High Speed Photography
No post production effects can duplicate the magic of film shot at high speed.
In 35 mm we are still limited to about 360 FPS for the film stopping for each
exposure. The rotary prism cameras go to 2500 FPS. In 16 mm, pin registered
cameras go to 500 and rotary prism to 10,000. FPS but the image is not very
good. The problem with rotary prism camera is that only a smaller image in the
middle of the frame is sharp, the maximum shutter angle is 73 degrees and registration
We needed a shot of a tennis ball hit by a tennis racket at 2500 FPS. The odds
of a player hitting a ball exactly in the right frame and place on the racket
were not very high. We built a spinning racket rig to increase our odds of hitting
the ball right.
At 2500 FPS with a 73-degree shutter you need a lot of light, about 8 stops
more exposure than at 24 FPS with a 180-degree shutter, not including reciprocity
failure. Sunlight is a great source. If shooting a smaller object plan to move
everything with the sun's movement. If you need more than sun power, you can
build a set of mirrors on 1" x 4' x 8' plywood that focus the light of 3 ea.
47" x 31" mirrors. Each single strength mirror will bend and focus light toward
a 2 by 2-foot area at the foci of the mirrors. The mirrors can be tuned with
pads glued to the back of the mirrors and adjusted with studs through the 1"
We used sunlight as fill and our concentrating mirror panel as 3/4 back rim
key. This rig was heavy and not easy to move and focus. Today's faster films
will help, pushing is acceptable because the image is not that sharp anyway.
We used this rig to shoot tennis balls hitting the "sweet" spot of two tennis
rackets spun at serving speed of 60 MPH. We dropped 30 tennis balls into the
spinning rackets and got one or two direct hits per each take of 500-foot film
take. Few of the balls hit the rackets at the right place and balls flew all
over the set. Each ball was dropped from a chute with the label facing camera.
A solenoid metered the balls one at a time out of a plastic tube. The rackets
were attached to a bearing mandrel with a 3/4 HP motor turning the whole thing
to give a racket speed of 50 MPH.
To split arrows we built a micrometer adjustable bow that could be re-adjusted
each take. Care was taken to load each arrow exactly the same. The notch of
the receiving arrows was removed and the back opened some to help the arrow
hit the hole. We got about one hit out of four. The hits pealed the arrow back
like a banana in all directions. Our actor had split arrows before, but we couldn't
afford to shoot hundreds of takes to repeat his feat. We lit the shot with two
separate 10 K Vegas with hot lenses to light the arrow, feathers, arrow head
coming in and back ground to show.
To save print cost at the lab we always asked them to "print action only (described
what the action is) plus and minus 10 feet".
16-mm tests can help find problems.
There are a few assistants who know high-speed cameras, but most of the time
it is wise to get an assistant very well acquainted with these cameras.
Focusing the rotary prism camera is not easy. Using a sun burst (Siemen's Star)
pattern focus chart will help.
The 35-mm cameras are very heavy and removable dovetails, heavy supports and
heads are helpful.
Always consider double printing or video finish to get a higher apparent speed.
Try to avoid the rotary prism camera if the 360-FPS camera will do the job,
its image is so much better.
Be sure that you have the right 3-phase power, heavy cable and enough power
to run these cameras. Not all 3-phase generators will run them.
A parallax video camera can be a little aid.
Everything that can be done to make each shot repeatable and controlled is worth
the effort. To make cereal flow smoothly out of a box and not clump up we built
a small motor driven conveyor belt to deliver the flakes smoothly out of the
box. The belt was a strip of film driven by a DC gear motor with film sprockets.
The cereal box covered the conveyor. The box and conveyor well supported.
When shooting you can save film by timing "action" not by the assistant calling
"speed", but by the person triggering the event when he hears the camera hits
speed by sound. The assistant can then shut the camera off and apply a brake
it if possible to stop it faster.
© Copyright 1999-2004 Ron Dexter. All Rights Reserved.