Out of all the great things music apps do for us, music identification is one of the most helpful, not to mention amazing. No matter where you are, if you hear a song you like but don’t know, a click and some taps will usually solve your problem.
Fortunately — or unfortunately, depending on whether you’re a glass half empty or half full kind of person — you have a few to choose from. Which one is best at identifying music, especially in challenging environments, like a laundromat or a restaurant?
We resolved to find out with some rigorous, real-world testing.
Using the free versions of all three, on both iOS and Android, we tried to identify fifty songs under varied conditions.
We tried to get them to ID songs in various New York and San Francisco locations including coffee shops, bars, restaurants, stores, and homes, to simulate the way the apps are typically used. The music included good old-fashioned American rock, pop, and hip hop, but also instrumentals and international songs, including a few in foreign languages.
We tested all three apps with each song, to maximize fairness.
Based on the number of successful identifications, Shazam and Soundhound blew musiXmatch out of the water, correctly identifying nearly twice as many songs. Admittedly MusiXmatch is primarily a lyrics app with a music ID feature built-in, so this isn’t too surprising. MusiXmatch, we still love you (especially in Spotify).
So which is the best: Shazam or SoundHound? The short story: It’s a toss-up since they both performed well with more mainstream stuff, and when the music was much louder than the ambient noise, although things got more interesting when we challenged them with some truly tough examples.
All three apps had mixed results with foreign-language and international music, so if that’s your bag, you’re not going to be too happy with any of them. At Jamaican restaurant Miss Lily’s in Manhattan, one of our testers tried to identify several reggae, ska, and dancehall songs on Radio Lily, the restaurant’s homespun internet radio station, which also plays there. Only Soundhound had any degree of success there, but it was only able to identify one song: “Everybody Needs Love” by Slim Smith. Another tester gave the apps a spin at her local laundromat, which plays Spanish radio station La Mega 97.9, but all three came up empty there.
Some international music fared better. All three called Italian band Equipe 84′s cover of “Bang Bang (My Baby Shot Me Down),” while Shazam and Soundhound correctly identified Icelandic instrumentalist Ólafur Arnalds’ “Near Light.”
Mashups also presented difficulty. We tested Wait What’s “Juicy-r,” a mashup of Notorious B.I.G.’s “Juicy” and The xx’s “VCR,” in a few different places, to give the apps multiple tries. When testing a part where Notorious B.I.G.’s sample was dominant, all three apps identified the mashup as “Juicy.” In parts where The xx was more dominant, none of the apps identified the song. When we tried to ID a transitional sample of the mashup so that the apps could hear both songs, they picked “Juicy.”
Since these challenging examples did not help us determine a clear winner between Shazam and Soundhound, we looked for extraordinary performance by either app in other ways, so we went through the results looking for instances when only one app successfully identified the song, tallying how many times each one won.
We’re a little reluctant to draw hard and fast conclusions here, because the results were so close, and we couldn’t always test each app on the same exact song sample, due to the real-world nature of the tests. Changes in ambient noise (like a friend yelling in your ear) may have affected the results here and there. Still, there are some conclusions to be drawn.
Based on the sheer numbers, there was precious little difference between Shazam and Soundhound, the top two music identifying apps. Out of fifty attempts (again, we designed the test to be quite challenging overall), Shazam correctly identified twenty-seven songs, while Soundhound identified twenty-eight. On that level, your choice is basically a simple matter of which design you prefer, from a cosmetic standpoint.
However, in quiet settings — as reflected by the fact that all three apps were able to “hear” the song well enough to make an attempt — Soundhound edged out Shazam, indicating that it might have a bigger database when it comes to more obscure music. In noisy environments, Shazam proved the better option, indicating that it’s probably better at filtering out background noise.
If you want to cover all your bases in truly optimal fashion, install them both, and use Soundhound when the music stands out from the background noise, or Shazam if it’s noisy. And if you’re only going to use one, we’d probably recommend Shazam, because it’s a noisy world out there.
Basically, it’s a tie.
Download Shazam (Android, iPhone/iPod Touch, iPad, Windows Phone, Blackberry)
Download SoundHound (Android, iOS, Windows Phone)
Update: Massimo Ciociola, Chief Emotions Officer of MusiXmatch, responds as follows:
“1) People want recognition for getting lyrics. By simply recognizing the song people are not happy with this is the main reason mXm is blowing Soundhound and Shazam out of the water.
While people spend 30-40sec on on Shazam and Soundhound, they spend 3-4 min on musiXmatch.
2) We have a new version of music ID much faster and better
would love to run a benchmark ;-)”