My complete catalogue of Gold Noise is now available on goldnoise.org.

Gold Noise is the sound of one million pop songs playing all at once.

I’ve written a Gold Noise generator for Mac OS X. Download it to generate your own, personal Gold Noise: goldnoisegenerator.zip

Here is a screencast of Gold Noise being generated on my laptop,

Source code available here.

Bonus: here is a score for Gold Noise that can be performed by any Mac (OS X 10.5 or above).

Gold Noise is the sound of one million pop songs all playing at once.

Here are some samples of Gold Noise. They’ve been generated by a Ruby script that randomly chooses mp3s from my music folder and blends them together.

10 Songs

20 Songs

40 Songs

100 Songs

Gold Noise builds on a tradition of naming different types of audio noise after colors. Noise in audio is often defined as a random signal — that is, the antithesis of signal. However the source of the randomness that generates the noise gives the noise different “colors”. For example, white noise can be generated by a series of random, uncorrelated samples (essentially a series of random numbers); it sounds like TV static. Brown noise is generated by a signal on a random walk, that is, a signal that is changed by a random offset at each sample; it sounds like a waterfall.

Gold Noise is generated by a random sampling from the cultural landscape of Pop Music. Pop Music is the collection of sounds that are shaped by the cultural zeitgeist as well as commercial forces to be, in some sense, the most desirable sounds to human ears at a given time in history. The sound of Gold Noise subconsciously explores these forces that shape Pop Music.

Within the cultural oeuvre, Gold Noise builds on the institutional critique of Pop Music by groups such as The KLF. In 1988 — after releasing the novelty pop song “Doctorin’ the Tardis” which reached number 1 in the UK Singles Chart — The KLF wrote a manual on how to “write” a hit pop song. The manual lists Golden Rules, mostly concerning how to lift elements from other hit songs to create a new one. Pop Music as Gold Noise itself.

Further back, Gold Noise builds on the work of the Dadaists and later the Situationists. These groups, through physical and conceptual collage, explored the collective subconscious of human societies. As in Gold Noise, random sampling was a key strategy for this exploration.

For my Sound and The City final project, I will incarnate Gold Noise into several physical and virtual spaces.

Gold Noise Radio

A physical, self-contained radio consisting of speakers and some number of dials which can be separately tuned to FM radio frequencies. When turned on, the radio plays all of the tuned-into radio stations at once. The radio’s enclosure will be painted gold.

Gold Noise Transmission

A radio station in New York City will be enlisted to broadcast 3 minutes and 20 seconds of live Gold Noise. 3 minutes and 20 seconds is the length of a perfect pop song according to The KLF manual. The broadcast will consist of all of the other radio stations in New York City playing at once, live.

Gold Cities

Using the internet streams of radio stations, radio Gold Noise will be sampled from several large cities all over the world. These samples will be released for comparison on goldnoise.org.

Personal Gold Noise

The above Ruby script will be packaged into a downloadable application which will create personal Gold Noise from the mp3s in a user’s music folder.

Sundialtone - Drunk Dial 2 (Half Price Special).mp3

Sundialtone - Drunk Dial 2 (Instrumental).mp3

Sundialtone - Swimming in the Ocean.mp3

Sundialtone - Non-Stop Metro.mp3

Momus is one of my favorite songwriters. (He’s also a favorite of Belle & Sebastian, Of Montreal, and Vampire Weekend). I got into Momus through his connections with Cornelius and Kahimi Karie. He had a couple top tens in Japan with the songs he wrote for K.

Last night I saw Momus, accompanied by ipod, chronologically karaoke through a career-spanning retrospective. I knew about half the songs. All of the backing arrangements were completely fresh to me though.

Throughout the performance, Aki Sasamoto did interpretive dance dressed as a Kubuki stagehand. Aki is an amazing, intuitive dancer! I especially liked when she started taking all the spare mic stands in back and ceremoniously adjusting and arranging them on stage. She and Momus have a great dynamic together, playing with the fact that you know they’re making it up as they go long. They are fresh out of a long-running improvisational performance piece in New York.

I think what makes Momus special is that he can sustain an idea (usually a language-formal idea or literary-cultural comment) for several verses while consistently keeping the language pushed to its limit, never wasting a word. His melodies and arrangements always support the lyrics absolutely (he’s strongly words first, music second). And his choruses never disappoint (a problem with many other talented songwriters).

It always amazes me to see an artist present a singular concept while pulling all the stops constantly, effortlessly. Roman Polanski does this for me too. His movies are full of cinematic memes, like the creepy guy hiding in a private space glimpsed in a mirror and then he’s not there, the voyeuristic neighbor in the background who watches the main action until he’s spotted, or the camera following a character through a busy street with jazz in the soundtrack (all examples from Repulsion). But these effects are always done in service of the story, never for their own sake.

All too often in my own creative endeavors, I think of an effect and build a piece around that, whereas I should start with a concept and use effects to enable the concept to be born. I suppose the key is to internalize a repertoire through practice and experience, to make it intuitive.

Open source framework for selling music online

The (my?) problem with the online music marketplace is that almost all transactions go through a few large players (iTunes or Amazon, for example).

This state of affairs–the premise that large, centralized resources were needed to distribute music–is a legacy of the pre-Internet era.

I propose an open source project that makes it super simple to set up a self-hosted online music store.

There’s no reason this needs to be hard. Musician loads the software onto her website, opens up the admin UI. Uploads mp3s, graphics, etc. Sets prices. Save Changes and presto music store.

Now when her fans go to her website, they buy music directly from her. None of the money gets sucked into the vortex.

Well, there are still a few costs but there’s reason to believe these are trending toward minimal:

- She has to pay for bandwidth and web hosting.

- Out of convenience, she opts to use a third-party credit card processor. There is, after all, a plugin framework and a community of developers who make add-ons that let your music store integrate with other services.

Oh, and our musician is multi-media-talented, she also sells videos and other digits.

Sundialtone - Not Speaking Gibberish.mp3

For a blast of reality, use headphones to employ the space plan.

Observation, not judgment.

Sundialtone - Jay looks like a retard.mp3