Leo Villareal, Lighting It Up

Steven Johnson talks to the creator of San Francisco’s iconic Bay Lights installation about the possibilities of algorithm art, his embrace of VR, and the new challenge of lighting the bridges of London.

Steven Johnson
Oct 30, 2017 · 11 min read

Occupation: Light sculptor

Projects: “Multiverse” at the National Gallery in Washington, DC; The Bay Lights in San Francisco; soon, The Illuminated River in London.

Favorite tools: Unreal engine, Photoshop, Oculus, custom software for controlling LEDs

Workspace: Loft in Brooklyn’s Industry City.

How does a project begin for you? I mean, is there some very early stage where you’re actually sketching something out by hand?

Not really. I’ve never been much of a sketcher. It’s always been more graph paper — but mostly digital. I got pretty comfortable early on with Photoshop and that’s how I draw.

The fascinating thing about your work is that there are these two layers to it: both the physical shape of the piece but just as importantly, the algorithm that controls the patterns of the lights. How do you work out ideas for that algorithmic side?

My very first pieces were using simple micro controllers, programming them in BASIC. So zero is off and one is on, and I’ve got sixteen lights and I can make sixteen zeros and ones. So really lowest level, and then starting to build sequences up from that, maybe adding adding some randomization. And then trying it out, and seeing it, and then adjusting it as needed. Eventually I started working with LEDs and some off the shelf software from Philips that was kind of like working with a spreadsheet. And it was okay — I made a bunch of pieces with it. But I could see the potential for talking to the lights in real time. Having real time control of the lights was a big leap.

What tools allowed you to do that?

There were a lot of different tools: Macromedia Director, some software from ITP, eventually I got into Max, MSP, Jitter. Eventually when I did my piece for the National Gallery we were suddenly working with 41,000 lights and all the software we were working with at that level croaked. So then we had to go into custom development of an application that would do this. And that was a really exciting moment. It’s been an ongoing collaboration with many programmers, and we’re still developing these tools. I never thought I’d have to be managing software development. [laughs]

Do you have a coherent image in your head ahead of time, or are you messing around with the tools until something happens?

My second light sculpture ever was a portrait of John Conway’s Game Of Life. I really interested in that and how these really simple rules could create this patterning that felt like it was alive. It had so much richness to it that I thought: I’m going to work on my own sets of rules. So that was a very early interest, and in a sense I’ve been pursuing that ever since. And so a lot of that is about not really knowing in advance what the sequences are going to be, and trying things out and being very improvisational with the code. It’s a lot like playing an instrument, and waiting for that one percent of the time where something interesting happens, and capturing those moments.

In my amateur musician life, I used to have this great little device that you’d attach to your acoustic guitar, and every time it detected you playing the guitar, it would spring to life and record on a thirty-minute loop, without you having to consciously turn it on, press record, etc. It would just record in the background, and if you played for more than thirty minutes, it would recording a new loop and erase the old stuff. But if you played something interesting, you could stop and say — save those last thirty seconds, or two minutes, or whatever. It was a really interesting idea, just passively capturing all your noodling but letting you pinpoint the good bits.

Yeah, definitely. I would love to do more of that. The problem is now with the new stuff I’m doing, one minute is like 4GB of data. But in terms of the sequencing data, it could be possible. Now I’m thinking more about doing generative work, things that aren’t recorded. I love the generative idea, but I still need to be there as an editor, but I don’t really trust the software to look good all the time. It ends up going off the rails pretty quickly.

Yeah, you don’t want to show up at the Bay Bridge and say, “What the hell are you doing?” (laughs)

Right! Of course, I could write myself out of a job and let the AI make the art.

(Villareal tuning “The Bay Lights.”)

Let’s talk about The Bay Lights. So when you were working on that in the early days, you would sit there in front of the actual bridge and experiment?

Yeah, we put 25,000 lights on the suspender cables of the bridge and spent months doing it, testing, debugging, all in full view of anyone who could see the bridge. I mean, we couldn’t really put a curtain in front of it. (Laughs) So we had to be comfortable exposing all the glitches, and trial and error. But finally when it was working I had to figure out what the piece was going to be. And then only way to do that was to sit in front of it and try things out. And so I did that for months. I had a wireless link to the computer that was controlling it. It was kind of the ultimate party trick. People would realize I was connected live to the Bay Bridge and suddenly there’d be people around me and much excitement. It was fun.

When you reinstalled the LEDs a few years ago, did you redesign the patterns as well?

When we did that, the LEDs themselves had evolved since we originally put it up — they were more than two times brighter. So I had to kind of balance it out and think about what the original lights looked like, because we didn’t want to have all of sudden something that was just three times brighter. The balancing aspect is really important — the tuning — you want it to be there and present and you can look out of your window and see it. But it’s also ambient, adding another layer to things without being overwhelming and demanding that you look at it. So you can kind of dip in and out of it, and because it’s randomized you’re never going to see the exact same progression twice. And depending on where you physically are, it changes based on your perspective, what time of day it is, whether there’s fog or not. So that gives it a lot of variability beyond what I’m doing with the sequencing part of it.

You used that word before — “tuning.” It’s a really good word.

You know, with my software I have total control over the lights. Zero is off and 255 is the brightest level. So you have very precise control over each light. So when I’m working on a site specific piece, I’m really working with it, looking at it, adjusting — especially with a piece like The Bay Lights, it was really important to see it up close and far away. Somethings were illegible from a distance but up close you could see finer detail. It was really important to be thinking about all the different audiences and baking it so that it was working for everyone as much as possible. And I think that’s what draws people to my work. Many many pieces I’ve done have started out as temporary works and then they become permanent because people just love them. And I think that’s because they’re done in a way that connects to their site and is responding in a way. It’s not literally responding, like there’s a sensor and its responding to traffic. But all that stuff that’s happening is filtering into the sequencing. Like with the Bay Lights, the water, the traffic, the atmosphere — all that stuff is in there. And there are certain moments where you might see a connection between those two things, and your brain can’t help but match them together. For that reason, the pieces almost feel like mirrors of what’s happening around them. And that all requires a bunch of tweaking and adjusting. Moving light is something that can go badly quickly.

Right — your eye literally cannot not look towards it if it’s too hot a signal.

Exactly. It can be used in a really aggressive way. That was one of the big concerns for the piece at the National Galley. The people there were like, wait, 41,000 lights in an IM Pei tunnel — I mean, are people going to have seizures, etc? I had to do a lot of reassuring.


Let’s talk about how you prepared for the big show you just had at Pace Gallery in Manhattan…

So I joined Pace Gallery recently, and I was going to have my first show there in 2018. And then we got a call in February of 2017 saying that another artist who had been planning to have a show wasn’t ready, and so there was a slot open — and so the question was: did I want to do a show in May? We’d been in a good zone here in the studio and over time I’ve been able to build up my team and have amazing help with visual effects and fabrication. So we had some discussions in the studio about whether we should do this. And one of things that’s been incredibly useful is that we’ve started to use Unreal Engine as a quick way to visualize what an installation or a exhibition would look like. So we jumped in immediately and modeled Pace —

You built a 3D model of the entire gallery in Unreal?

Yeah, I mean, it wasn’t very complex. But there were some interesting architectural details — this vaulted ceiling. So we thought about that space about what would be appropriate in it. And so we got to try it out. I got to put on the head-mounted display and walk around the show in VR. The visual quality was really astounding. We were able to take actual sequences I was making with my sculptures, put them on the virtual versions of the pieces, and then start to move around. The largest and most complicated piece in the show is called Ellipse — it’s this hanging rod structure that we tried maybe thirty or forty different iterations of different arrangements of rods, because again it was a complicated space. It was a remarkable way to be able to try these things out.

So you tried those thirty versions entirely virtually?

Yes. In VR. We could move around the space, and we were using these mirrored stainless steel rods, so we could simulate the reflection of the LEDs, and see it in response to the space itself. With my regular old 3D tools I would have never been able to experience all that. The next step will be to be able to author the piece directly inside the virtual world. And in Unreal, there’s some of the beginnings of that where you can actually be in the world, using the tool to make the world. I can imagine going full-on Inception mode in there and like never coming out. [laughs]

And then there’s London — a slightly more complicated space to render.

Yeah, in London we’re working on fifteen bridges, from Tower Bridge to Albert Bridge. I won a competition in December for this project called the Illuminated River. The idea is to both illuminate the bridges and create a sense of cohesion between the bridges. Some of the bridges are iconic works of architecture, and others are concrete road bridges that get you from A to B; others are rail bridges.

(Rendering of the Illuminated River.)

And you’re doing the pedestrian bridges too, right? Some of those are beautiful.

Yes, they are. Millennium Bridge, designed by Fosters, is amazing, and the Golden Jubilee bridges designed by Alex Lifschutz, the architect I’m working with. So we are working with some contemporary bridges, but also Westminster Bridge, which is a UNESCO heritage site. So there’s a lot of apprehension, but also excitement too. I think our approach is to put light where light has always existed on the bridges as a starting point. And then add the element of sequencing to it. Very subtle movement over time. Because you know, the Thames itself is so remarkable, and it rises and falls five meters twice a day. It really is this living thing snaking through the city, so the idea is to take that as an inspiration for some of the sequencing that we’re doing. Make it almost as if it’s a mirror of the activity of the water, and the traffic and movement of people.

So what we’ve done is that we had our team in London scan the bridges, and we got these models and put them into a virtual world and created these panoramic environments around them, and then we have a real-time Thames that we can control the speed of and its reflectivity. It’s a beautiful thing to be able to simulate the water and have these real-time reflection. And there’s one-to-one mapping on the lights. So every single place we are putting a light on the real bridge, we have a simulated light on the virtual bridge. And we’ve done some testing with putting actual test lights on the bridge, and comparing how it looks to the VR version, and it’s extremely accurate. So that is allowing me to do a huge amount of work in advance, with visualizations of all fifteen bridges. I mean, nothing will replace being able to go there and sit in front of the bridges for weeks and months, tweaking and adjusting. But this is an extraordinarily powerful way of working.

You’re working with color now, right?

Yes, though with some of the more modern bridges, like Millennium, with those I feel like we need a more monochromatic treatment of those bridges. I just think colored light doesn’t look good on them. But there are other places like London Bridge or Waterloo that can be more saturated because they’re in more lively parts of town, where there’s a lot of nighttime activity.

There are some weird lighting on the south bank side, isn’t there — on the National Theater, right?

Yeah, we have a rule that we’ve banned primary colors. We’re doing all intermediate colors. We’re hoping to set a good example for others. As LED gets more prevalent and inexpensive, every surface seems to be encrusted with lights. I mean, there are lots of really bad uses of light. We’re trying to figure out how to set a good example, and inspire others to use some moderation — I mean, just because you can make sixteen million colors, you don’t have to use every single one!

(Villareal at his Industry City studio.)

Thanks to John Battelle

Steven Johnson

Written by

Writer. 11 books. (The latest: Farsighted: How We Make The Decisions That Matter The Most.) Host, How We Got To Now. Three kids, one wife. Marin/Brooklyn.


About this Collection


An interview series about the tools, techniques, and habits behind the creative mind. Hosted by Steven Johnson, featuring conversations with Liz Phair, Rebecca Skloot, Kevin Kelly, and more. Produced in partnership with shift.newco.co.

An interview series about the tools, techniques, and habits behind the creative mind. Hosted by Steven Johnson, featuring conversations with Liz Phair, Rebecca Skloot, Kevin Kelly, and more. Produced in partnership with shift.newco.co.

Welcome to a place where words matter. On Medium, smart voices and original ideas take center stage - with no ads in sight. Watch
Follow all the topics you care about, and we’ll deliver the best stories for you to your homepage and inbox. Explore
Get unlimited access to the best stories on Medium — and support writers while you’re at it. Just $5/month. Upgrade