Reminiscence: My first SIGGRAPHs
After reading the 50th Anniversary reminiscences about SIGGRAPH, and, attending some of the SIGGRAPH 2023 retrospectives, I thought I’d share what I often think about when I think about my early years at SIGGRAPH, and how one’s experience changes over time.
1996
In 1996, I drove the six hours from Houston to New Orleans to attend SIGGRAPH. During my grad school visits, several professors had recommended that I go. I’d just graduated from college with a double-major in art and computer science, two passions I viewed as entirely separate. I had spent many hours in college writing code and doing math, and, separately, many hours drawing and painting pictures. And I loved many other things as well, including movies, literature, experimental animation, video games, fine art. I had no idea what I would do in the future, except that, having had a real job before, going to grad school in computer graphics seemed preferable to a real job.
SIGGRAPH felt like a wormhole to a wonderful new world. Movies, ideas, math, technologies, animations, games, experiemnts together, all at the cutting edge, all things I could see nowhere else, all driven by people sharing the same weirdo passions as me. The very first papers session I attended, including Steve Seitz’s View Morphing, Paul Debevec’s Façade, two light field papers blew my mind. This “Image-Based Rendering” seemed like magic, that you could make new 3D views just from photographs.
In the Electronic Theatre, I watched experimental animations—often edgy or dark, experimental uses of graphics technologies you couldn’t see anywhere else. And these sat side-by-side with Hollywood visual effects reels, and educational scientific visualizations. In the E-tech and art gallery I saw all kinds of other experiments. I recall spending a lot of time playing a video game demo from some HP engineers.
Throughout the program, artists used new technology to create something new. Fine art and popular art all lived together, without the usual judgement that one kind of art is “too popular” or “too commercial” or that another is “too weird, too pretentious.” As long as it used or explored these evolving technologies, it was welcome. At the same time, scientists and engineers demonstrated entirely new ideas and technologies, new ways to make art.
I wanted to be part of this community. I saw people on stage presenting their new works, and I wanted to do that too, and grad school offered a path to doing so.
A high school friend flew down for SIGGRAPH and we explored the city a bit together. Apart from that, I don’t remember having any conversations that whole week that lasted more than a few sentences.
The next few years
I started grad school excited about Image-Based Rendering, the magical idea of making 3D navigation just from a few photographs. I’d been inspired by a demo that Leonard McMillan showed me during by grad school visits, in addition to that first session at SIGGRAPH. It built on computer vision, and the computer vision courses I could take were mostly 2D-focused, hoary old techniques like Hough transform and Gestalt grouping. So I spent much of my first grad school years teaching myself the relevant 3D computer vision, and also drifted into reading papers on machine learning, which was also not taught at my school.
But at SIGGRAPH 97, I sat in Pete Litwinowicz’ talk on his painting algorithm, and, when he said “this is how artists paint,” I thought, “this isn’t how artists paint.” Or, at least, it wasn’t how I paint. I went back to the lab and started playing with his algorithm, and Haeberli’s Java demo, and at some point I thought it might be worth trying to publish what I had. My little algorithm got accepted to SIGGRAPH 98.
Because of my paper, I really started to meet people at SIGGRAPH 98. I remember bonding with Paul Rademacher, another grad student publishing his first paper, also a single-author paper like mine. I met many lifelong friends and colleagues in subsequent years. For example, the chair of one of my paper sessions in 2000, David Salesin, emailed us paper authors suggesting that we meet over breakfast at the (to-me) ungodly hour of 7:30am. David later hired me for three different jobs at three institutions, including the one I’ve been for the past decade. Before SIGGRAPH 2001, fellow grad student Alyosha Efros introduced himself because we had similar papers in the same session. I’m still friends with both of them.
Downtown LA sure has improved since then—and I’m glad the conference goes to Vancouver now—but there’s just one thing I miss. Because there few were places to hang out at night, it seemed like everyone sat around the pool at the classic Hotel Figueroa. After dinner, if you were on your own, you could always go to the Figueroa pool and find people to have a drink with. Now there are so many, better options—especially since the Figueroa’s renovation—and I don’t think such a focal point exists anymore.
The SIGGRAPH week was always full. For the day after SIGGRAPH 99, I made plans to go hiking with a college friend. When I got to her apartment in Pasadena and we caught up, I fell asleep on her couch for like five hours.
By the time I graduated, I started reviewing papers, which felt like an honor; I had lectured in SIGGRAPH courses, had finished several internships in computer graphics research labs, and had published four first-author SIGGRAPH papers with coauthors from four institutions. I felt like a real part of the community.
You Can’t Attend The Same Conference Twice
At my first SIGGRAPHs, everything excited me. Being there made me so happy to see all the new art and technology in so many forms. I remember one conversation where a colleague told me she loved my enthusiasm for the conference. I made a point of seeing all the entire animation festival (not just the Electronic Theatre), going to all the art gallery and other exhibits. Every year, it all got better and newer, with technical and artistic explorations surpassing the previous year. And I met artists and animators, and people theorizing about the nature of digital art, and debating hard questions about whether it’s even possible to tell real stories in an interactive medium.
But, when I spoke to older attendees, I kept hearing jadedness: “all the papers are incremental this year, there’s nothing exciting.”
Within a decade, I heard myself also saying the same thing: “Nothing’s exciting this year, all the papers are incremental.”
With some experience and seniority, it became easy to judge everything as obvious and predictable. I would read papers and think “I already had that idea.” As a paper reviewer, I probably treated submissions as guilty until proven innocent.
With enough years at SIGGRAPH and learning about the history of science, I grew to understand the importance of “incremental” research. Moshe Vardi told me the analogy that science is like building a cathedral: each brick is necessary.
Psychology research tells us that we tend to most value the artistic experiences from our formative years. This is why some people say “there’s no good new music”—and all the good music came out by the time the speaker was a young adult. We often seem to be unaware how much our artistic judgements and taste are tied to our coming-of-age experiences. Sometimes it seems like conference planning is governed by the expectation that every paper should make us feel like we felt about our first favorite papers.
Over the years, I’ve urged people to go to SIGGRAPH. I want them to love it as much as I do. But no one’s going to have the same experience that I did when I was 24 years old, attending SIGGRAPH 98. I don’t have that experience now.
SIGGRAPH today
I often wonder how it seems to newcomers today. We have a daily stream of amazing new research on arXiv. The experimental art and animation don’t seem so singular compared now that you can find so much on YouTube and Vimeo. Computer graphics has taken over feature film and television production; you regularly see all kinds of stuff in movie theaters that you once could only see at SIGGRAPH. And, apart from arXiv, so much of the energy around new image synthesis research is at the computer vision and machine learning conferences. The world now is different from 1996.
But I still love witnessing what happens when tinkerers and enthusiats for both art and technology start exploring, how lucky one is to be around as new art forms develop, even more the few who create them.
I’ve attended every SIGGRAPH since my first, 27 years ago (virtually in 2020 and 2021). I’ve stopped attending the Electronic Theatre in person, and often just take a quick browse through some of the other exhibits. The papers program this year still seems really good, lots of great new research happening, and I’m really glad to have coauthored two new papers that I couldn’t have published at vision conferences. Attending this week in LA, and reconnecting with so many old friends and colleagues, has been a blast.
At times, there are other conferences that seem more current and exciting, but SIGGRAPH feels like my “home conference.”