The Echotron

Having read Paul's post yesterday, I instantly applied for a beta invite. In fact, while writing this blog post I'm listening to a very pleasant stream of music from the echotron. I'm impressed! It couldn't be much easier to get started. It's very easy to quickly add a bunch of artists I like to my profile, and I even get to listen to any search results. The recommendations are pretty good, definitely a lot better than many others I've seen out there... and they are sometimes rather different from the ones I get at Last.fm (in a good way). However, despite being biased, I'd still argue that Last.fm's recommendations are better ;-)

Anyway, it seems the echotron could easily turn into more than just a site built to showcase the echo nest's APIs.

Jobs@Last.fm

If I'd know how to code C++ I'd instantly apply for this job:

Software Engineer, Data and Recommendations

We’re looking for someone who can help us improve our recommendation engine, someone who knows how to handle large amounts of data, and knows C++ really well. You’d be joining a small, enthusiastic, and flexible team.

Crunching terrabytes, dealing with computational complexities on a daily basis, developing applications where you can hear the difference (and can measure the feedback from millions of users), working on some of the most elegant C++ code every written, designing and building the next generation of Last.fm's recommendation engine, ...

Automatic accompaniment

Here's a great video demonstrating how much fun automatic accompaniment systems can be:



The video demonstrates Andrew Robertson's (C4DM) live drumming controlled automatic accompaniment system: "James Sedwards (guitar) and Jeremy Doulton (drums) improvise a rock track by recording loops of bass and guitar. This is done in Ableton Live with B-Keeper controlling the tempo so they stay in time even when speeding up."

Here's a link to one of the publications describing the system (PDF).

Applying audio-based similarity

Applications of audio-based similarity still seem rather rare nowadays. Once in a while a startup claims that they're using it for recommendations, but looking at their results might suggest that they are just using metadata instead (for example, see this example which Paul recently blogged about).

Anyway, here is an example where 100% pure audio-similarity is being used: FM4 soundpark. The feature launched today. For anyone who doesn't read German: soundpark is the number one place for new Austrian artists to expose themselves online. Soundpark has been around for ages. Long before Myspace or similar sites became popular. Soundpark has a devoted community and is well integrated into one of the most popular radio stations in Austria (the FM4 station which reaches out to younger demographics favoring alternative/indie music).

However, attention is usually only paid to new releases and old ones can get buried quickly. To help users navigate and find content researchers from the legendary OFAI (including Martin Gasser and Arthur Flexer) in collaboration with Gerhard Widmer's department of computational perception (including Dominik Schnitzer) have helped FM4 integrate audio-similarity into their system. There is a bit more information on OFAI's page here and some more information in German can be found on orf.at.

To each song on Soundpark users will get 3 acoustically similar songs, and there seems to be a feature that allows creating a playlist by defining a starting and end song (but I haven't found that feature yet).

Soundpark hosts about 5000 artists with each a few tracks in average. (If I'm not mistaken they got about 8800 tracks.) They are constantly growing. According to the press releases they have a number of plans to add additional features to make it easier to navigate their content and discover interesting artists. Great news! Makes me especially happy knowing that it's former colleagues who are building this.

Tag of the Day: Death Country

Today I learned from Mischa about death country. Apparently it's related to gothic americana... which btw sounds nothing like gothic.

The Echo Nest on Techcrunch

The Echno Nest made it on Techcrunch!

Interestingly Techcrunch seems to think that The Echo Nest is about genre classification (second sentence in their blog post). I doubt they heard that from Brian.

One reader raised an interesting question: "[...] do we really need a computerized pandora?" It's nice to see something like this discussed outside of the music-ir mailing list!

I really like The Echo Nest's very open approach to demonstrate the technologies they got. Besides being of interest to anyone in MIR I'm sure it also makes it a lot easier to attract interest to their APIs and communicate what kind of problems can be solved with them. But I'm not sure if I like the redesign of their webpage, I somehow had gotten used to the simplicity of their original one. I'm very curious what their next demonstrations will be... maybe some web crawling based stuff?

See also Paul's post about this is my jam.

Thoughts on Autotagging

I've recently spent a lot of time looking at tags. I'm fascinated by how creative listeners are when it comes to inventing new tags for the music they care about. And I'm fascinated by how small communities adapt these newly invented terms as quickly as they tune into a new radio station. This constantly changing and seemingly endless vocabulary of words to describe music makes me wonder if research on automatically classifying or tagging music will ever be able to catch up.

If you think research is catching up, check out the 22,703 shades of metal, or the 11,929 shades of pop, or the 29,940 shades of rock at Last.fm.



See also Paul's experiments on creating a taxonomy of metal and gothic tags.

See also Mike's recent results on autotagging of music (which includes some impressive numbers).