After my first test in weymouth house I found a problem with the speed of the video being an issue. So today I tested my project again after I changed the video to the original version to see if this would be an improvement. The original video seemed to work better when a face was seen, but due to peoples faces not being recognised from further away it still made it difficult for people to understand what was happening. Without moving the position of the camera I am unaware of how to fix this. Unfortunately moving the camera away from the screen was not possible as it was connected to the screen by a wire.
My predictions for the testing were that when lots of faces are seen, my piece may not work as well, as the video will constantly be playing because of all the faces. As well as this, when there are lots of people around I don’t think people will feel comfortable walking slowly past a screen and interacting with it. During the quieter time I was hoping that people may be more curious to investigate. During testing I compared my predictions to what was actually happening. Due to the faces not being detected very well when not in close proximity, there was no problem with the video constantly playing when there were lots of faces in the room. When the room was busy I did notice that when people were walking through the space they would move quickly instead of dawdle.