The Echo Nest Blog

The Echo Nest Powers 'Resonate' from Ad Age, Spotify

Fri, 14 Jun 2013 14:54:00 -0400

Advertising firms like music — not just because it brings life to their creations for television, radio, the web, and mobile, but because they are people, after all, and people have ears.

In the spirit of friendly competition between ad agencies, the Ad Age trade publication for the advertising industry has launched a Spotify app powered by The Echo Nest that encourages the advertising community to set up profiles to showcase their musical taste — or even that of an entire firm.

To use it, you simply sign in with Facebook and answer a few questions, like which song you like to sing along to in bars, and a song that was played at the first concert you ever went to. Because the app hooks into Spotify for its music, it also has access to your listening history there.

And that’s where The Echo Nest comes in. Our Musical Identity technology is incredibly adept at figuring out who you are as a music fan, which is how the Resonate app manages to include four graphically-intensive charts at the top of every profile page in Resonate, powered by The Echo Nest:

Tempo Over Time (which depicts how fast the music is that you listen to over the course of the day)
Favorite Genres
Favorite Artists
Music Tastes (for individuals)
Company Compare (for ad firms)

The Resonate app looks great, with design from Portland, Oregon-based studio The Brigade; music from Spotify; concept from Ad Age; and profile stuff from Facebook, but we like to think that the charts, powered by The Echo Nest, are the neatest part of all.

The resulting musical social network for the ad industry will allow it to answer, finally, in the words of Ad Age, “What’s the most popular song at Starcom USA? How about Universal McCann? Or Droga5?”

That’s just the beginning of what the ad industry can now see about itself, firm by firm, and person by person. To up the ante, Ad Age will send out the top tracks of one agency per week in its Friday newsletter.

If you’re in the advertising industry, or you’re just interested in what it looks like when an industry builds its own musical social network, check out Resonatewith.me.

How We Understand Music Genres

Fri, 07 Jun 2013 12:00:28 -0400

Everyone from Laughing Squid to NPR is talking about “Every Noise At Once,” an ambitious exercise from The Echo Nest principal engineer Glenn McDonald that lets people explore arcane and general genres of music via an awesome word map. It’s an incredibly simple, deep way to explore all of music, as well as searching for bands to find out where they fall, and exploring the additional genre maps to see what bands in each genre sound like,from happy hardcore to indiepop and beyond.

The breadth of information in this chart is rather breathtaking — in fact, it channels a good deal of all the work we have ever done towards understanding the world of music. Yet the interface is a marvel of simplicity, masking most of the powerful stuff going on under the hood.

Let’s take a peek, shall we? How did Glenn fit the world of music onto a word map that won the internet? How do you teach computers to understand genre? Glenn himself explains:

A “music intelligence platform” ought to be able to play you some rock music. Hopefully it does more, of course, but let’s start there.

Our music intelligence platform responds in great depth to a wide variety of structured inquiries. Among many other things, this means we can ask it for the 10 hottest artists commonly described with the word “rock.”

This seems like a reasonable structured formulation of the question “what is rock music?”, and it takes advantage of our wide-reaching and sophisticated calculations of “hotness” and descriptive terms like “rock.”

Except here are the results:

Rihanna
Daft Punk
Justin Timberlake
Bruno Mars
P!nk
Taylor Swift
Macklemore & Ryan Lewis
Demi Lovato
Lil Wayne
Fall Out Boy

These are defensible answers to our query, because for all of them, “rock” is among the dozen or two most common terms with which people describe the artist, and we have insanely detailed data to prove it. But this is really not what you and I mean by rock music.

In fact, if we take out “rock” and just ask who the 10 hottest artists of any kind are right now, we get the same top 9, only swapping Pitbull out for Fall Out Boy. And if we switch the term from “pop” to “rock”, we trade Macklemore & Ryan Lewis for Pitbull, and keep Fall Out Boy.

These words don’t tell us enough on their own. The computers are doing their literal-minded best, but we need to ask them a different question.

When we say “rock music,” of course, we aren’t talking about term-frequency in a corpus of descriptive text, we’re talking about a kind of music. It’s an amorphous, evolving, imprecisely-delineated genre of music, to be sure, but still, if we were talking in person about this idea of rock music, we could straightforwardly clarify: “You know, rock music, man! Guitars, drums. The Stones, The Who, Led Zeppelin, that kind of thing.” Or maybe we’d say Nirvana and U2, or maybe we’d say The Allman Brothers and Lynyrd Skynyrd. Maybe we really mean classic rock, or album rock, or alternative rock.

We can argue about any of these permutations of genres and bands, but that’s an insight in itself: Out of genres, artists, and their cheerfully imprecise relationships, we can build a more accurate view of the world.

In fact, we can build a nearly infinite number of views of the world. Here’s one:

This is a map I made from outputs of the system we built to help our computers answer questions about genres of music in a way that’s much closer to what we, as music fans, would expect.

Now if you ask us to play some rock music, we might suggest:

Blue Öyster Cult “Burnin’ For You”
Queen “We Are The Champions”
Cheap Trick “Surrender”
Led Zeppelin “Black Dog”
Deep Purple “Highway Star”
Kiss “Strutter”
ZZ Top “Sharp Dressed Man”
Boston “More Than A Feeling”
AC/DC “Back In Black”
Lynyrd Skynyrd “Simple Man”

These might not be the exact same 10 songs you or I would pick by hand to define “rock,” but you and I probably wouldn’t pick the same 10 songs anyway. At least this is a far more plausible list than one with Rihanna and Daft Punk.

More importantly, perhaps, it’s more internally consistent. This is a coherent introduction to a conception of “rock.” And we can also now generate plausible and internally-consistent introductions to any number of other conceptions.

Here are 10 songs from our “alternative rock”:

Urge Overkill “Girl, You’ll Be A Woman Soon”
The Lemonheads “Into Your Arms”
Nirvana “Smells Like Teen Spirit”
R.E.M. “Man On The Moon”
Pavement “Cut Your Hair”
Weezer “Island In The Sun”
Built to Spill “Car”
Violent Femmes “Add It Up”
Soul Asylum “Runaway Train”
Meat Puppets “Backwater”

This, assuredly, will not end all arguments about what constitutes alternative rock. But I don’t think it’s an egregiously worse starting point for those arguments than whatever list you or I might personally propose.

Our computers can now enter plausibly into arguments over almost 500 genres, from a cappella to zydeco. Rock is the biggest and most central; we calculate a centrality score, of course, because that’s the kind of thing we do, so we mean that quantitatively. The least central genre is “skweee,” which most of us hadn’t heard of before this chart, either.

The most coherent — and thus hopefully least-debatable — genre is comedy. We don’t expect anyone to grouse that an artist we call “comedy” really plays tango or melodic hardcore. The most debatable genre is probably moombahton, a genre that could come as news to half the bands in it.

The genre with the least popularly-familiar artists is skiffle, which younger readers can be forgiven for thinking is a candy. Older readers might know that skiffle is why nobody would remember Lennon and McCartney if they’d kept playing it instead of forming the Beatles.

The calculations and machinations with which we build these genres involve layers upon layers upon layers of data-collection and synthesis, and a carefully considered (and mercifully manageable) amount of editorial guidance. For example, we decide what to do with naming variants like “nu soul” and “neo soul” (we went with “neo”), and whether we have enough data for the computers to produce a substantial and satisfyingly distinct body of music for any given thing, such as “indie folk” (yes), “sertanejo” (yes), or “ziglibithy” (no, not yet).

We almost never make up genres, but we could. With great power comes great responsibility. The approach allows us (or our customers) to seed, and then organically grow, a new genre or style from essentially any inspiration. In a couple peculiar cases, we’ve gathered an initial artist list, let the computers give us some songs, and only then listened to those songs to find out what kind of music we were even talking about. (Take a listen to the Brazilian country-music style “sertanejo”, which consists so overwhelmingly of male duos called something like “X & Y”, or in Portuguese, “X e Y”, that it can be checked pretty effectively purely by sight.)

The resulting system’s crucial, pragmatic quality is that it is dynamic and self-regulating on an ongoing basis. Bands appear in or disappear from genres automatically, as they come into and out of prominence or relevance. Rankings can change as often as daily. Genres scale automatically, according to our internal data-density and the artists’ inter-relatedness, so the more central genres like “rock” get more artists, while the more peripheral ones like “jug band” automatically get fewer, without anybody explicitly saying (or even knowing) whether a given genre is one sort or the other, or having to answer any existential questions about where a genre should end.

As data science, this is pretty unruly. There’s no imposed taxonomy of genres, and we have no objection to genres that overlap in small or even large part if they represent a subtle distinction that somebody, somewhere might care about (e.g. “gothic metal” vs. “symphonic metal” vs. “gothic symphonic metal”).

In the same spirit, any artist can be in as many different genres as apply. The genres aren’t even of the same sort: “tekno” is a very particular dance-music style, defined by tempo and historical circumstance; “wind ensemble” is a configuration of performers; “Christian hip-hop” is philosophical distinction; “Slovenian rock” a cultural and geographic one.

All of these are totally fine with us. Spend some time wandering around the map and you’ll get a better sense of how these genres vary.

You might also get a sense, perhaps more vividly than before encountering our map, of the contours of the overall space of musical possibility. Roughly speaking, genres at the top of the map are more electric, while those towards the bottom sound more acoustic. Genres on the left are sonically denser, the ones on the right sonically sparer and spikier.

There are other ways to plot music genres, of course, than the one used by the map at any given time. We use 10 dimensions internally, and two completely independent measures of genre similarity. I’ve flipped the map once already since first publishing it, and I might do so again without warning if I find another configuration that seems more interesting.

The point of the map, as with the genres, is not to resolve disputes but to invite you to explore music. It is an attempt — however uneven, idiosyncratic, and incomplete — to embrace this new state of the world, in which nearly all of humanity’s recorded music is streamable or downloadable, and give you a way to find out what you don’t know you don’t know.

Click any genre in the map and you’ll hear what we think is a representative song. These aren’t always ideal, but they’re close. Click the » next to a genre for a similarly-clickable audio-map of the artists we’ve extrapolated for it. Hit “scan” at the top of any genre page, and you’ll take a randomized car radio-style journey through that genre. Hit “scan” at the top of the main map and your car will careen wildly around the entire planet.

Or, if you’d prefer a more orderly, guided experience, links at the bottom of each genre page can take you to introductory genre playlists on Rdio. The “»” links for each artist go to their Rdio pages. The same Echo Nest data that powers this map also powers the “related artists” links in Rdio, so if the guided tour takes you somewhere interesting, you can always veer off onto your own path at any point.

Maps are, after all, as much machines for getting lost as they are for finding yourself. There are probably things on this map you’ve never imagined. It probably contains things that you don’t yet realize you love, and branching points where you will be amazed and thrilled to have veered.

I say this from dizzyingly-repeated personal experience.

I’ve long believed that music is what we humans do best, and the main lesson I’ve learned, after all of this exploring and mapping, is that I was right about that, but had wildly underestimated the magnitude of music.

Follow any path, no matter how unlikely and untrodden it appears, and you’ll find a hidden valley with a hundred bands who’ve lived there for years, reconstructing the music world in methodically- and idiosyncratically-altered miniature, as in Australian hip-hop, Hungarian pop, microhouse or Viking metal.

You might not want to abandon your old life and stay there with them forever, but you’ll go home knowing that there are other ways to live.

We make maps to mark treasure when we think treasure is rare, and then, later, to remember where we’ve been, once we start to realize that there are treasures everywhere. Eventually, these maps become something to do with our hands while we listen.

How We Resolve Artists on the Internet

Tue, 04 Jun 2013 15:45:00 -0400

The Web is full of text. Some of that text is about music.

People love to read text about music, but only if it’s about an artist they’re interested in. They tend to get annoyed if that text is about the wrong artist, and a mistake like that makes the service where the text appeared look silly and untrustworthy.

One thing we do at The Echo Nest is analyze the online conversation about music to actually understand the content. Every day, we index and analyze over 10 million daily new blog posts, social media discussions, and music reviews. We apply natural language processing and machine learning to contextualize all of this fan activity to understand music culture and identify music trends.

It might sound simple, but it’s not.

Consider this snippet:

“The founder of the ’70s band Boston has more than a feeling that his ex-guitarist is ripping him off… and that’s why he’s suing the guy.”

versus:

“Billy Borgioli’s long and storied history as a Boston rocker began in the late-1970’s with his tenure as the lead guitarist for the one of the greatest bands in Boston rock history: The Real Kids.”

The first one is about the ’70s rock band Boston, while the second concerns a ’70s rocker from the city of Boston. You could forgive someone for confusing or conflating them — and it’s no piece of cake for software, either.

Band names that also refer to common things prove the most problematic: Chicago, Train, Air, Atmosphere, Pink, and The Game all appear in our language far more than the band name versions.

So, how then do we filter out the noise? How can we know if something is about “Pink” the pop music maker or “pink” the color, let alone “Pink” the Victoria Secret lingerie collection? As with many things at The Echo Nest, our solution combines technology and curation.

Before we dive in, let’s define what we’re after. For a given piece of text from the Web, we want to find words that refer to music artists. We want to avoid confusion with common words and non-music-related entities. And we want to do it all reasonably fast, understanding tens of thousands of blog posts all over the world each day — a figure that potentially bursts to millions of posts.

This is really an entity extraction problem, with a bit of entity disambiguation thrown into the mix, but we treat it more like a classification problem: “For a given piece of text, assign that text to zero or more categories.”

Here, our categories are artists and as we’re fond of pointing out, we know about a couple million.

First, our algorithm picks out the artist names within the target text. We use Apache Solr for this, with an index of artist names and their variants. We developed custom Solr components to help us highlight artist names in a bag of text, and you, as a developer, can access that functionality via our public API by making an artist extract call.

We assume that a piece of text cannot be about a music artist if it does not mention at least one of the artist’s name variants. While not strictly true, we find this assumption works for our purpose here: to prune down our set of possible artists quickly and dramatically.

We typically identify between 5 and 20 artists in that way, depending on how much text we’re looking at. As usual, things get a bit more complicated from there — in this case because we see lots of playlist-like data embedded in our target text, and can thus find over a hundred possible matches. Even so, choosing appropriate categories from a few hundred artists is much easier than from our full set of 2.4 million.

Once we have identified the artist names in our target text, we check them for ambiguity. The goal here is to determine how much more work we need to do. Some artist names, like “Aesop Rock” or “P!nk” are quite unique, and their occurrence is enough for us to make a definite association. A piece of text that contains “P!nk” is almost certainly referring to the pop songstress Alecia Beth Moore. However, when her name is stylized as “Pink,” as often happens, it is ambiguous and prone to confusion with the color pink, the “Code Pink” peace activists’ group, Nicki Minaj’s album Pink Friday, Victoria Secret’s “Pink” lingerie collection, and more.

To determine the ambiguity of an artist name, we identify its frequency in a corpus of “general purpose” text — a set of news articles and movie reviews we’ve assembled, which we believe are not about music.

If an artist name appears with high frequency in this corpus of general-purpose text, we consider the name to be ambiguous. It’s worth noting this is the first place — but not the only place — where curation by people enters into the system. If we incorrectly guess the ambiguity status of a name, we correct that slip-up by hand, and then make the correction “sticky” such that future iterations of the algorithm remember how to treat that name.

For example, in our general purpose text we found the name “David Bowie” frequent enough that, according to our thresholds, the name “David Bowie” is considered ambiguous. However, in practice, mentions of the name “David Bowie” are almost never ambiguous. This false positive is a result of the same, multi-talented person appearing as an actor in films such as Labyrinth, The Man Who Fell to Earth, and Twin Peaks: Fire Walk with Me. Using manual curation, we have marked the name “David Bowie” as being unambiguous. We can, and have, applied similar corrections in the other direction — that is to say, marking names as ambiguous when they manage to slip through the algorithm.

If an artist name determined by our our ambiguity filter to be unambiguous appears in our target text, we assign those words to that particular artist, and continue on our way, looking at any additional artist name variants in the target text. However, if the artist name is considered ambiguous, we apply yet another step. We’ve now reduced the problem to a binary classification: Either “yes,” the text mentions the artist, or “no,” it does not. We then treat this as a supervised classification problem — and that is where our machine learning kicks in.

Here’s how that part works. For any given artist in our system, we have a set of text we can consider positive examples for our classifier. These include biographies, reviews, song titles, release titles, and the names of band members or collaborators — the presence of any of those indicates that our classifier has gotten it right.

And that, in turn, lets us build a baseline classifier using a simple heuristic. This heuristic approach works in an intuitive fashion: If, and only if, the target text contains mentions of terms strongly associated with the artist (e.g. song titles, release titles, band member names, etc), then the target text mentions the artist. The terms associated with the artist are passed through the same ambiguity filter, such that song titles such as “True Love” or “Brooklyn” are dropped. This approach tends to give us good precision, in that the pieces of text we keep are usually about the artist.

However, this method tends to give us poor recall. We drop lots of text that actually does mention the artist. The answer to this thorny issue is a bit of manual curation. We can (generally) beat the heuristic approach using a naive Bayes classifier — a classic spam filter implementation. Words (or “n-grams”) are used as features, and then Bayesian inference determines if a post is actually mentioning the artist or if it is “spam.”

We experimented with a variety of text classification implementations using the open-source scikit-learn library, python nltk, and a hand-rolled Bayesian classifier. We landed on a naive Bayes implementation that provides a good mix of performance, speed, and transparency. It enables us to look at which features the classifier thinks are most important, and get an idea of what kind of additional training we might want to do. As our human curators add new training data, our results here get better and better, and the cycle continues, adding to the intelligence of our database, the largest about music in the world.

If you’re interested in seeing some of this in action, register for an API key today and use it to see the results of some of our blogs and news calls. At the moment, the artist extract call only applies the first step of our entity extraction process, so only the Solr-based name highlighting is applied. In the future, we might expose an API call that applies the entire system described above, which is currently available only to our internal use — and that would allow outside developers to extract ambiguous artist names from any arbitrary text.

For some instant gratification (and some good old-fashioned json), check out the blog posts we’ve found for Boston, fun., or Pink. It probably won’t take you long to find some false positives (and as mentioned above, we have great ways of dealing with those), because we’re always improving our models, and adding more training data, but for the most part, you’ll see that the results really are about even ambiguously-named artists, and music services, apps, projects, and products can rest easy, safe in the knowledge that they are including news and blog posts about the the right artists.

- Kurt Jacobson, Senior Knowledge Engineer, The Echo Nest

Listen to 'Every Noise At Once' on NPR

Wed, 22 May 2013 10:48:00 -0400

Mark Fraunfelder, the founder and editor of BoingBoing, one of the most widely-read blogs in the world, appeared on National Public Radio’s Bullseye with Jesse Thorn yesterday to talk about two of his favorite things of all time.

One of them was Every Noise At Once, created by The Echo Nest principal engineer Glenn McDonald. This intriguing web app leverages our data to let you hear any of 467 genres of music, some of them impossibly arcane, by simply mousing over the words.

His creation also lets you search for bands to find where they fall, and then hop over to a whole map of bands in that genre (exhibit: chillwave).

To hear Mark Fraunfelder tell NPR about Every Noise At Once, powered by The Echo Nest, skip directly to 1:32 in the clip below. The whole segment is worth a listen if you have five minutes:

Music Hack Day Philadelphia Was Delicious

Mon, 20 May 2013 15:02:34 -0400

We organized Music Hack Day - Philadelphia this past weekend along with Drexel University’s ExCITe Center, sending a full crew of Echo Nesters down to help people build stuff with our API, make sure the event ran smoothly, and of course enjoy some delicious cheesesteaks.

The Echo Nest awarded prizes to the following tasty hacks among this hackathon’s 25 awe-inspiring creations:

1st Prize: Multivoice Jazz Companion

Philadelphia is a city rich in musical history, but enough about ’90s space rock — it’s also a longtime hotbed of the jazz scene. Designed by Julie Borgeot and Ray Migneco for use by musicians trying to work out how to play particularly difficult parts, Multivoice Jazz Companion for iPad, uses The Echo Nest’s Analyze data to figure out exactly what’s going on in a particular track, and outputs a piano roll-style map of the notes, with the instruments represented separately (thus “multi-voice”).

It’s not for everyone, but it’s pretty amazing.

2nd Prize: Are You Happy Now?

Our platform knows more about music than any other database in the world. Are You Happy Now? is a fun example of what happens when you mash that intelligence against an idea, such as “how happy or sad is the top song in the country trending on every May 20th over the past 72 years?”

As it turns out, our favorite songs from each year have been getting progressively happier, although the details do reveal some anomolies. Whether that means people in general are happier is another question, but at the very least, we’ve been preferring happier music over time. Are You Happy Now?, created by Benjamin Rose at Philly Music Hack Day, uses The Echo Nest’s valence and energy data to rate the top-charting hit from each May 20th from 1941 to present

The Echo Nest organized Music Hack Day - Philadelphia with the exCITe Center at Drexel University.

Sponsors included ExCITEe Center, The Echo Nest, Spotify, SoundCloud, tokbox, Rdio, musiXmatch, Gracenote, Fame House, SendGrid, MailChimp. Participating companies included The Echo Nest, Spotify, SoundCloud, tokbox, Rdio, musiXmatch, Gracenote, and Free Music Archive.

Congratulations to all of the hackers who made this weekend such a stunning success, and to the ExCITe Center — and finally, be sure to check out all 25 hacks, if you haven’t already, because there was plenty of great stuff — everything from an emotionally-appropriate tweet singer to radio stations based on the personalities of the characters from Arrested Development.

Photo: Matthew Santiago

The Echo Nest Partners with SeatGeek

Mon, 06 May 2013 13:33:00 -0400

SeatGeek, a ticket search engine, and The Echo Nest have partnered to incorporate SeatGeek artist IDs into The Echo Nest’s Rosetta ID mapping layer making it possible to use SeatGeek IDs directly with the Echo Nest API.

This makes it easier for you to use the SeatGeek and the Echo Nest APIs together. For instance, you can call the SeatGeek API, get performer IDs in response, and use those IDs with the Echo Nest API to present more context about the artist.

Here, we make a call to the SeatGeek API to get the set of top artists that are performing nearby:

http://api.seatgeek.com/2/events?geoip=true&range=50mi&sort=score.desc&taxonomies.name=concert&per_page=10

Each returned event includes a list of performances, like so:

performers”: [{”name”: ”Rolling Stones”,”short_name”: ”Rolling Stones”,”url”: ”http://seatgeek.com/rolling-stones-tickets/”,”type”: ”band”,”image”: ”http://cdn.chairnerd.com/images/performers/2597/rolling-stones-baba31/huge.jpg”,”primary”: true,”slug”: ”rolling-stones”,”score”: 0.80756,“images”: …”id”: 2597}]

We see the entry for the Rolling Stones. The SeatGeek ID for The Rolling Stones is 2597. We can use this ID directly within The Echo Nest calls; to get The Echo Nest info on the Rolling Stones using the SeatGeek ID, we can make an artist/profile call like this:

http://developer.echonest.com/api/v4/artist/profile?api_key=N6E4NIOVYMTHNDM8J&id=seatgeek:artist:2597&format=json&bucket=biographies&bucket=blogs&bucket=familiarity&bucket=hotttnesss&bucket=images&bucket=news&bucket=reviews

To show off the integration of SeatGeek and The Echo Nest, we’ve built a demonstration app that shows a list of top SeatGeek concerts (generated via the SeatGeek API) for any city. The app allows you to play music by the artist (via Rdio) and shows you the biography of the artist.

The application is live at this SeatGeek Demo, and the source code is on github.

The SeatGeek API is quite easy to work with. No API key is necessary. There’s lots of good data, including artist images, and an easy to understand TOS. See the SeatGeek page in The Echo Nest Developer Center for more info on the SeatGeek / Echo Nest integration.

The Echo Nest's Global Expansion Goes Nordic: WiMP and More

Fri, 03 May 2013 12:41:37 -0400

When we closed our latest round of funding, we said we were excited about expanding internationally to share our industry-leading music intelligence with music fans all over the world.

On that front, we’re happy to announce partnerships in the digital music stronghold of Scandinavia and surrounding countries, beginning with…

WiMP’s New Radio and Discovery Features

Norway’s WiMP announced a major upgrade to its iOS app for music fans in Denmark, Germany, Norway, Poland, and Sweden (more countries to follow).

Our technology powers the biggest upgrades to the app, including the ability to turn just about anything on the service into a personalized streaming radio station. Making any song or artist into a radio station is a fun, convenient way to discover your next favorite music makers.

“We believe this new feature will help our users discover even more artists and get tips even beyond our editorial recommendations,” said WiMP Global Editor, Thor Martin Jensen, about this Echo Nest-powered feature. “It’s all about convenience.”

We’re also the brains behind WiMP’s new Discovery Radio feature, which builds a bespoke station based on each person’s favorite tracks and artists, mixing in a healthy dose of new releases to combat music collection stagnancy.

More from The Northern Lands, powered by The Echo Nest

Norwegian Viamo, has apps that let users check in to radio and television programs to receive more information about whatever they’re watching or listening to on their “second screen.” They’re using our smarts to provide detailed information, including biographies and real-time news feeds, for each artist.
Xite, based in The Netherlands (granted, technically not “Nordic,” but nearby) is using The Echo Nest to create dynamic playlists that react to user behavior to keep them happy. It leverages our Taste Profiles to understand each of its users — how musically adventurous they are, among other things.

There’s a lot going on for us in this part of the world, with more to come there and elsewhere. To keep tabs on our latest developments stay tuned.

How We Cope with Spammers, Fakers, and Cloners

Fri, 26 Apr 2013 14:50:20 -0400

Here, The Echo Nest Senior Software Engineer Aaron Mandel explores the sneaky techniques used by musical spammers to “game the system” in music services — and how we stop them from succeeding.

—

The Echo Nest knows approximately 2.4 million artists as part of our database of music information, which is the largest in the world. However, we also keep a list of artists to ban from our system intentionally, so they never get recommended on any of our clients’ services, or in their apps.

This isn’t a matter of taste (so Coldplay and Raffi are in no danger). It’s because those banned artists are spammers.

Musical spam is much less familiar than email spam, but it works the same way: If it’s too hard to find the 10 people who might enjoy a shady or questionable product, spammers go for sheer volume, in their attempts to spoil your online experience with unwanted email or music.

Let’s take a tour through the tawdry world of musical spam, including a few exceptions that we choose not to ban.

If you want to follow along on our tour, listen to this Spotify playlist, which includes most of the tracks mentioned in this post. Be warned, though, that just like pop music and spam, this playlist may contain taboo words.

Even if you don’t mind a bit of bad language, you might want to listen on headphones anyway, because a lot of this music is terrible.

Okay. Ready?

Those moccasins someone else has been walking in*

The very best type of music spammers — the ones whose music elicits the best mix of hysterical laughter and retributive threats when you play them for friends — are the cloners. These groups record their own versions of popular songs, replicating the originals as closely as possible with whatever time and talent they have. (Spoiler: Often, that’s not very closely.)

These cloned songs are credited to “artists” such as The Hit Crew, Hip Hop’s Finest, #1 Hits Now, DJ New Release — names that could, and often do, pass for compilation titles. They might be named after the very song they’re cloning (“Call Me Maybe,” “Thrift Shop”) or a lyric from it (“Here’s My Number,” “Party Rock Is In The House Tonight”). The name doesn’t matter, so long as it’s close enough to fool people into clicking on the track without thinking twice.

A handful of cloners have managed to introduce their music into digital distribution networks with the original artists’ names attached — the ones they are ripping off — and that is very bad news.

If you stream music (who doesn’t?), I recommend investigating some cloners for yourself. Pick a recent popular song that you know well — the kind that you can identify after just a few notes. Search for it by its title on a streaming music service, without the artist’s name, and look for the aforementioned generic artist credits, or song titles with notations like “as performed by Macklemore” or “tribute to Macklemore.”

Then hit play. (This is fun whether you like the song or not, but choosing one you do like might protect your sanity.)

You won’t have to explore many of these before you find something awesomely Just Not Right — a clone of a song based around a single big riff which gets that big riff completely wrong (like The Hit Crew’s clone of “Gangnam Style” or Soundclash’s clone of the Caesars’ “Jerk It Out”), or one where the vocalist gives up on faking their accent partway through the first verse, like the band Call Me Maybe’s clone of the song “Call Me Maybe.”

The most amazing one that I’ve found to date is Charts Hits 2013’s clone of Macklemore’s “Thrift Shop.” It features a vocalist who makes no attempt to sound like Macklemore, and even so, he’s in way over his head. Clearly, the guy is reading from a lyric sheet. He says “mezzanine” like it rhymes with “nine” and says that he’s draped in a “Leonard mink.” The horn riff is also agonizingly squared off, with every note played at exactly the same volume.

This “Thrift Shop” clone is on the Spotify playlist of musical spam we mentioned above, but really, it’s way too good to miss, so here’s a YouTube link too.

Okay, so… where do clones come from?

Some are made by openly proud cloners. Drew’s Entertainment created The Hit Crew and other clone artists. Their website will happily explain to you all about how their CDs are a convenience for party hosts and DJs who want to hear their favorite songs without regard for authenticity or quality (not the exact wording).

Drew’s Entertainment has even landed some of their releases on Billboard charts recently, and — credit where it’s due — cheerfully includes Billboard’s befuddled quote on its website (above in blue).

That’s an exception to the rule. Most of these cloners are secretive, so we have to guess at their motivation.

Some clones are sold alongside karaoke versions (i.e. same track, no vocals), suggesting that the cloner’s main business lies in selling karaoke backing tracks, and the clones are just a byproduct of that — a case of “Why not?” Maybe having someone sing a scratch track helps keep the band together while they record, and these cloners simply decide to release both versions.

Others seem to appear when a hit song hasn’t been officially released in a particular country, so a clone steps in to fill the gap in the market. This type has a long history, going back at least as far as the fake Beatles records that hit America before the originals had been issued here.

There’s more. Let’s keep moving.

“Happy Birthday, Keanu!”

In terms of the sheer bulk of the spam they produce, the worst culprits when it comes to musical spam are “personalized music” artists, who make a practice of releasing hundreds or even thousands of the same song with various names spliced in.

The group responsible for these “Happy Birthday, [Your Name Here]” songs, Birthday Song Crew, apparently didn’t want to shell out licensing fees for the familiar “Happy Birthday” song, nor risk legal action for using it without permission.

Instead they simply wrote some new tunes with deceptive titles “Happy Birthday (Reggae)”, “Happy Birthday (Jazzy),” and “Happy Birthday (Hillbilly),” each of which is a depressingly bad genre pastiche. They are not the “Happy Birthday” you’re looking for, and on top of that, they’re musically awful.

You may also notice in the screenshot above that some names appear twice. That, as far as I can tell, is because the “Happy Birthday, Jerome!” at the very beginning of the song has a slightly different intonation on some tracks, and the duplicated names are the ones where both intonations are available.

Some of these artists, especially the ones aimed at parents, claim to be personalizing entire albums for your special child. They aren’t. They have one personalized birthday song, followed by a dozen tracks that are identical to what every other kid is getting.

The Birthday Bunch and The Teddybears, among others, are guilty of this. Maybe their full-album approach comes from the days when CDs ruled the roost; if you were buying hard copies of these albums, your kid’s name would be printed on the front cover, in addition to being pasted into one track a few times. So there’s that.

To be clear, The Teddybears are spammers. Teddybears, also known as Teddybears STHLM, are not; we love them and have been known to blast “Get Mama A House” on the office stereo. We do not particularly love, but are also not going to ban, The Teddy Bears (note the extra space), a doo-wop group best known for their 1958 hit “To Know Him Is To Love Him,” which a Google search just told me featured a teenage Phil Spector. They’re fine.

The process by which these spammers choose their names appears to be manual, even if the recording is more, shall we say, automated. Birthday Song Crew have five or more tracks called “Happy Birthday Jay-Z,” even though Jay-Z has not yet cracked the top 1,000 baby names for boys. At least here in the U.S.

Money for nothing and your clicks for free*

What could be easier than personalizing music by splicing various names into the recording?

Think about this one for a second before reading on, because the answer is pretty brilliant.

That answer: Personalizing music by NOT including names in it. Or any vocals at all, for that matter.

Silhouette is a shadowy figure responsible for hundreds of copies of the same instrumental track, completely unchanged, other than the name:

Lest anyone misunderstand this spammer’s intent, a smattering of the tracks have “(Dedicated to My Love)” added to the end of the title. They also appear to only name this track after women. One might wonder whether that’s more insulting to: women, or people who date women.

One might also wonder why “Doris (Dedicated to My Love)”, “Yvonne”, “Maite” and “Beyhan” — and none of the others! — are marked “Explicit” by some services. But too much wondering can corrode the spirit, so let’s leave that alone for the moment.

Today, I don’t feel like doing anything*

What if you have already convinced all the people you know that you are utterly devoted to them and/or are aware of their first names?

Is there anything left to achieve in your life that music spammers want to help you with?

Well, yes, obviously.

Brainwave Mind Voyages is one of several artists in digital music catalogs who, rather than music, sell the promise of subliminal upgrades to the human mind. These are long white-noise (or even silent!) MP3s that supposedly teach you things, change your habits, or make your teeth grow while you hang out and listen. I’m not sure if they’re supposed to work while you’re asleep, though that would probably be more convenient, if no more effective.

We got our hands on several MP3s from one of these outfits, once upon a time, and determined that they were all bit-for-bit identical. In other words, the same white-noise MP3 would subliminally teach you to speak English, Spanish, French, German, Tagalog, Swedish, Lojban, or hundreds of other languages, while also making you a successful CEO, a successful computer programmer, a lottery winner, and the life of the party. Brainwave Mind Voyages also offers to help you with card counting, thinking outside the box, time travel, and tooth regeneration. Among other things.

It’s worth pointing out that, as with birthday music, even people who want to listen to these tracks on demand presumably do not want them showing up in radio playlists. We are more than happy to help with that.

I wanna rock and roll all night* and then for 200 additional hours

Even among non-cloners, some spammers actually make music. A lot of music. Like this:

Why Not is a person or band (hard to tell) that has released hundreds upon hundreds of albums of unremarkable psychedelic rock. All of those albums consist of the same few hundred songs, repeated in various combinations. No two albums are the same, but they all draw from the same pool of material — a large pool, but not enough to support hundreds of albums without including the same songs over and over again.

Perhaps to mask this, most of these albums have titles suggesting greatest hit anthologies, or other compilations. A dozen or two are named after countries: Rock Over Peru, Rock Over Norway, etc. I haven’t been able to discern anything particularly Norwegian about Rock Over Norway, though maybe I missed it.

Then there are the eccentrics.

Sir Juan Mutant produced an epic-length discography that recycles the same tracks many times, in one instance giving the same track nine different names, on the same album. (The album, Cash The System, is 11 hours long.) This artist also frequently reuses the same title for different tracks, adding to the confusion and repetition:

The Echo Nest does not always ban these faux-prolific, song-cloning artists. They seem to be creating original music, and it’s not always possible to distinguish objectively between a band like Why Not and, say, Rick Springfield or Fleetwood Mac, both of which have (re)released the same songs on a staggering number of anthologies and reissues.

There’s also little danger of our deeming Why Not to be similar to other artists, so he’s pretty much out of circulation to begin with. Nobody is talking about him online, let alone using terms that would link him to music that people really are listening to. He’s benign, in our grand scheme of things, so we leave him alone.

Stop — Hatin’ Is Bad*

Why does the music clone issue matter, and what are we doing about it?

The main reason it matters, as mentioned earlier, is that it’s very easy to predict whether a listener wants to hear any of these tracks. The answer is almost always “no,” so we want to leave spammers out of recommendations and playlists — particularly the ones who impersonate music that people really do want to hear, because that’s extra annoying.

Another reason it matters to us is that our systems are always optimized for real-world music. Hundreds of near-identical releases could potentially slow or confuse our system; if it were music we needed to know about, that would mean it was time for some clever engineering to solve the problem — which we are perfectly willing to do when the situation calls for it — but for spam, why even bother? We leave most of them out.

When it comes to finding spam in order to ignore it, our tools vary by the the spammer’s technique.

In some cases, pure track counts are enough to bring someone to our attention. Spammers The Birthday Bunch have a catalog comparable in size to Frank Sinatra or Johnny Cash, which raises a flag.

In many other cases, our audio fingerprinter tips us off, whether because it tells us that tracks with wildly different names are actually the same music (as with Sir Juan Mutant or Silhouette) or because a big-name artist is weirdly associated with one, very unpopular song (as with the cloners who manage to attach their work to a legitimate artist’s name).

The process isn’t fully computerizable, only because new types of spam are always appearing. I haven’t even gotten into the horrors of:

joke ringtones (such as the artist Comedy Ringtone Factory Funny Ring Tones, Phone Humor — despite the comma, that’s a single entity);
identical new-age instrumental albums sold under different names as baby brain music, yoga background music, and a romantic soundtrack; and
quick-buck compilations of popular songs re-recorded as dubstep, “workout mixes”, or piano bar music.

Nor has this post touched on bizarre, one-off cases like:

the strange case of Deborah Weissbuch, who, as far as we can tell, simply uploaded Beatles recordings as her own; and
Slim Shady, the rapper who borrowed Eminem’s alias and song titles for totally non-Eminem songs. He is not the real Shady.

I could go on here, but there are only so many hours in the day, and besides, some of us have already spent hours compiling lists of our favorite Sir Juan Mutant song titles, such as “Did You Put That Man On Fire” and “I Mean Sorcery And All That.”

Have fun. And happy birthday, whoever you are.

The Echo Nest Thoroughly Enjoyed Music Hack Day Paris

Tue, 23 Apr 2013 15:56:54 -0400

Music Hack Day Paris took place this last weekend and The Echo Nest was there, of course. Not only are we a founding sponsor of the event, but… come on, it’s Paris.

Ellis presented our API (above) at Deezer headquarters, which hosted the event, and which provided an excellent assortment of food including at least six varieties of pastry. On to the hacks!

Approximately 120 attendees built 32 projects. More of them used our API than that of any other company, which is becoming something of a trend at these events. Hackers also had the chance to build stuff for the humanoid robot invited to the event by our co-founder and CSO Tristan Jehan.

The Echo Nest awarded prizes to…

Jigsawng: This original, well-implemented iOS puzzle game asks the user to rearrange music chunks of a song in the correct order, similarly to how a picture puzzle game works. It was original and pretty nicely implemented. It should be available in the iTunes app store soon.

This Is My J: Inspired by This Is My Jam and Twitter’s Vine short video sharing app, This Is My J brings newfound brevity to the social music scene, allowing you to share only the most “interesting” six seconds of any song as determined by our Remix API.

We were also really impressed by…

Podmapper: Designed for people who like to listen to music podcasts and DJ sets but don’t always have time to listen all the way through, Podmapper analyzes podcasts and mixes to find the songs contained therein, and translates them into a neat and tidy Rdio playlist.

WoMEn (World Music Energy): Based on the logic that GDP is a subpar way to compare countries against one another, WoMEn analyzes the energy levels of music listened to by people in various countries. The result is available in map form.

Brutalize Me: Not for the faint of heart, Brutalize Me is designed to destroy pop music by turning the singer’s vocals into a death metal growl, or by inserting heavy metal instrumentation. “Brutalness” settings include light or deadly.

Tunopoly: This nicely-designed iOS game tests your knowledge of electronic music on tiled board consisting of album covers. Players move around the board, as in Monopoly, answering questions. The first player to the center of the board wins.

RemixStation: Connect a MIDI keyboard to this interactive remix app, and you’ll be able to play various music segments, reassembling them into a complete song.

On top of those, we have Tristan’s own hack…

The Soulizer: A revamp of his video-based visualizer, The Soulizer first creates a highly-danceable, beat-matched DJ mix out of any selection of songs using our Remix API, but that’s just for starters. It then dynamically resamples “Soul Train” videos from YouTube to make them sync up with the audio. Try it at your next dance party.

(Photo: Tristan Jehan, The Echo Nest)

The Echo Nest Teams with SiriusXM for MySXM

Mon, 15 Apr 2013 10:06:57 -0400

SiriusXM Radio has more paying subscribers than any other music service. We’re fired up to be powering the latest enhancement to SiriusXM Internet Radio: MySXM.

MySXM combines SiriusXM’s world-class music programmers and DJs with The Echo Nest’s Taste Profiling and Dynamic Playlisting engine. We’re confident this is the most sophisticated combination of expert human music curation and data-driven music discovery the industry has seen.

When users first start listening to a MySXM channel, they hear programming similar to the SiriusXM broadcast channel. But with MySXM, any listener can take the reins and easily personalize what they hear thanks to the power of The Echo Nest.

For over 50 SiriusXM channels, listeners can customize the programming with easy to use sliders. Moving the sliders creates a new personalized version of the channel, based on SiriusXM’s expert programmers and the user’s own taste. For example, say a user is listening to “80’s on 8,” but they want more early 80’s dance tunes. A few slider adjustments and they’re there.

The combination of professionally curated channels and the ability to fine-tune the content gives listeners an entirely new level of control. And, users can reprogram the channel with the sliders at any time, offering a degree of control that is impossible with a traditional or satellite broadcast.

For a slightly deeper look under the hood, our Taste Profiles are building a deep (anonymous) understanding of each MySXM user’s music preference and listening trends, while our Dynamic Playlisting engine helps to generate these curated and personalized channels based on hundreds of musical attributes, cultural attributes and custom radio programming rules established by SiriusXM’s expert curators.

With the launch of MySXM, the programming knowledge of SiriusXM’s experts becomes dynamic, allowing SiriusXM to launch a highly-interactive, personalized internet radio service that maintains the music programming style that helped SiriusXM build the largest subscription music service in the world.

The collaboration between SiriusXM and The Echo Nest has been one of the most interesting, innovative and eye-opening partnerships we’ve entered into. After working with SiriusXM’s music programming and digital product team for over a year, we’re confident this approach will re-shape how music services think about personalized radio.

(screenshots via siriusbuzz)

The Echo Nest Wins '50 on Fire' Award

Fri, 05 Apr 2013 12:58:53 -0400

The Echo Nest is a global presence, with offices in San Francisco and London and customers everywhere.

Our headquarters are in Boston. As such, we’re happy to announce that our city’s tech publication of record, BostonInno, announced its “50 on Fire” list for 2013, honoring “the city’s hottest game-changers, disruptors, and innovators,” and we’re on it.

The committee considered the Advertising, Arts & Entertainment, Design, Dining, Education, Fitness, Media, Marketing, Retail, Sports, Healthcare, Medicine, and Tech industries.

The Echo Nest joins Actifo (data management), Crashlytics (iOS/Android problem solving), Digital Lumens (intelligent lighting), ReThink Robotics (automated manufacturing), SessionM (mobile rewards), and UTest (software testing) as winners in the Tech section.

Boston is a great place to grow a business, and we’ve built a talented, passionate, world-class team here. It’s with no small amount of hometown pride that we accept this award.

The BostonInno 50 on Fire Awards

Microsoft, Spotify, The Echo Nest, DJ Team for MixShape Contest

Mon, 25 Mar 2013 20:31:46 -0400

Cassettes were great for making mixes, and they will always hold a place in our hearts. These days, the options for making and sharing “mixtapes” online go far beyond what used to be possible.

One of the most advanced methods we’ve seen is MixShape, a collaboration between Microsoft, Spotify, world-renowned DJ James Lavelle, and The Echo Nest.

Designed to showcase Microsoft’s tablets and touch-sensitive computers, MixShape lets you take any of your Spotify playlists — up to 300 tracks — and analyzes them using our technology, sorting the tracks by their acoustic properties to create a mix with the optimal flow between songs, depending on what kind of playlist you want (romance, party, work, or exercise).

Then, it displays a beautiful graphic that you can actually touch, so you can reorder and “reshape” the mix with your fingertips or your mouse:

MixShape launched today in the UK, but it’s accessible wherever Spotify is used.

This is a fantastic new way to create a top-notch mix out of your favorite music, but it’s also a month-long contest. DJ James Lavelle, who has created four MixShapes of his own, will judge the UK entries to find the best playlist. The winner will receive a Windows 8 tablet, Xbox 360, Windows Phone HTC 8X, a year’s subscription to Spotify Premium, and a print of their MixShape autographed by James Lavelle.

We’re happy whenever someone uses our stuff to make something, and even more so when that thing feels entirely new. Dragging your fingers across a screen to shape the flow of an intelligently-created playlist definitely qualifies.

You can try MixShape here.

Reebok FitList: Spotify App for Getting in Shape

Thu, 21 Mar 2013 09:43:11 -0400

Exercise does wonders for your mood, health, and physique. The biggest barrier is getting motivated, and then staying that way.

Science proves that music helps people work out, and we all know we enjoy it more with the right music. The Reebok FitList Spotify app unites any exerciser with the perfect music for their taste, activity, and the length of their workout. It’s free, and it makes Spotify playlists you can use on all of your Spotify devices.

Powered by The Echo Nest’s deep musical intelligence about every song in the world, Reebok FitList for Spotify (Spotify link) includes a powerful playlist generator that lets fitness freaks and former couch potatoes enter the type of exercise they are going to do; the length of their workout; desired intensity; and a favorite artist upon which to base the whole thing.

Reebok FitList then instantly generates the perfect playlist for that person and what they’re doing, with high-energy music at the appropriate tempo. It includes music from that artist, but also similar artists, for just the right mix.

Any Spotify user can use Reebok FitList to create as many exercise playlists as they want, for free. Premium Spotify subscribers can sync the music to their smartphones — perfect for the gym, jog, yoga mat, or wherever else their quest for healthy living takes them.

Reebok recently launched another exercise app: Reebok Fitness, a workout planner with exercise videos.

ArtistX Offers Astounding View of Any Band

Mon, 18 Mar 2013 15:16:00 -0400

The Echo Nest director of developer community Paul Lamere put together a nice new app called ArtistX over the weekend. Enter any band name, and his creation, which runs on The Echo Nest’s data, presents you with graphs and other science sure to amaze and amuse. Oh, and it also lets you listen to everything on Rdio.

This is a guest post from Paul’s Music Machinery blog:

—

There’s no hackathon this weekend, but that’s no excuse not to write some code. I’ve been wanting to experiment with amcharts, a Javascript charting package so I wrote a web app that shows lots of charts and graphs for artists. The app is ArtistX. It is an artist explorer that lets you look at all of the Echo Nest song parameters for any artist. For instance, you can look at the Energy Distribution of songs by Weezer:

You can look at the tempo distribution of songs by The Rolling Stones:

Or you can look at scatter plots that show 4 attributes at once (X, Y, size and color). Here’s a plot of all of Muse’s songs showing the energy, loudness, hotttnesss and liveness:

You can interact with the plots – click on a bar or point in a plot to listen to songs (via Rdio).

The app lets you explore across 11 different song parameters: energy, loudness, danceability, liveness, speechiness, hotttnesss, tempo, duration, key, time signature and mode.

You can use the app to find all sorts of interesting things. Want to listen all the stage patter for an artist? Create scatter plot for the artist with liveness and speechiness as the X, Y parameters. The songs in the upper right-hand corner of the plot will be the ones you are looking for. Try it with an artist like Elvis Presley or Dean Martin.

Give the app a spin here: ArtistX. The source is at github/echonest/ArtistX

Groovebug Concert Vault Delivers Thousands of Shows to iPad

Wed, 13 Mar 2013 17:50:53 -0400

Until someone invents a time machine, the best way to experience legendary performances of the past is through recorded audio and video. Out of today’s selection of devices, the iPad is arguably the best place to experience them.

The Echo Nest has teamed with Groovebug and Wolfgang’s Concert Vault for a fantastic new live music app for the iPad (screenshots below) that lets music fans experience thousands of performances by everyone from Jimi Hendrix to Bon Iver, from the comfort of their living room, commute, beach, or wherever they find themselves.

Groovebug Concert Vault includes live audio and video performances from over 3,500 musicians from the rock, indie, blues, jazz, country, folk, and bluegrass genres, with more added daily. For $3.99 per month or $39.99 per year, music fans can access all of these shows on multiple platforms — and now, the iPad, offering a great combination of portability and screen size. The app includes native Apple AirPlay too, so you can watch them all on Apple TV with ease, listening to the audio on your surround sound system.

To see whether the app is worth their cash, fans can sign up for a seven-day free trial, allowing full access within the app as well as on the website.

The Echo Nest’s technology is all over this app, powering similar artist suggestions and more — but perhaps most importantly, helping fans find the performances most likely to set their spines atingle from this large selection of high-quality recordings. It works by analyzing your listening history to figure out what you might want to watch or listen to next.

Do you want to see The Grateful Dead at the Winterland in 1977? Perhaps you’d prefer R.E.M.’s outstanding set at the Capitol Theater in 1984, St. Vincent rocking the Great American Music Hall in 2009, or Mumford and Sons stopping by Daytrotter last October? Groovebug Concert has all of that and much more. It’s a lot to choose from, but The Echo Nest’s technology assures that the app gets better the more you use it.

Nothing replaces the feeling of being at a show as it happens. But until that time machine gets invented, Groovebug Concert Vault for iPad is the next best thing.

Here’s the Home screen:

Here’s Jimi Hendrix live at the Fillmore (this show is audio only, but some contain video too):

You also get a bio of every artist, as well as news and similar artists.

Install Groovebug Concert Vault for iOS.

DoubleTwist Magic Radio Lives Up To Its Name

Tue, 12 Mar 2013 12:00:20 -0400

DoubleTwist Player, an incredibly popular Android music player and media synching tool, broke “two million unique users per day” last May and has been downloaded over 10 million times.

Following today’s launch of the Magic Radio feature, the doubleTwist Player app becomes even more powerful with a brand-spanking-new, ad-free streaming radio service sporting advanced, easy-to-use features you won’t find anywhere else.

Of course, it is powered by The Echo Nest.

Magic Radio, available as a feature within the latest doubleTwist Player Android app, is better than Pandora for the following reasons, according to doubleTwist:

“It’s bigger - much, much bigger: Magic Radio is powered by a repertoire of more than twelve times the number of songs on Pandora’s music library. This means you can enjoy more eclectic artists and encounter far less repetition.
No ads: Magic Radio covers the streaming costs and music royalties using an affordable monthly subscription, without intrusive banners or voice ads.
It’s tunable: With Magic Radio, you can modify a station on the fly by changing its tempo or the familiarity of the songs.
It’s personal: Magic Radio can look at the songs you have in your music library and uses them to tailor stations to your taste. It’s like a personal DJ.
It’s smart: With Magic Radio you can create stations based on words such as ‘Woodstock,’ ‘Coachella.’ or ‘Brazilian music.’ It’s magic!”

DoubleTwist Magic Radio can import the music on your phone, allowing The Echo Nest’s deep musical intelligence build an immediate “music I love” station that plays forever. You also get helpful defaults such as Relax, Discover, Hot Artists, and more, to get you started.

That’s just the beginning of what this powerful, next-generation radio app can do using the strength of our data. You can build incredibly specific playlists out of over 100 microgenres tracked by The Echo Nest - then fine-tune them not only with thumbs-up and thumbs-down ratings, but other factors like how familiar the music is likely to be, or your mood and energy level.

Another taste of this “magic”: You can build a radio station out of any regular doubleTwist playlist, pulling from 7Digital’s entire catalog. In other words, it’s your musical taste, extended over 22 million songs, effortlessly.

Regardless of how you build your stations, you can always add artists to them, and fine-tune them in all sort of ways, even as they play. On top of that, all of these stations are shaped by your own unique Taste Profile, so the more you use it, the better it gets at all of this stuff, where your specific ears are concerned.

Like they said: “magic.”

DoubleTwist Magic Radio is available to U.S. users for free, as a seven-day trial within the regular doubleTwist Player. Users can subscribe to this completely ad-free service for $3.99 per month.

For more, see magicrad.io.

The Artist's Hack at SXSW

Tue, 12 Mar 2013 10:22:00 -0400

Sunday at SXSW was the Artist’s Hack - where passionate developers from around the world gathered to build cool stuff. Artist’s Hack was organized by Backplane and Spotify and is dedicated to building the future of music, art, video and collaborative though on the web and mobile during SXSW. It is a first-of-its-kind hackday with a focus on bringing music, video, interactive, and any other type of artist at SXSW together with the music technology magic that never fails to emerge from these events.

The hack was held at Raptor House – a short walk from downtown Austin. There was plenty of bandwidth, good food and beverages for the 8 hour hackathon. APIs were in abundance: Spotify, The Echo Nest, SendGrid, Twilio, Youtube, Klout, Paypal, Gimbal, SeatGeek, Aviary, Etsy, Topspin, Chute, Dropbox, Music Dealers and others were all there in force offering their technology for hackers to use.

Hackers built around 20 hacks during the event. Some of my favorites are:

biomuse – creates playlists based upon your biometrics. This was built on top of the biobeats platform. Quite neat stuff. Winner of one of the Echo Nest prizes.
Jamblot – visualize your song history in a creative way to commemorate any period of your life that affected your music choice. Jamblot draws your song history for you. Winner of one of the Echo Nest prizes.
Party Together - ambient automatic shared playlists for your party. Winner of one of the Echo Nest prizes.
We browse in public – Stream all of your browser activity live to others. Chat with others based on their activity.
Bundio – Monetize dropbox.

My hack is A longer life for post-rock fans. This was my first time using the Twilio API. It was a lot of fun to build. The Twilio API and whole developer experience is awesome. Any company with an API should try to emulate what Twilio does.

One novel aspect of the event was that Cory Booker was one of the judges. Here he is watching Danny Kirschner give the Bundio demo

Cory is a pro – when there was a power outage that delayed some of the demos, Cory conducted an impromptu ’interview’ with one of the founders of Backplane while the crew scurried to restore the power.

All in all, the Artist’s Hack was great fun, with lots of creative hacks. Well done Spotify and Backplane! - Paul

Reposted from Music Machinery

The Echo Nest CEO Jim Lucchese To Talk Musical Identity at SXSW

Fri, 08 Mar 2013 11:38:37 -0500

SXSW kicks off today in Austin, Texas. As always, The Echo Nest will be there.

In his talk, “Musical Identity: What Music Taste Says About You,” on Thursday, March 14, at 12:30pm CT, The Echo Nest CEO Jim Lucchese will discuss a trend in music intelligence that we feel will come to define the next year in music apps: understanding musical identity.

One of the first selections by the SXSW committee, his talk will revolve around the question of what your music taste says about you — and how a better understanding of you as a music fan will enable better music experiences.

As many readers of this blog know, we’ve been working on understanding musical identity for a long time. We’ve posted about Taste Profile Attributes and even the connection between music taste and political affiliation.

Jim plans to answer the following questions next Thursday in Austin:

How are applications enabling people to share their musical preference and identity today?
How online music services and other apps (social discovery, dating and others) are applying an understanding of music your music taste to create smarter, more intuitive online experiences.
What else does your music taste tell us about you? Can it predict your gender, your politics, what about your taste in movies or books?
How will music preference play a role in overall social interactions online?
The creepiness factor — how to make apps smarter in understanding musical identity while respecting a fan’s privacy.

You can add it to your SXSW schedule here.

Tech and Music Join Forces for 'The Artist's Hack' at SXSW

Tue, 05 Mar 2013 14:50:20 -0500

As a founding sponsor of the Music Hack Day series, we’re always impressed by the hacks that emerge. Infinite Gangnam Style! Invisible Instruments! Echo Run!

Music Hack Days are great at bringing together coders and hardware wizards from all walks of life, and uniting them with the world’s best APIs around music, including, of course, our own. But one important community has been absent from most of them: the musicians!

They return each March like salmon to Austin’s SXSW, along with a healthy portion of the overall tech scene. It’s the perfect time and place for a Music Hack Day that loops the music community itself into the Music Hack Day equation.

Enter: The Artist’s Hack, a first-of-its-kind hackday with a focus on bringing music, video, interactive, and any other type of artist at SXSW together with the music technology magic that never fails to emerge from these events (latest examples).

The Artist’s Hack takes place on Sunday, March 10, at Austin’s Raptor House venue. Of course we’ll be there to present and support our API, which usually figures into a large percentage of Music Hack Day creations.

Artists and technologists need each other. The Artist’s Hack will help them meet up and make some cool stuff. Both parties can RSVP here. We look forward to seeing the results.

Echo Run Takes Jogging Apps to the Next Level

Fri, 01 Mar 2013 12:07:17 -0500

Tufts had a Music Hack Day last weekend, and The Echo Nest head of developer community Paul Lamere was in attendance. Every hack was interesting in its own way, but one struck him as particularly advanced, with possible commercial viability.

It’s called EchoRun, currently awaiting Apple’s approval for distribution in the iTunes app store, for free.

Why was he so impressed? Well, we’ve seen plenty of jogging apps that use the accelerometer in today’s smartphones to monkey with the music being played to facilitate a workout, but none goes so far as EchoRun.

This as-yet-unreleased iOS app adds gamification, smart playlist generation, audio manipulation, social elements, and more to the well-established concept that the right music can make exercise more effective — and, of course, it uses our musical data (which, as you might know, is the biggest database about music in the world).

Created by Tufts students Diego Carranza, Charles Holbrow, J. Foster Lockwood, Will Millar, and Stephan Panaro, EchoRun turns jogging into a much more musical — and competitive — experience.

We’ll let J. Foster Lockwood explain how it works:

“EchoRun provides incentive for those who find their morning run lacking,” writes Lockwood. “Using a user-created playlist in the default [iOS] music app, EchoRun analyzes each song, using The Echo Nest API, for tempo and energy levels. As you start running with EchoRun playing your music, it detects the rate at which you run at (via the accelerometer) and if you aren’t up to pace, it will slow down the song currently playing and fade out.”

“Once you’ve caught your breath and started running again, it will speed the song up to its original tempo and give you a cheer for recovering,” he added. “The website currently shows a live leader board with the names of who is running and what song they are listening to, as well as a score out of 100 representing their pace.”

So not only does EchoRun keep you motivated by manipulating music from your own library — even including a cheering section — but it lets you race against other users of the app, see how they fast they are, and what they like to listen to as they jog.

This was just one of the great hacks to emerge at Tufts. All in all, it was a great Music Hack Day.

“We know lots of students have interesting ideas, but with school work and everything else, it is hard to set aside time to work on them,” said Alden Keefe Sampson, one of the organizers of the event. ”We started the hackathon to give students a chance to get their projects off the ground, learn knew technologies, and find new friends to work with. We were very impressed by the creativity and quality of the projects that teams were able to create in less than 24 hours.”

And thus, another excellent Music Hack Day drew to a close — and another promising app made its way to the iTunes App Store.

Screenshots: