Deeper Behind the Work: Chris Milk on Arcade Fire's The Wilderness Downtwon

By Published on .

Last week we heard from B-Reel on the creative, technical challenges of Milk Koblin's Wilderness Downtown interactive music video for Arcade Fire. Here, Chris Milk steps in to discuss the director's point of view on bringing the project to fruition—and what it was like expanding his viewpoint onto a digital canvas.

Director Chris Milk
Director Chris Milk

How did you get involved in this project?
I'm friends with both Aaron Koblin at Google and Win from Arcade Fire. The band was recording their new album, and Google was interested in doing something to showcase what was possible on the web using HTML5. It seemed like a good fit.

What was the initial brief-from where/whom did it come—Google or Arcade Fire? Presumably, it was designed to showcase not just Arcade Fire, but also the Chrome Browser—how did you go about addressing both of these things in your treatment?
There was no real brief. Google wanted something cool in HTML5, Arcade Fire wanted something cool. That was really the extent of it. I was more concerned with finding something that would emotionally resonate with people, without bogging them down in the technology. It's easy to lose the humanity when you start showcasing tech. Google maps and Streetview embody the contradiction though. It's cold high-tech that can be incredibly emotional when used in the right context. The whole piece is full of contradictions. It's human nostalgia produced by the most advanced technology available today.

Can you describe what aspects of the project you were involved in— it was called a Milk+Koblin production, but there were other partners involved, including B-Reel.
I conceived the idea and directed it through the film production, post-production, and web production. Milk+Koblin is Aaron Koblin's and my little company for special projects. We've done three so far—the other two projects were Johnny Cash's final video, and a section of the current tour visuals for U2.

So The Wilderness Downtown grew out of our relationships at the beginning— Aaron's with Google and mine with Arcade Fire. But there was an army of talented dedicated artists that brought it all to life—probably the biggest being the interactive team at B-Reel, programming extraordinaire Mr.doob, and the producers at @radical. Media. @Radical really goes above and beyond when it comes to supporting me in these crazy endeavors. We did the entire Johnny Cash Project in-house for essentially nothing budget-wise.

How did you and Aaron Koblin first hook up? Difficulty-wise, how did this project compare to Cash?
Aaron and I met at a technology conference in Portugal two years ago. This project was equally as difficult as Johnny Cash, if not more. Being that it was all HTML5, we were trying to do a lot of things for the first time on the platform.

How did the collaboration with B-Reel work? What was it like working with multiple creative partners?
Working with B-Reel was great. They really took what was a crazy idea and figured out how technically to actually bring it to life. It was far from simple. There were things at the beginning that everyone said were impossible, but they figured it out. Our producer Nicole Muniz and creative director Ben Tricklebank (both of B-Reel) brought so much to the table and really made the whole project possible. Also on the team, as I mentioned before, was our mercenary super programmer Mr. doob, who Aaron and I also worked with on Johnny Cash.

Working with multiple creative partners is an integral part of filmmaking. It's the same as collaborating with a cinematographer, a production designer, or an editor. It's a collaborative art form. You find the most talented people you can to help bring the overall vision to life.

As a director, what were the most interesting aspects of the project? What were the most challenging parts/biggest obstacles, and how did you tackle them?
The most interesting part as a director is working in a new medium, with a new set of rules. It's a completely different canvas to work with physically, and that in turn steers things conceptually. The browser as the canvas allows for such a larger dialog with the viewer. There's actual two-way communication going on between the art and the observer.

It's much different than film which is a very linear one-way medium. Working with the browser as your broadcast medium though presents a whole set of pitfalls you never have in film/TV, because everything you do is ultimately programming code. In film if you change something in post in one of your shots, you might make it a little better or a little worse. In code, if you change one thing in one shot, suddenly your whole piece doesn't play on a Mac anymore, and it takes three guys all day to figure out where the code got messed up. That can get a bit frustrating.

Showing off the features of the Chrome Browser—was this foremost in your mind when you were conceiving the story? How did it affect the way you approached the project, compared to doing a more traditional music video?
The project was not really trying to showcase the Chrome browser specifically. It's about HTML5, which Chrome runs extremely well. But the piece will play great in most modern HTML5 compatible browsers. It's really smooth in Safari.

No one ever asked me to use any specific Google technology or products. I wrote Maps and Streetview into the concept for their emotional resonance rather than being a commercial for Google. Plus I knew this might be the only project I'd be able to get away with it given what we needed to do with the dataset.

Also it's important to note that we haven't purposely excluded any browser. That seems to be a big misconception. This is not a walled garden we built. Some browsers are just not HTML5-compliant yet. Amongst the browsers that are, Firefox doesn't render canvas elements fast enough and IE9 doesn't support a JavaScript feature which we've used throughout the site. They will in the future I'm sure. The site will work on most everything I would imagine by the end of the year. And it will run like a hot knife through butter when browsers start getting webGL.

Also Chrome 6 came out on Thursday and everyone should upgrade. It runs like an absolute dream on there, even with tons of tabs open. I kind of wish we had launched four days later now.

Production logistics—how long did this take, from conception to finish?
I had been talking to Win about it since January. At the time they didn't have any songs finished yet. I listened to an early unmixed version of the track in March. Went to see the band in Montreal in April. Wrote the concept in May. Started physical production and web development in June. Shot in July. Finished at the end of Aug.

What were the biggest lessons you learned from this project? If you were to do again, what would you have done differently?
I would have started production in January to finish in August.

Anything else interesting we should know about in regards to this video?
Where it gets really interesting with this project is the second level of social interaction, beyond the interaction the observers have with the piece. This will unfold over the next few months as people are able to anonymously connect with each other through the postcards created in the film, and output through a number of both digital and physical mediums—one being the traveling Wilderness Machine.

Blueprint for The Wilderness Machine
Blueprint for The Wilderness Machine

The machine itself is a bit of a surprise I'd rather not give away completely right now. The website provides a bit of a hint: "A postcard is created by an analog signal: you. This site takes that postcard and converts it to digital. The Wilderness Machine brings it back to analog. Look for it on tour with the band in North America. If you're lucky enough to get someone else's postcard from it, plant it. A tree will grow out of it."

Logistically speaking, it takes the message in a bottle scenario to its most unnecessarily complex technological extreme.

Most Popular
In this article: