Monthly Archives: March 2012

open source hardware for dance music made in China by westerners

Jon Phillips and Wolfgang Spraul are doing hardware hacking on dance music devices. They’re both westerners living in China.

One project is Milkymist:

It’s easy to create an entertaining video installation with the Milkymist One. No computer needed – everything is included in a small device that has it all. Connect a camera and a video projector, press the power button, and seconds later, everything you film becomes live psychedelic effects of color and light.

Another project is the Laoban Soundsystem:

a 6,000-watt massive soundsystem fabricated in China, designed by Matt Hope and produced by Jon Phillips.

The whole thing is so far ahead of me that I don’t know where to begin.

To begin with, a good rule of thumb for hardware development is to not do it. There be dragons. Like, you’ll have to manufacture the things you design, which means dealing with shops in the third world, which means being very far away from people who you are totally dependent on. Also there’s the issue that the factories are staffed by people who are brutally exploited. You’ll have inventory. Your latency will go through the roof.

Hang on, Jon and Wolfgang moved to China. So it’s ok, the factory is within reach, they can speak the language, etc.

Oh, wait, it’s ok, it’s open source. Whaa? Doesn’t it matter that you’re using atoms instead of bits? At the least there’s a whole new set of dynamics that apply to open source atoms. It affects inventory – you’re not actually the person doing the manufacturing, so you’re not paying for the stuff you made to sit in a warehouse. Whacky.

Our devices are open source hardware and software. In fact, we go great lengths to apply the open source principles at every level possible, and is best known for the Milkymist system-on-chip (SoC) which is among the first commercialized system-on-chip designs with free HDL source code. As a result, several Milkymist technologies have even been reused in applications unrelated to video synthesis.

Not that I have any idea what to do with this stuff except respect that it’s a whole new frontier. It’s the kind of over-my-head innovation that makes me love my chosen field. Inspiring stuff.

Muve is the model

From the marketing brochure page at

Muve Music from Cricket is a game changer for everyone. By tightly integrating the music service into the handset and the billing plan everyone in the value chain benefits and consumers have a complete music service where the phone is the hub not the PC.

And Muve is doing very well as a business. My memory is that the subscriber numbers are in the hundreds of thousands, which puts it well ahead of Rdio or MOG.


You may know this already, but Beats is basically the music recording industry moving into the music hardware business. It’s a similar strategy to 360 deals: abandon ship. Key people at Universal Music Group saw that the money was in iPods, not MP3 files.

The label people see the same writing on the wall as everybody else.

Centralization and MOG acquisition

Wolfgang Spraul posted a long and interesting comment which is worth reading in its entirety. The gist of it is that increasing centralization of the music industry is bad for musicians and listeners.

I agree that less choice of vendors is a bad thing. If Beats is aligned with a single distributor – MOG – then they will make life harder on customers of Rdio, Rhapsody, Spotify, etc.

But it’s really important to create a healthy foundation for the recording industry. The evolution of clusters of related businesses like { HTC + Beats + Universal Music Group + MOG } is about the ecosystem reconfiguring to match changes in the environment.

This is the change we are seeking. The business of recorded music needs to leave physical media behind, and that’s precisely why { HTC + Beats + Universal Music Group + MOG } exists.

That doesn’t mean this is the final form of the business. A device has to play all the content in the world, so companies like Beats and HTC can’t lock out distributors aside from MOG. They have to allow Rhapsody et al into their world.

How will Beats go to market with MOG? My guess is that the MOG brand will be abandoned in favor of the Beats brand. MOG’s software, like the in-browser app and mobile apps, will continue to exist, but will be named “Beats.”

When you get an Android phone with “Beats”, that will mean it comes with a subscription to MOG. You’ll be able to add a subscription to Spotify or whatever other service interests you, but you’ll have already paid for the MOG subscription as part of the purchase price of your device. That means phone prices going up to cover the subscription. Where will the money come from? The subsidy paid by your telecom. The cost is about $10 a month. I suppose that will be added right on top of your phone bill.

magically transferable label deals

A subtle thing about the MOG/Beats/HTC deal just occurred to me.

Companies like MOG, Spotify, and Rdio rely on negotiated deals with record labels. Those deals usually contain clauses that make the deal non-transferable in case of a change of ownership. When MOG changes hands to ownership by Beats, it has to start over again with the labels.

It happens that one of the major stakeholders in Beats is Jimmy Iovine, who is also a lead at Universal Music. That’s not to say that in this deal UMG is getting ownership of MOG (they already have equity anyway). It is to say that getting Universal’s blessing is pretty much a done deal. And where Universal goes, Sony and Warner go, because Universal is the 600 pound gorilla of labels.

The deals not being transferable depresses the value of a company like MOG, because the deals are a key asset. That’s why Myspace Music went for so little. But Beats doesn’t have the same problem with buying MOG that other suitors did. For Beats the deals actually *were* transferable.

the brutish and short life of a commodified complement

On-demand subscription services don’t earn a profit for their owners. All the money goes to other parts of the ecosystem. The subscriber fees are passed to the labels. Electronics used for listening make a profit for the manufacturers.

Think of iPods. Most of the money for buying tracks goes to the labels. Apple keeps the money from selling iPods. There is no independent “music store.”

So how to interpret the news that Beats, which is owned by HTC, is buying MOG? Beats and HTC are device companies. They make good money by selling equipment on which to listen to music. In addition, Beats is affiliated with Universal Music Group, the biggest label.

MOG makes little money. In that sense it’s no different than other online music distributors, including Pandora and Spotify. However such companies do make money for device makers and labels. The content, equipment, and distributors are complementary goods. HTC, and sorta kinda UMG, are buying a complement in order to enhance the value of their primary line of business.

This move is ultimately good for the internet music industry, because it puts the industry on healthier footing. MOG’s product makes more sense as part of a larger service which makes a net profit than it does as a standalone which breaks even at best. The underlying economics are getting better.

Information design and “I’m Gonna Start a Graveyard of My Own”

When you have a song cheat sheet with chords and words (and maybe notes) on it, how come it’s always black and white? Well, because color printing is more expensive, obviously, but also because the value of color is underestimated.

I wrote up a cheat sheet for a song the band was going to learn live, performing it the very first time they played it. So the cheat sheet needed to be excellent, and therefore the cost of color ink was worth it.

Written music is an information graphic like a subway map or graph of annual GDP. There’s visual design being used to communicate quantitative information.

When you think about it that way the value of color isn’t controversial. The GDP of Romania is a blue line, the GDP of Poland is a red line, the color contrast helps you compare and contrast the data.

So here’s my chart. The chords are the most important thing when a band is faking it, so I used color to distinguish the chords from the lyrics. In addition I communicated the arrangement by putting labels for sections on the right margin in all caps.

I'm Gonna Start a Graveyard of My Own lead sheet (click for full size PDF – good for printing out to jam on).

I accommodated transposing instruments like trumpet (needs to see a “d” note to play a “c” note) by naming chords according to their relative position within the key, and went with the convention of using roman numerals for showing those chord names. In the key of G a “g” chord is “I”, a “c” chord is “IV”, a “d” chord is “V” or maybe “V7”.

This would be better if I had actually notated the pitches but I didn’t have enough time.