Dial-a-Song Names Your Tune
LIANE HANSEN, host:
This is WEEKEND EDITION. I'm Liane Hansen.
As families gather together this holiday season, musical tastes that don't cross generations will sometimes collide,and sometimes mesh well. Sometimes Mom or Dad or Aunt Dora or Uncle Howard will hear a tune that's new to their ears. And because it's very appealing to be cool in the no-oldsters, now when the kids play a song that's new to you, you can pretend you know the tune, or you can call 411-SONG, hold up your phone to the stereo speakers and let the machine on the other end of the line name that tune. Sunjay Guleria founded 411-SONG, and he's in our New York Bureau.
Welcome to the program, Sunjay.
Mr. SUNJAY GULERIA (Founder, 411-SONG): Thanks, Liane. Thanks for having me on the show.
HANSEN: You're quite welcome. You have to tell us what exact--411 I know means that's the number you dial for information. So 411-SONG, you're getting information about a song, but how does it work?
Mr. GULERIA: We have millions and millions of songs in our database, and what a consumer is doing is essentially calling the phone number, holding their phone to the music, and we're interpreting that music on the other end and matching against our database of millions of songs. And basically there's one unique song for each sample that we hear, and we return back the result. We send the user a text message containing the song information, and we actually include, also, a link to the ringtone.
HANSEN: What if it's a song that's been done by different people?
Mr. GULERIA: Every individual prerecorded song will have a unique fingerprint. That's the underlying technology. It's audio-acoustic waveform fingerprinting. And if there's a unique song, we will identify it with a unique fingerprint and we will deliver back a result. So if it's the same song done by multiple artists--say a Bob Dylan song redone by the White Stripes--then each individual song will have its own song information contained. So we will know the difference between the Bob Dylan version and the White Stripes version.
HANSEN: And what kind of music do you have in there?
Mr. GULERIA: We have all kinds of music: pop, hip-hop, jazz, Latin, rock obviously, international world music. So it's really a very deep and broad database of music. What we have difficulty with is classical music. Classical music is a little more difficult for us to identify because you'll have different versions of the exact same sonata or symphony done by different artists or performers, and it's next to impossible for us to detect. So we prefer to stay away from classical music, and predominantly it's more pop- and rock-oriented music that we're identifying.
HANSEN: What about live performances?
Mr. GULERIA: Unfortunately, live performances we cannot identify. It's taking a unique fingerprint of prerecorded music and trying to do a match against our database. And in the case of live performances, the performer's performing real time, and there's nuances that we aren't able to pick up.
HANSEN: Sunjay Guleria is the founder of 411-SONG, a music recognition service, and he spoke to us from our New York bureau.
Thanks a lot for your time.
Mr. GULERIA: Thanks, Liane.
HANSEN: All right, let's see if this thing works. Let me first call their number, which is 1-8-5-5...
(Soundbite of "All Along the Watchtower"; phone being dialed)
Mr. BOB DYLAN (Singer-Songwriter): (Singing) ...said the joker to the thief. `There's too much confusion.'
Automated Voice: To ID your song, just wait for the beep or press `pound' for help and information about how to get unlimited IDs. You'll be charged 99 cents per song identified, but if we can't ID your song, we won't charge you. Ready? Identification begins now.
(Soundbite of "All Along the Watchtower")
Mr. DYLAN: (Singing) ...none of them along the line know what any of it is worth.
HANSEN: Here's the message. This is "All Along the Watchtower" by Bob Dylan. Worth 99 cents.
(Soundbite of "All Along the Watchtower")
Mr. DYLAN: (Singing) There are many here among us who feel that life is but a joke. But you and I, we've been through that, and this is not our fate.
NPR transcripts are created on a rush deadline by Verb8tm, Inc., an NPR contractor, and produced using a proprietary transcription process developed with NPR. This text may not be in its final form and may be updated or revised in the future. Accuracy and availability may vary. The authoritative record of NPR’s programming is the audio record.