Sowmya Rajendran: Projection Proposals

Sowmya Rajendran

___________________________________________________________

Concept
For this project, I primarily focused on projecting onto sites and architecture located in the northern portion of the UT Dallas campus. I rarely venture into this part of the campus, so I decided to explore the spaces and architecture in this area. Most of these buildings and spaces are pretty bare and exposed, thus giving the perfect opportunity to project onto these surfaces.  With my compositions, I wanted to focus on the synthesis of the physical and digital world. I tried to emulate layered patterns present in the architecture itself, through explorations in similar visual structures seen in the digital space -- such as pixels, mosaic fragments, and grid structures. 

If these compositions were to be executed, I would love to find a way for viewers and passersby to become an active participant in the piece. A possible scenario would be to place sensors that would record and react to surrounding movements. This would, in turn, affect how the visuals are represented on the structure itself and create a space for viewers to interact and converse with the architecture.   

Adrien M and Claire B are French artists who utilize data sensors to create reactive projections according to movements in the given space. To see more of their work, head over to their Vimeo channel.
 

Process
First, Maryam and I went location scouting to find suitable spaces for projection mapping. The three images I choose were located near the Founder's and Jonsson building. The first space had a really interesting shape to it, with its stacked and geometric form.  For the second space, I was mostly interested in the jagged, zig-zag formation of the walls and wanted to explore how to create a sense of movement within this formation . Lastly, the staircase was also interesting with its step lights and glowing walls, submerging viewers as they ascended/descended the steps. For this space, I wanted to create an immersive experience for viewers with explorations in abstract textures and animations.

After finally deciding on the spaces I wanted to activate with light, I took these images into Photoshop. They were cropped according to the resolution size of the projector. Before I could start playing around in MadMapper, I had to plan out which spaces I wanted to project onto, so I could separate and mask overlapping objects, such as trees, trash cans, or light poles. In addition, my first photograph was taken during the afternoon, and I wanted to transform the image into an evening/dusk scene, so that the projections would contrast better when composited. This was resolved in Photoshop by adjusting curve levels and applying a blue-toned photo filter effect to the photograph. 

All the images and masks were exported into MadMapper and applied as quads. I setup my scene with the appropriate quads I needed for the space. For content, I aggregated all the videos I made in After Effects into Modul8; however, I encountered lag as I was importing media into the program. I remedied this by compressing my videos via QuickTime. These videos were further edited with different filter effects offered by Modul8 to create the desired, pixelized patterns & textures I was going for. I increased the size of the output window in Modul8, so that I could fit three videos into each composition layer. After I achieved the desired designs I wanted for each video, I syphoned the content to MadMapper and applied this content to the individual quads. I then moved the quads in the UV field to achieve the desired compositions. The final renders were recorded through Syphon Recorder and edited with YouTube Editor. 

Finished Pieces