Bloccupy Revisited

I recently updated <Bloccupy>

it’s an interactive media installation with kinect/video input, which I built for Projektil, Zurich in 2015.

It wouldn’t run anymore for various minor reasons. plain old code-rot and no devs around to prevent it.

while getting rid of some dust and freshing it all up, I did a nice screencast I wanted to share 🙂


I also wanted to summarize the basic building blocks and concept used for Bloccupy.

As most graphics-related things I did in the past years, it relies on crossplattform C++ and OpenGL.


pin toy metaphor


The underlying metaphor for the concept is a pin-toy.

Like in other projects I really fancy the procedure of downsampling something close to the point where it becomes unrecognizable and exploring this borderland.

To achieve the blocky look and elevation behaviour I manipulated a simple array of points in 3D space.

Those are fed into a GLSL geometry-shader and transformed into cuboids.

As input we used a raw depth-camera image in combination with video-footage, the update of point positions is performed with a couple of OpenCL kernels.

The depth values and/or color information from video is mapped onto the elevation of the cuboids and their color.

The scene is shaded and lit with a couple of point-light sources, to be able to visualize the intricate details of the resulting geometry.



optionally a colomap can be applied to better highlight different elevations.

I used this one, called “hot iron”:



nothing more to add, enjoy this outtake from Carl Sagan’s Cosmos, rendered by the Bloccupy Engine.


hope you like it,