More selected projects


Luke Dash

I always wanted to try making a piece that involves the identity of the spectator very directly. I find it fascinating in galleries, how sometimes a sculpture has a reflective surface, and you can watch dozens of people take a moment to snap a photo of their reflection in the work.

Daniel Rozins mirrors were a big influence.

Aiming to make a mirror that reflects the way we try to curate and capture ourselves in the context of social media, I wanted to try and project a more fluid representation of a viewer. Something amorphous, and hopefully a little bit pretty. To me there's emotional value in being shown to yourself from different angles, offset against familiar contexts.

The approach I took visually comes from playing with the inspiration for the mirror (a smartphone). In particular, finding long exposure apps (I think intended for light trails) and turning the camera on myself. The code aims to reflect that simple experiment. Different layer blending modes are procedurally applied to a buffer of frames captured over the last few seconds. The result is a strange light painting of the spectator, whose palette and forms are modulated slowly and asynchronously.

Since the main thing influencing this work is hooked into social media, as well as bringing fluidity to the spectators' presence, I wanted to try to work with the presentation a little to bring a bit of an obtuse interruption to the self-image. So, on occasion, the buffer of past frames gets shuffled and spammed over the face of the spectator. They see themselves obscured, but in full view simultaneously. Fluid, present, and a bit displaced. A similar behavior which resembles a broken scrolling news feed where the images from the buffer are randomly cropped is also possible.

Since art gallery selfies seem to be a pretty commonplace thing to do, my hope is that the piece can be a little provocation pointing back to the viewer, nudging them to consider how they might be looking for themselves when they look at art.

The project was facilitated through experimentation with a piece of code which establishes a delayed buffer of video frames in real time, kindly provided to our class by Theodoros Papatheodorou.

The program also uses simple face tracking to make the chaotic visual elements more connected to the live input. This is achieved using an openFrameworks addon called ofxOpenCV.

This project felt like an interesting first step toward a body of work that might begin to gently interfere with discussions surrounding identity and privacy, so I was pleased with the outcome to that end. I think the main thing with this piece in particular that I wish I had more time to try is the delivery and presentation. I’m now pretty excited to try something similar on a swath of different mobile devices. I think this would elucidate the concept to a spectator much more naturally.