Music Ally on Music AR

Music Ally replied thusly (paywall) to my post yesterday on musical versions of Pokemon Go:

 Pokemon Go’s success is about Pokemon more than AR: it’s hard to think of music brands that would generate a similar buzz, rather than (for example) story-based brands like Harry Potter or Star Wars.

My experience with my kids yesterday confirms that Pokemon is critical to Go.  They loved the game so much I almost couldn’t get them to glaze over in front of the TV instead of hunting Pokemon around the neighborhood. A big part of the appeal was imagining specific Pokemon characters existing in the real world. If they didn’t love the characters, they wouldn’t care.

That said, I think the product challenge for augmented reality music is to find a good reason for that particular sound to be in that particular place, and to be virtual rather than physical.

Annotating the space is a use case that satisfies both requirements.

As a practical, rather than artistic, example, people in a restaurant might review menu items.

The sound needs to be in that particular place because it is about that place; using the location to access the sound makes the app easier to use. The location establishes the context, which frees the user from having to tell the app what the context is.

One reason that posts need to be audio because talking is easier than typing . Another is discoverability: the user doesn’t have to be looking down at a screen to discover that relevant audio exists; the sound can just start as soon as it becomes contextually relevant.

Why does the sound needs to be virtual, as opposed to running over physical speakers? Eliminating the need for distinct speakers in every location creates economies of scale. Some types of audio – such as comments about menu items in a restaurant – are only relevant to one listener at time.  The owner of the speakers would only want flattering comments.


All this said, Music Ally’s thought about buzz suggests a marketing issue. Music is for art. Whether it works is a creative question. So what we need to figure out is the creative and artistic value of location-based music.

Dancing or going to a show is obviously a social activity, and location abets social gathering. So then we want to ask about the benefit of virtuality. Why is virtuality crucial?

Noise is one possibility. To the neighbors a virtual concert is indistinguishable from a silent one.

But I don’t know. This is all speculation. You’d have to do it to find out.

AR Music Layer

A Music X Tech X Future post on Pokemon Go speculates on augmented reality and music. In practice, what would that be like? I’ll try to visualize it.

There is music – well, audio – associated with a location. When you enter the location the music starts. When you leave it stops.

This is different than a speaker making physical noise in a couple ways. One, it is opt-in. Silence is an option, listening to your private earbud stream is an option, joining the public AR stream is an option. Two, you don’t need physical infrastructure. You can do this in places with no electricity and you can do it without anybody’s permission.

It can very specific to location, down to the finest physical resolution of the device. The magazine section of a drugstore might have the sound of celebrity voices, while the food section had hip hop. Or it might be very broad.

The soundtrack could be programmed by anybody, like a shared bulletin board. You might be standing in a dark alley and think of pinning up creepy footstep sounds. Or you might go into the park on a sunny day and decide to pin up “Walking on Sunshine.” More than one person might pin up songs in the same location, creating a conversation. People could leave spoken notes.

Music could be planted in very public spaces, or in very private ones.

As you drive down the street you might get a feeling of the places you are passing through from the associated music.

Some people would be better at it. You might go hunting for songs pinned by a specific person.


This feels fun and credible to me. If I had the tool I’d enjoy taking it for a spin.


Postscript —

Location Aware Albums as Apps

Washington DC-based duo Bluebrain’s latest release is not a traditional album — it can’t be listened to passively in one sitting or, for that matter, at just any location. ‘Central Park’ is a site-specific work of music that responds to the listeners location within the stretch of green of the same name in New York City. Available only as an iPhone and iPad app, the album will be released starting October 4th, 2011.

‘Central Park’ is the second in a series of site-specific app-albums, following ‘The National Mall’, released last May, designed to work within the boundaries of the park in Washington DC.

Both albums work by tracking a users location via the iPhones built-in GPS capabilities. Hundreds of zones within the landscape are tagged and alter the sound based on where the listener is located in proximity to them. Zones overlap and interact in dynamic ways that, while far from random, will yield a unique experience with each listen. The proprietary design that is the engine behind the app stays hidden from view as the melodies, rhythms, instrumentation and pace of the music vary based on the listeners’ chosen path. Unlike other music-related apps, these are not musical toys or instruments. Nor are they a compliment to a traditional album release. The app is the work itself. A musical ‘Chose-Your-Own-Adventure’ that does not progress in a linear fashion but rather allows the listener to explore the terrain and experience music in way that has never been possible before now. Chris Richards wrote in a Washington Post cover story, “Bluebrain (‘The National Mall’) is helping to redefine what an album actually can be…it’s truly magical”

Hacker’s Guide to DMCA Takedown Requests

Are you a developer who needs to handle DMCA takedown requests? It’s not trivial but it’s doable. Most of what you need to know is programmer-friendly. The law is basically a flowchart.

So I created this flowchart for developers to use. It’s big – you’ll need to click through, open it, and expand to see the details.

Reference for startups - what to do about DMCA takedown requests
Reference for startups – what to do about DMCA takedown requests

Warnings: I am not a lawyer, I am a technologist. This is not legal advice. This is my personal attempt, as a technologist, to collate available information and share what I have learned. I have taken care with this, but I am just some dude on the Internet. I’m very happy to make corrections if you see an issue – I already found and fixed one major error.

I’m happy to help explain bits of this or otherwise help out.

Earbud implant

Man implants magnets to make his ears into their own headphones

If headphones are too bulky and ear buds make your ear canals hurt, why not surgically transform your ear itself into a speaker? That’s what body hacker Rich Lee has done, by implanting rare-earth magnets in his ears, so he can listen to music or amplified sounds even when he’s not wearing headphones.

In addition to music, he looks forward to connecting these embedded bio-speakers to a directional microphone or a voice analysis app, so he can do surreptitious spy-like activities, like listening to conversations across the room and detecting whether you’re telling lies or not. He’d also like to connect his setup to a Geiger counter, so he can get ambient readings on radioactivity, or perhaps use it as part of a digital echolocation system of some kind.

He first implanted magnets into each ear’s tragus (the flap of skin and cartilage that sticks out in front of your ear). No licensed surgeon would be willing to do this procedure, so like many transhumanists, he had to get a friend to do the operation for him.

To make these magnets into a speaker, he wears an induction coil around his neck, connected to an amplifier and, through that, to his phone. The varying currents in the coil cause the magnets to vibrate, exactly as they do in a speaker — except in this case, the magnets are part of his actual flesh.

That was building on the Instructibles project Make Your Own Invisible Earphones

Open projects in open music

– Cross reference artists, albums, and tracks in Discogs, Musicbrainz and Wikipedia, so that an entry in each one is linked to the corresponding entry in the others.

– Finish the genre graph implied in DBPedia. Extract the graph and QA it as a whole. Enable multiple distinct genre graphs, so that multiple cultures are supported. Create graphs for non-EU/US cultures, e.g. for use in China.

– Cross reference Encyclopaedia Metallum with Discogs, Musicbrainz and Wikipedia.

How to ask for a raise

Here is how to ask for a raise:

Ask for a raise.

Just asking is the most important thing. A lot of people don’t. Not everybody who asks gets one, but almost nobody who doesn’t ask gets one.


If you want to get fancy, have good reasons. What are good reasons? Go make a friend in HR and ask. Here’s a starting point:

  • Underpaid relative to other companies who might hire you
  • Long time since last raise
  • Outstanding evaluation
  • Credible threat to leave
  • Been at the company for a while
  • Increase in responsibilities without increase in pay. Implicit promotion.

Also: a promotion is the single best way to get a raise. Go ask for one. Make your desire to move up known.

Open Genres

There have been many open-ish projects working with music genres, but nothing on the scale of Musicbrainz or Discogs as far as I know.

  • Maintained on an ongoing basis
  • Global reach
  • Translated
  • Professional-level musicology

This territory is held by proprietary interests, including data vendors like All Music Guide (Rovi) and Gracenote, as well as music service providers like Spotify and Apple Music.

Maybe that shouldn’t surprise me. The great care and patient labor to create such a data set is generally not a good match for crowdsourcing. Crowds can do amazing things, but they are not methodical.


The Baroque entry in DBPedia is an interesting parallel.

Maybe I have bias from my time in the commercial music data world, and DBPedia is highly competitive. The entries for Jump Blues and Rockabilly, and their relationships, are excellent.


DBPedia has a single genre map for the whole world, I believe. (It might be scoped within a language, so that genre relationships don’t span more than one language).

This is a problem because genres mean different things in different cultures. Highlife is World Music in the US but Palm-wine music at home.

There could be a graph (“genre map”) which is a superset of all graphs, but it wouldn’t be coherent. There would be multiple roots or starting points, the same genre would appear in multiple locations, the same genre name might be used in different ways.

In terms of data modeling, there needs to be one simple change in DBPedia: add an identifier for the cultural context to the the music genre class.