Interview by Tony Ryans 

July 2017

1. A fascination with surveillance-camera footage is probably universal, but the desire to use it to tell a story is quite unusual.  Can you remember how and why you first had this idea?

I first had the idea in 2013. I believe I was watching a ‘reality’ program about legal issues on TV which contained some surveillance-camera footage. I remember finding surveillance-camera material rather striking and intriguing. Whatever it was that caught my eye must have been something one doesn’t find in everyday photography and video. And what could that be? I think there were two things: the people in the images were unaware that they were being photographed, which generated a strong sense of reality, and the images themselves had a unique perspective and compositions. This uniqueness has nothing to do with the traditional compositions or aesthetics of photography. Rather, it’s something which underlines the reason that surveillance cameras are installed in the first place: to record as much information as possible within the frame. And so surveillance-camera images have qualities which are quite novel and non-conceptual. It struck me that using this kind of footage to make a feature-length fiction film would be a very valuable project.

I thought this idea was worth a try, so I embarked on the project. I asked friends, security guards and TV station employees to collect some surveillance-camera footage for me. To see if my idea would work, I experimented with footage from a hospital parking lot; I tried to make up a story based on the events and human activities seen in that video. The footage came with no sound, so I was able to add whatever dialogue and plotting I liked. The experiment made me fairly certain that my idea would be viable if I could muster enough resources. Unfortunately it was very difficult to access surveillance footage back then. The little that was available came from illegal sources. We had no choice but to give up the project. It was not until 2015 that a huge amount of surveillance footage appeared on the internet. There were even some websites streaming live video caught by the kind of surveillance cameras you see everywhere. So I resumed the project.

2. The film reflects not only the streaming of surveillance-camera material but also the rise in Chinese video-blogging and site-hosting.  Are the two phenomena related?

I didn’t set out to directly discuss surveillance in this film. The fact that we can narrate a story vividly by using surveillance footage tells us something about the way we relate to surveillance these days. It encourages us to think more about the human situations we encounter today.

During the four years we worked on this project, surveillance video developed rapidly. Live streaming online is developing even more rapidly. You asked if these two phenomena are related, and I think both of them have roots in the emergence of surveillance cameras which feed data into the cloud. But they’re different in certain ways. The use of surveillance sprang from the desire to control crowds and to monitor certain locations. Surveillance serves to acquire factual evidence and to govern it. These days, though, people’s attitude to surveillance and to its fundamental purpose has changed. At both the governmental level and the popular, first-person level, surveillance is now used very differently from the way it was used in, for example, the cold war period. Clearly many people now try to build connections with the world by presenting themselves online, hoping that this will change their lives. This is a fast-expanding field in China, and there are even training centers and textbooks teaching people how to become Internet celebrities. In some places the Internet celebrity phenomenon has made a significant contribution to the local economy. What’s special in this film is the way it uses materials which are constantly growing and changing.

3. The possible relationship between Qing Ting and Ke Fan is never really fulfilled or consummated – at least, not in the conventional sense:  there is a mismatch between her desires and his. Was it your wish to explore such a story from the start? Or did this story somehow arise from the material you were collecting?

Both Ke Fan and Qing Ting are serious people.  Ke Fan is a latter-day Don Quixote who fights against ‘the system’ and standardization, but his values don’t match Qing Ting’s. I think that kind of collision between ‘reality’ and the emotional desires of individuals has been a constant in human history. The mismatch between Ke Fan and Qing Ting is pushed to the limit by today’s realities.

Since we were working outside the usual conventions of filmmaking, our story and the materials we drew on grew side by side.  Each ‘discovered’ and supplemented the other in a continuous process. We began by collecting footage and then devising our story; as more visual materials became available, we revised the story again and again. In these last few years we’ve downloaded and analysed a huge amount of surveillance footage, in the process acquiring a growing understanding of situations right across China and, indeed, beyond China. Our understanding of the relationship between humans and their surroundings also grew, and we came to a new understanding of the so-called “boundaries of reality”.

Sometimes a surveillance-video feed remains hauntingly silent for hundreds of hours, and sometimes it takes only a couple of seconds for it to capture something unexpected, mad, and utterly irrational. A fragment like that can stretch our current knowledge to the limit or even change our view of history. It provides evidence of a phenomenon which we cannot understand or explain, although we know it really did occur. It makes me wonder what bizarre things happened in earlier eras without being recorded. Thanks to surveillance cameras, we can now see them. It would be marvellous if these objective records could be preserved for future generations to see, but much of the footage is wiped soon after it’s been recorded. Such thoughts which struck us during the process of making the film helped us to enrich and deepen the film’s themes.

In fact, the film is filled with a kind of tension. It’s a tension that springs from the gap between our privately felt emotions and the ‘invisible’ crises which are hidden in our day-to-day realities. Our film shows a world no-one knows about, a world seen through the ‘dragonfly’ eyes of surveillance cameras. What separates these fragmentary video recordings and what we think of as ‘reality’?

4. There is a sense throughout the film that the world beyond the story is dangerous and chaotic, full of man-made and natural disasters, worse than anything that happens to Qing Ting and Ke Fan. Should we call this expressionistic? Or just realistic?

The film isn’t actively expressionistic or realistic. First and foremost, it’s a poetic and metaphorical fiction. Seen that way, it has obvious affinities with expressionism. However, every single frame that we see depicts something that has actually happened. So you could equally think of it as more ‘realistic’ than anything else. In that sense, it’s an ultra-realist film which blurs our existing knowledge and affects our judgments of ‘reality’.

5. Why does the “Dragonfly” computer voice speak English?

Because it mirrors global search engines. The computer’s rather awkward use of single nouns – “man”, “woman” and so on – reflects the limits of today’s artificial intelligence. At the same time, though, it evokes ancient Chinese poetry and Japanese haiku. Aside from that, I wanted to distinguish the computer’s ‘vision’ from the story itself. I was looking for a slightly distanced structure, not unlike the transitions which occur in Chinese opera in between the stages and twists of the plot.

6. Nearly all the film’s sound has been created and added by you and your team. Can you sketch how this process worked?  Did it change or develop as the work progressed?

The sound design and dialogue in the film are very complicated and special. It may be one of the most “soundtracked” films ever made!  I realized early on that the sound would be crucial, since it would allow us to tell and amplify the story when we couldn’t originate new footage or work with actors in the visuals.  It also helped us to build scenes and characters in ways that the visuals alone could not.

By layering our sound we were able to create a sense of space. Nearly all surveillance camera is fixed, and provide an unchanging perspective on the space they cover, so they serve to distance the audience from the action to some degree. The first surveillance footage we accessed was silent, but more recent material includes sound recording too. That meant that we could make partial use in some scenes of the original sound.

I’m always interested in materials and techniques which change and develop.  In the case of this film, that interest took us into some unforeseen directions, as in the changing type of visual materials we collected. The footage originates anytime, anywhere, but our work in the studio is in sync with what’s happening in China at this moment, even with what’s happening further afield. We recognize no distinction between what goes on in our studio and what goes on outside of it. In that sense, our ‘cameramen’ – the surveillance cameras – are stationed all over China and even all over the world. They produce remarkable footage for us 24/7, non-stop. There are some 20 computers in the studio and we used them to search for and to download the surveillance videos we needed. One example:  we needed footage of a car driving along a mountain road in the rainy night. First we checked the weather forecasts to find out where it was due to rain. Then we locked onto a surveillance channel in that area. The next day we’d check to see if we’d managed to ‘harvest’ the images we needed.

The process affected the members of our crew psychologically. By the time we were some way into the project, some of us became very cautious whenever we went out. After sifting through all the surveillance footage, we developed a deep feeling which we had no control over the environment we lived in. Anything could happen.

It’s the synchronization of the film with the social environment which makes our work rather innovative. Today’s world is like a spectacular film studio, continuously photographed by ‘our’ surveillance cameras. I always say that our working method is like Uber’s. Uber doesn’t own a single car, but so many vehicles in our cities work for the company. So I feel that our working methods are in tune with the ways our society and technology are developing.

7. Do you feel any connection between the film and your other work in visual arts? Or is the film a new departure for you?

I’m a visual artist and this is the first time I’ve made a film. If I have more ideas for films in future, I’ll pursue my filmmaking and this film will mark the starting point. But I don’t know. I don’t believe that anyone’s artistic path is designed beforehand. Rather, it’s discovered afterwards:  “Oh, it turns out that I’m interested in things like this and I work like this”.

However, I can certainly say that this film is a continuation of my working methods and my attitude towards art. All of my art projects look different, to the extent that it seems they haven’t been made by the same person. But deep down they share the same spine, and they explain and highlight aspects of each other.  Dragonfly Eyes is no different. Take my past work Background Story, which transforms the invisible into the visible and vice versa. The viewer looking at Background Story sees a classical Chinese landscape painting, but it’s actually not a painting made on paper or canvas; it’s made from abandoned materials which are arranged to create light effects mimicking landscape painting. I’ve always liked to use small, even meaningless materials in a serious way to produce a substantial ‘fact’, which is actually a fiction. Thirty years ago I made Book from the Sky, a series of scrolls covered with ‘fake’ Chinese characters of my own invention. I also like to play with the boundaries of human knowledge and concepts. In my work Square Word Calligraphy there’s a writing system that looks like Chinese characters but is actually English. When you look at such calligraphy, none of our current knowledge is much use. Similarly, we cannot define Dragonfly Eyes as either fiction or documentary. Years before creating Dragonfly Eyes, I made Book from the Ground, a storybook written in signs and icons that everybody can read. As with Dragonfly Eyes, it took us years to finish the work; the widespread popularity of emojis eventually helped.  (There’s more about this at www.xubing.com if you’re interested.)

I don’t approach art through any past art style or school of art, because schools and styles are invented by artists as ways to approach the issues of their time. To discuss today’s issues, I need to use today’s way of talking. I can’t find that in any pre-existing system. I can find it only by exploring today’s lively social contexts.

As I finish answering your questions, there’s one more thing I want to say.  As the process of making this film drew to a close end, I felt a strange unease. What was the relationship between us, the film crew, and the strangers who appear in the footage? There’s a new kind of relationship between the watchers and the watched that we can’t yet fully understand. Where is the boundary? This drove me to track down the people who appear in our images and to ask about their real lives and their attitudes to privacy. We spent a lot of time and effort analysing the information from the websites; we found most of the people from the images, and obtained their approval for using their images in the film.

I’m very grateful to everyone in our team, and to those who find their images co-opted into our work. Together, we created something out of thin air. No such film has existed before. Thanks to our work, it does now.