For one of the elements of , I filmed a group of people reaching out with their arms. In the exhibition, a person can stand up against a wall and from there, the arms will be projected onto them. To an on-looker, it will look like the person is being attacked by the arms.
This footage was shot on a plain black background, and the actors moved their arms to the side and waved them around. When filming this, in order to get the arms to interlock from either side, the actors themselves had to be in the shot. For the final version of this footage, the actors will have to be cropped out, leaving just there arms in the footage.
For said crop, I tried two different techniques of masking them out; one that went straight down, cutting the scene in one place all the way along, and one that followed the layout of the actors tightly, creating more of a wavey cut. These were then projected and compared to see which one looked the most effective. The straight cut made the arms appear too harsh, whereas the tight crop made the arms seem more natural, due to them all having different origin points. Both sets of this footage can be found here:
However, this was not the only part that needed to be tested. Different background elements were tested for the projection. The options I had were: to removed the masked area, but leave the arm’s background, repeating this process, but using a clean slate to put behind the now empty mask space, or keying out the background completely from both the mask and the arms.
When projecting the footage, the shadows created by the actors made the wall behind their arms appear different to the recorded clean slate. This made the divide between the arms and the clean slate too visible, resulting in this method not being usable. This footage can be seen here:
The version without the masked background, but with the background still present was also not effective. This was because there was too much of a difference in the appearance of the footage between the ‘blank’ area and the area with a background. However, the ‘blank’ area looked the best when projected, and made the version where the entire area was ‘blank’ appear to be the most promising of the techniques.
As expected from the previous test, the best version of the footage was that with the background keyed out completely, leaving only the arms themselves. This made the arms appear from nowhere, and left no division between the space around the arms, and the place that was masked out. This information, and that of the straight/wavey test, revealed that the best version of this footage to use is the wavey cut one with no backgrounds at all; and thus it is the version that will be displayed during the exhibition. This footage, and the straight cut version, can be found here:
Once the final version of the footage had been decided upon, it needed to be tested on a real person, in order to test that the effect had the desired effect. If not, the footage would have to be re-recorded or further edited to create the desired result. Although some of the videos feature audio, this will not be used during the exhibition, so the audio will not be an issue in any instances of the footage.
When testing onto a person, the arms created the desired effect. The arms could be seen on the model in the testing environment, which was brighter than I plan to have the final location. This means that the arms will be clearer on a subject during the exhibition. From this test, it was clear that the arms would appear more clearly depending on what the model was wearing. However, this issue can not be avoided, as it is unknown what the audience members will be wearing during the exhibition. The area of the model covered by the arms can be changed with the positioning of the projector. Although the projector will be set up to a standard height, moving the projector can change where they are, in order to make it work on all the guests, regardless of their height.
