Blended Multiple Exposure

While this is definitely a post processing technique, I think it works very well in sports photography.

The technique I use is pretty simple if somewhat labor intensive. When shooting, a tripod is ideal, but you can handhold. You want to capture the ball (or person/object that is going to move) from beginning to end. This usually means a wider shot that normal. Then, motor drive through the action, keeping the camera still. This is a great technique to try when you are at a blowout game, or just find yourself shooting the same old thing.

Back at the computer choose a first and last photo, and stack them as layers. Blend these being careful to choose beginning and ending players as well as ball positions, with soft transitions in places where the photos are identical. Now add each frame between these two one at a time, selecting the ball with some room around it, soften edges, invert delete. (removing everything but the ball and a little of the environment around it.) If the position is a little off and the background does not match, use the arrow keys to place it in the right position.


REX14047
by RexPhoto91, on Flickr


Swoosh
by RexPhoto91, on Flickr
 
In the cliff jump photograph, I can estimate the frames per second shot by the camera based on a few assumptions:

[list type=decimal]
[*]The man is of average height, 1.77 meters.
[*]There is no significant vertical component to the initial velocity of the jump.
[*]The first frame is taken at precisely the moment of the jump.
[/list]

Then we estimate the total height of the jump by measuring (using Photoshop) the vertical distance in pixels from the top of the man's head in the first frame to the corresponding point in the last frame just prior to the splash, and divide this by the height of the man in pixels. This gives me roughly a ratio of 5.04, and under assumption 1, this means the vertical distance of the jump to the water is about D = 8.9 meters.

Using the formula D = (1/2) g t^2, where D is displacement, g = 9.8 m/s^2 is the acceleration due to gravity, and t is time, we find that the jump took a total time of approximately 1.35 seconds. This is probably a bit longer than the actual time, which I think should be closer to 1 second.

There are about 12 frames that were shot in the composite. This gives a frame rate of 12/1.35 = 8.89 frames per second for the camera that shot these images.

More interestingly, given the frame rate of the camera, I can estimate the approximate height of the jump, and therefore, the height of the man. I could even give error bounds for these heights.
 
Upvote 0