Consider this animated gif of a racquetball hitting a wall. The motion has the appearance of a video clip of a single event, that is, one ball striking the wall and rebounding from it. This might be what one would obtain, for example, from a high-speed motion picture camera. However, that is not the case here. Each frame of the animation is of a different event. Thus, it is more correctly called an animation rather than a video in the sense that the sequence is produced artificially. The events themselves are real. The sequence is constructed to show as closely as possible how the shape of the ball changes during the collision. For a larger version of the animation, click on the image.
Here's how the animation was created.
- Filming: The collisions were filmed with a Hi8 camera using the techniques described in the Filming section. The flash unit was triggered by two metallic plates that were forced into contact during the collision, thus closing the trigger circuit. A timing circuit was used to delay the flash discharge by a different amount for each collision in order to capture the ball in different stages. (The delay circuit is described here.)
- Capture: The frames containing images of the racquetball were transferred to a computer using the freeze-frame capture method.
- Image editing: Since each image was of a different collision, there was some variation in where the ball struck the wall. A photo-editing program was used to bring all the images into registration to the extent possible. The images were also given a uniform size, density, contrast, and color balance. The individual images, assembled into a collage, may be viewed here.
- Animation: A gif animator program was used to sequence the images for the animation.
Video clips were also produced using Adobe Premiere. Here's an example:
These clips used 26 images to provide a smoother appearance of motion. In order to make the transitions between images even smoother, dissolve transitions were used in Premiere. This also causes the clip to run slower. The results are shown in the following files:
All of the above clips are one-quarter the size of the original video frames.
Note: The technique described above for creating animations only works if the event of interest is reproducible in both time and space. That would not likely be the case if the racquetball were being struck by a racquet. For these photos, the racquetball was shot from a large slingshot, which was always stretched initially the same amount. Thus, it was possible to aim the ball toward the contact trigger on the wall and to sequence the collisions a quarter of a millisecond apart (0.00025 s). Experimental development work was carried out by Sean McGrew and Eric Deren, who also took several series of photographs.