Midomi, a new company in Silicon Valley (Sunnyvale), has just launched an impressive musical search service.
If you hum, or sing a song, Midomi will likely recognize what you’re singing. So if you get a song in your head, and you don’t know what it is, go to Midomi. It will show you the most likely results, based on the signatures of songs it has in its database, and lets you go to a store to buy them.
That’s just the beginning. Midomi lets you see the people who have hummed the same songs, showing you those who sang most similar to you. It lets you chat with those people through a message box. It also lets you store your tune in its database for others to hear.
For example, see the screenshot below for Midomi’s results when I sang “Morning has Broken” into a microphone. At right, it shows that Cat Stevens was the original singer, and lets me buy the song at Amazon.com. At left you’ll see the profiles of the people who sounded most like me. At bottom, it gives other possible results. I can click on their profiles, listen and message them.
Midomi blows away Nayio Media, the San Mateo company that launched a humming search last month. If Nayio recognizes your tune, it cross-references with Napster’s library of songs. But the trouble is the recognition part. We tried humming “Morning has Broken,” but the song wasn’t represented on the first page of results. On Midomi, it was the first result, even when we hummed. Midomi gets better when you use words.
It was launched by two engineering PhDs from Stanford, Keyvan Mohajer (who studied music recognition), and Majid Emami, along with Stanford science grads James Hom and Michal Grabowski. The technology is self-built, called MMARS, or multi-modal adaptive recognition system. Midomi looks at pitch variation, like Nayio does, but it goes deeper, looking at tempo variation, phonetic content and location of pauses. If you hum a wrong note, it will often find ways to override it, by filtering it with other close signatures in its system. It isn’t perfect. While it recognized me humming “I Just Called to Say I Love you” immediately, it failed to do it on a later try when I hummed it quickly, without care. But once I added the words to my tune, it picked it up even when I sang quickly.
It gives you a studio, too, where you can pick out songs to sing — to help, it lets you play a 30-second track and see the lyrics — and lets you tag them, so that they are stored in the database. See a tour of the Midomi service here.
For copyright reasons, Midomi won’t let you upload your own songs, though that may change in the future. We did encounter a few bugs. Some sample tracks (it has licensed two million 30-second clips) didn’t play correctly (the audio didn’t come through). And one major caution: When we tried to download a software required before you can buy the songs, Midomi shut down the entire browser without warning. We still haven’t been able to download that software. We’ll update once Midomi fixes this.
The company’s goal is to become world’s most comprehensive database, Mohajer tells VentureBeat.
The Midomi service is their first product. They’ve formed a company called Melodis, and they want to move into other areas, such as speech recognition that will beat existing players like Nuance or Tellme. They’ve already implemented text recognition in Midomi. If type a search term “Ooops Britni Spirs,” it will recognize you are searching for the song by Britney Spears. Midomi also wants make the service easy to use for mobile devices.
The company has 15 employees in Sunnyvale, and another dozen developers in Eastern Europe and India. It has raised “multiple millions” from Global Catalyst Partners, and angel funding from Amidzad and former Googler, Aydin Senkut.
17 Comments
-
Brett said:
I don’t get it - why would I care to connect to other people, who song the same song, or who sung it in a similar way. That is trying to put community (Web 2.0)in for no good reason.
-
Richard Smalls said:
You know we are setting up another Silicon Valley bubble when a ridiculous company like this gets funding. This is just lame.
-
ASG said:
In fairness, I’ve met Keyvan and the Melodis guys and they’re extremely smart. The Midomi demo was very impressive and there are a number of attractive opportunities in the mobile music sector, as just one example.
-
ea said:
Forcing community on http://www.shazam.com/music/portal ?
They have been doing mobile for 5 years now.
-
LAila said:
I love this website and whoever said those bad commmetns about it is just weird.
-
anon said:
@Laila - coming from someone linking to their site…oh brother…poor saps who invested in this. Did Paris Hilton dump some “hard earned” cash into this one?
-
Mark Cramer said:
Hmm, I wonder where they get the music from? Is it licensed or unlicensed?
Something smells funny about this…
-
Matt Marshall said:
Yes, it has licensed 30-second clips from 2 million songs (I mentiond this in 4th paragraph from bottom).
-
Kim Seluga said:
I think I may be a bit tone-deaf since I tried to get the site to recognize 6-7 songs but to no avail
-
Bjorn said:
this is fascinating technology. as a music listener, i have always wondered how else how i could find the song name of a tune i can only hum.
But i retain my scepticism for a humming search as a business. that does not mean these guys’ startup should be scorned upon as another bubble. without the current healthy VC climate, how else could you test this kind of useful service to its intended consumer audience without taking VC money? who knows, if it really helps sell songs, amazon, apple or the other music sellers would buy it. perhaps humming recognition could be the new pre-purchase song sampling at record stores.
-
Erik Schwartz said:
Yet another company trying to get a small slice of the crappy margins in the music business.
The most successful music store in the last few years is ITMS. ITMS has no interest in making a profil selling music, they exist to support the 50% gross margins apple makes selling iPods.
-
Ved said:
I think users would like to see audio search functionality eventually gets incorporated into Google and yahoo search engines.
-
MSG said:
This is a FANTASTIC idea, if it would WORK. I tried humming several songs, and the only one it got was the Star Spangled Banner. WTF?? It couldn’t even recognize a simple Chopin melody, let alone a popular Gn’R song. This idea would be wildly popular, I believe, if it worked atleast 75% of the time. Too bad, cause there’s a song that Snoop Dogg and DMX did a while ago that I can’t find the name of.
-
Brett said:
To Anon’s point Laila’s link is right back to their site. I guess the only positive thing one can say is that their fake chat board postings are so badly done that there is no real attempt to hide the association.
BTW I love the idea of being able to search but the community aspect is babble - I agree with Ved’s thoughts that the search engine integration might be the big idea here. The old sell out to Google.
-
Gene Linetsky said:
Looks like it’s already becoming yet another collection of “this is me” pics (and now “this is me singing” clips). No matter what you do, kids will find a way to MySpacize it :)
On a serious note, melody extraction, indexing, and search is very hard. The state of the art (and there’s a lot of ongoing research on this, believe it or not) is very far from 75% precision. There have been many attempts to do QBH (query by humming), and there are dozen of QBH sites that are still live. None work well.
-
Sean said:
I think many of you are missing the point of the community aspect. It’s not just social networking for the sake of social networking: user contributions power the search engine. For the same reason, it wouldn’t be effective just to slap the search algorithm onto a Google application. You need a massive database of songs to bring search success rates up, and amateur, a capella, varied user contributions are better than licensed samples, because they’re far closer to the search inputs (not to mention cheaper for the company).
This aspect of the site is essentially free karaoke on a massive scale, and even if you don’t understand the appeal of that, you must admit that there are many who do, since karaoke remains a popular pastime.
The other aspect of Midomi is a user contribution-driven reference site like Wikipedia. However, it’s better than Wikipedia in at least two ways: one, people are motivated to contribute for reasons beyond basic altruism — it’s fun, and they can share their singing with peers and receive feedback and recognition while building the search engine’s database; and two, poor renditions of songs don’t damage the search engine’s effectiveness (as incorrect information on Wikipedia harms its usefulness and credibility), they just waste space on the company’s hard drives.
The community is far from arbitrary. It’s an integral part of the search engine, as well as a way for the company to provide two distinct services with one site.
-
tom said:
hum,hum,hum,i say humbug. that part of the programme is really irrelevant.if you do not know any songs then what the blazes you doing on that website?
all one has to do is type the song you wish to give ones rendition of and the technology comes into play for you to carry that out.on top of that one can chat about music etc. it is fun. sing your heart out and enjoy life, tomi.
2 Trackbacks
3:04 pm
[Question] Get Some Help With Tracking Down Those Hard To Find Songs - Page 6 - Allofmp3 Community Forums said:
[...] RedSmurf, forget the aligators. there plenty such sites available, one that come to mind is -> Hum along, and Midomi recognizes your song VentureBeat <-. this should lead you on. actually, the same people that invented/developed MP3s (Fraunhofer [...]
6:10 pm
Melodis matches $7 million with its music discovery service » VentureBeat said:
[...] catchy tune but not know the name of the song or the artist? Midomi, a web site and mobile service, lets you play a recorded song or sing a snippet of it yourself, and then it matches that information…. Then, you can listen to the recommended samples and go buy the song on iTunes or watch related [...]