should be using this with Apple Music. Instead, it's using scattershot matching, which results in lots of tracks showing up as being from different albums, from compilations, or totally different versions of songs."