Old Vinyl by fensterbme

The world of online music is growing rapidly.  Every day, thousands of new tracks by new artists are uploaded to the Internet. Soon all recorded music will be online and available for you to listen to anywhere at any time, on any device.

However, this seemingly endless supply of new music creates a problem for a music listener. How can you find music that you will like when there are millions and millions of tracks to choose from?

In this blog, I write about the world of online music discovery and recommendation.  I look at the tools available to help people find music.  I examine some of the issues that can make music recommendations go bad.   I also write about things that I find generally interesting including programming, data visualization, playing games,  and (of course) music.

About the author:  Paul is the Director of Developer Platform for The Echo Nest, a music intelligence company located in Somerville Massachusetts.  en_logo_250x200_lt1Paul serves on a number of program committees including the International Society for Music Information Retrieval and UMAP 09. He is the Industrial Co-Chair of RecSys’09.  You can reach Paul by email at Paul [at] EchoNest.com.

Music Machinery banner image by Maschinenraum (cc)

  1. #1 by Joel on March 5, 2009 - 11:43 am

    I read your blog about the drummers that use clicktracks in the studio, have you tested songs from Rush and Neil Peart’s drumming?

    • #2 by Geoffrey Heinzel on February 18, 2014 - 6:37 am

      Paul, you seem to have a long experience in music recommendation. All streaming services offer a huge catalogue of music with millions of songs, but when I try to listen to some new music that sounds similar to what I know and like, not one offers the user a way of interacting, for example to enter a seed song and generate an auto playlist based on the acoustic fingerprint and not on genre or similar artist. Maybe I expect too much, or am not aware of who offers this, or how to do it, but a real world app or service that works without being a hacker is missing. Thanks for helping, and sorry if I posted this in a wrong blog or format.

  2. #3 by ryangrus on March 17, 2009 - 12:30 pm

    Hi Paul,

    I also stumbled on your blog after reading about the click track post. I really enjoy everything you’ve written so far.

    I have a blog where I post a new drum loop everyday and a question has arisen in an area where you seem to have some expertise in:


    Basically, are there any current methods for converting an audio track into a transcribed piece of music. Much like audio to text conversion, but obviously much more complex. Would love to hear your thoughts!


    • #4 by plamere on March 17, 2009 - 11:16 pm


      It is not a solved problem, but it is a very active area of research. Google for “ISMIR transcription” to see the state of the art.


  3. #5 by matt on April 11, 2009 - 12:46 pm

    i stumbled across this blog from your sun blog post on the api for yes.com. i was annoyed that the website didn’t have an rss feed for song logs or other such easily exportable format. (i had to go select a station, date, hour, click submit, change hour, submit…) then i saw the link to come over to this blog and yow. this is such interesting stuff. keep up the interesting posts!

  4. #6 by Troy on April 16, 2009 - 7:21 am

    Hi, re Spotify – My other half is worried about using it because it’s peer to peer, but does that mean people are playing music from my pc ? like – sharing, in the way we usually do with peer to peer.
    I thought the music files are all at a Spotify server.
    And no chance of viruses ?

    cheers !

    • #7 by plamere on April 16, 2009 - 7:24 am

      @troy – a bit off topic? but no, spotify does not serve music from your PC to other people.

  5. #8 by dbEsq on April 22, 2009 - 11:54 am

    Ryan and plamere,

    You may want to check out Melodyne, which recognizes audio as notes and can treat it like MIDI. I don’t know if it can render that info into a transcription or not. Hope this helps.

  6. #9 by Used machinery on May 12, 2009 - 4:52 am

    Introduction to the quietly proliferating problem of digital piracy in the literary world. “I thought, who do these people think they are?” Ms. Le Guin said. “Why do they think they can violate my copyright and get away with it?”


  7. #10 by john henning on July 16, 2009 - 9:10 am

    Seen http://www.sun.com today? This article was linked to front page. I saved a copy of the graphic, if you like. http://www.sun.com/featured-articles/2009-0710/feature/index.jsp

    • #11 by plamere on July 16, 2009 - 10:30 am

      Thanks John.

  8. #12 by J L van Os on October 19, 2009 - 4:50 pm

    Dear Plamere,
    Apart from subjects on this page I have seen your article: Searching for music by melody or rhythm, with the statement: (about Musipedia) The best query-by-melody, and query-by-tapping site I’ve seen.
    This was in 2006
    You might not yet have seen the Melodycatcher.
    We think this tool is clearly better (more accurate)
    but undoubtedly you can better judge that for your self.
    All questions, comments and suggestions are welcome.
    “We” is Florian Bomer and my self, as you can see on the website under “About”

    Best Regards
    Jan van Os

  9. #13 by Chris Castiglione on November 4, 2009 - 12:26 pm

    Paul…. great blog. I was also at Music & Bits, and so I think it’s great you recapped Brian’s talk.

    I was wondering if you’ve heard of a really simple online tap-metronome (preferably using Echo Nest) where I can tap in a beat for 20 seconds..and return the average. Do you think that is how EN would work – it’d be the avg? Or would it maybe just choose the first 5 seconds and decide the BPM?


  10. #14 by Seth on November 7, 2009 - 7:39 am

    Hi Paul – What happened to Snapp Radio? I loved that site? Is there anything similar you know of??


  11. #15 by Daniel Webb on January 22, 2010 - 6:21 am


    I’ve read your music recommendation app reviews over the years with interest, and mostly agree with your findings that most disappoint.

    I’ve just heard about the youtube music recommender (http://www.youtube.com/disco) – I’d be really interested to hear what you think of it.

    Can Google can bring the kind of crowd analysis to music discovery that they bring to translation and search (http://googleblog.blogspot.com/2010/01/helping-computers-understand-language.html)?

    (Sorry about abusing the about page, not sure how else to contact you)

  12. #16 by Rob on February 11, 2010 - 10:02 am

    Are you going to be doing the SXSW Artist Catalog this year? It was really helpful. If not, do you still have the code? Somebody should keep it going.

    • #17 by Paul on February 11, 2010 - 11:29 am

      Yeah – look for it around March 1

  13. #18 by Webby on May 26, 2010 - 6:56 pm

    Your music engine might need updating…
    I did link between Itaal Shur and Carlos Santana — it should have been one degree of separation, they cowrote “Smooth” and won a Grammy for it.
    Your program had it at 10 Degrees…

  14. #19 by Estrella Espinoza on July 10, 2011 - 10:20 pm

    hola, Hi, in Bogotá Colombia, I´m so happy wen a discover your musicmaze, the first vesion was simple but is a god idea… I hope that the new vesion still god and better….. congratulations

  15. #20 by maxim on August 20, 2011 - 8:59 am

    In case you’re interested, I’ve created a free smart playlist based system for iTunes that will learn song ratings over time as well as vary the play frequency based on those ratings. Just the thing for organizing huge libraries. Only had a couple people download it, but they seem to like it so far.


  16. #21 by Rory on October 6, 2011 - 1:54 pm

    Just tried out ‘ 6 degrees of black sabbath’ ; Todd Rundgren to Ian Tyson, curious to see where it took me. it charted a path that went through Rambling jack Eliot to Corb Lund etc.I believe that back in the early seventies Todd Rundgren produced the Ian and Sylvia Tyson album’ The Great Speckled Bird’ I may be wrong but I thought you might like to know. Cheers.

  17. #22 by Grace Asiado World of Music on July 1, 2012 - 3:58 am

    Yes, it’s really hard to choose what music you will listen to if there are millions of tracks to choose from. For me, I often watch TV shows on music channels.

  18. #23 by SoundEagle on November 26, 2012 - 10:28 am

    Hi Paul, thank you for all the information and research. Keep up with the writing!

    It is a pity that Infinite Jukebox does not work in some browsers. By the way, is it possible to save a copy of the generated output of Infinite Jukebox? In other words, can Infinite Jukebox save each new result as a downloadable mp3 file?

    • #24 by Paul on November 26, 2012 - 11:29 am

      no – sorry it is not possible to save the output. An infinitely long song would be too large to fit on all of the world’s hard disks

      • #25 by SoundEagle on November 26, 2012 - 3:31 pm

        That goes without saying! It would be really nice if Infinite Jukebox could have a “Save” feature to allow the user to save as much or as little of the output data on his or her storage device(s).

  19. #26 by Jesús Herrera on March 8, 2013 - 8:22 pm

    Paul, Hi.

    My name is Jesus Herrera and I am just an amateur audiophlie.

    I also have a (very humble) Blog at WordPress, which mainly focuses in reviewing songs, trying to analyze why I like them.


    Your post “The Looudness War Analyzed” was a true eye (ear) opener for me, thanks.

    Would you allow me to put a link on my Blog to yours? I think it might be valuable for my readers.

    Thanks. Greetings from Mexico

    • #27 by Paul on March 9, 2013 - 9:09 am

      Of course. — Paul

  20. #28 by Chris on May 7, 2013 - 8:46 pm

    There are so many clever tools on here. One I particularly like is the Map of Music Styles. How hard would it be to adapt this to highlight your own music library based on Spotify for example? And then create a playlist using the Boil the Frog algorithm by selecting one or more nodes?

  21. #29 by Rick Bolton on August 12, 2013 - 1:06 pm

    Hi, Paul… I’ve admired your work from a distance, and would love to initiate a conversation with you regarding discovery paradigms. I am working on film and television content discovery, which is totally different from music discovery, of course, for many obvious reasons, but I think we still have some things to talk about! Please drop me a note at my email address (provided to you via my login here)… hope we can connect soon.

  22. #30 by Dominic on November 20, 2013 - 7:21 am

    Hi Paul,

    Nice blog – I particularly enjoyed your post about the Hackathon in Finland.

    I write for Xyo’s mobile apps and games blog (http://xyo.net/blog/) and I’m writing an article about the best apps for making, recording, performing, and listening to/discovering music, and, as you’re a music lover, app hacker and someone who is interested in the tech-based side of music I thought I’d write to see if you’d be willing to share some of your wisdom.

    Please let me know if you’d be up for a short interview.

    Hope to hear from you soon,


  23. #31 by aondoe on February 16, 2014 - 2:12 am

    Very informative and interesting blog you have here. Thanks for your efforts, will follow.

  24. #32 by indie music on May 20, 2014 - 3:47 pm

    When I initially commented I clicked the “Notify me when new comments are added” checkbox and now each time a comment is added I
    get several e-mails with the same comment.
    Is there any way you can remove people from that service?


    • #33 by Paul on May 21, 2014 - 12:45 pm

      You have to manage this yourself (It is a wordpress.com setting that I have no control over). Go to your settings page and you’ll be able to disable notifications.

  25. #34 by David Johnson on September 17, 2014 - 11:00 am

    I am interested in your study of profanity in music. Could you tell me where you got your data from, and can you quantify the relative increase in profanity in music from the 70s to the present in a percentage increase?

    • #35 by Paul on September 18, 2014 - 5:05 am

      Data is from musicbrainz. The data show that the likelihood of a song with the word ‘fuck’ in the title has more than doubled since the 80s.

      • #36 by David Johnson on September 23, 2014 - 11:58 am


  26. #37 by Andrew on October 17, 2014 - 7:41 pm

    Hi Paul. I’m currently a college student majoring in Statistics and am doing a project on predicting people’s preferred music tastes given their characteristics. I’ve been inspired to do this project from reading your blog. Does The Echo Nest release datasets for academic purposes, and if not, where can I find relevant datasets to use? Thanks.

  27. #38 by lakyab on January 18, 2015 - 12:46 am

    Hi Paul. I just came across your blog. Things look great. I particularly appreciated your posts about the music data analysis, and the APIs.

    I also started a similar blog here: http://www.musichemicals.com . I am using the Last.fm API to do some data visualization, and we started writing some posts analyzing music data. I appreciate your feedbacks and comments.


  28. #39 by johnmac08 on March 7, 2015 - 3:15 pm

    Hi Paul – I just wrote a short review about your Where’s the Drama app. http://www.indoorcycleinstructor.com/icipro-instructor-training/music/spotify-app-shows-where-you-yell-go/

    This would be golden to us Indoor Cycling Instructors > especially if it would display a compete playlist in real time. It’s very effective for us to cue changes in movements/intensity at points of high drama in the track. It’s tough to remember where each is. love to discuss

  29. #40 by Louie Cesario on July 2, 2015 - 1:14 pm

    Hey Paul,

    Big fan of your blog- I was hoping that you could take the time to check out http://www.rocketsongs.com , We are a music discovery and distribution platform that allows users to license 100% original songs to purchase, record over and distribute to their liking. It is a one of a kind service, we haven’t been around for long- but we are growing fast and receiving a ton of excellent feedback from our users. I would love the chance to connect and chat about the opportunity iof being featured on your blog at some point. Feel free to reach out with any questions/comments. Keep up the awesome work!

    Louie Cesario

  30. #41 by Julia on July 16, 2015 - 4:11 am

    Hello, sorry for my english I’m french. I just want to say that your projects are amazing but we are not able to find the infinite jukebox page anymore. This is sad !

  31. #43 by rayofsanjose on September 16, 2015 - 12:18 pm

    Hi Paul,

    Just found your site and find it quite interesting. I look forward to perusing the contents.

    I was looking for an explanation on the web as to why music library management apps don’t allow sorting music by “Gender.” I currently use Foobar2000 and have lots of experience with iTunes (which I do not like for many reasons.)

    They allow you to cut data all manner of ways and yet one of the most basic, male or female performer, is not available, at least from what I can tell.

    Am I correct or have I missed something? I would like to sort music by say, “Female Country Singer” but trying to pick my listening music is made more difficult for lack of this “basic” tag.

    Any comments or suggestions you can provide? Thanks in advance.

  32. #44 by makingitrealhr on October 24, 2015 - 12:55 am

    Infinite jukebox doesn’t seem to work anymore. Everytime I visit the page it says “The service is not available. Please try again later.”

  33. #45 by keereell on March 3, 2016 - 11:11 am

    Hello, Paul. Thank you for blog and your web service. I work as PD on the radio station in Russia, so I found your posts very useful. Can you please help me with more detail information? As a programm director, I need some specific information about our tracks. I tried work with some APIs (the most I used is Last.FM API), but I didn’t found all information I need. Trying all of APIs can get a lot of time, so maybe you can tell me where I can find that info I need? If ‘yes’, how can I send you more detailed information?

  1. The Echo Nest Remix API | Superposition Kitty
  2. Ismir 2009 Music Industry Panel : Self Conscious White Noise
  3. Datachondria » Blog Archive » Discovered: Music Machinery
  4. Ismir 2009 Music Industry Panel « self conscious white noise
  5. Geeks Can Be Rock Stars Too « eleetmusic – Direction in music, marketing and business
  6. The Infinite Jukebox « NYC Expeditionist
  7. Recommended: 5 Great Talks on Recommender Systems | spider's space
  8. WordPressers Making a Splash: March Edition — Blog — WordPress.com
  9. Audio Cast: Determining Music Preference | kVibe
  10. Mapa de Estilos Musicais - Musicbit

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

%d bloggers like this: