Digital Stroboscopic Motion photography
Andrew Davidhazy
Imaging and Photographic Technology
School of Photographic Arts and Sciences
Rochester Institute of Technology
Stroboscopic photography for depicting the changing features of
subjects in motion is a technique that was enabled and popularized by
one of the "fathers" of high speed photography, Harold "Doc" Edgerton.
He applied this technique to numerous situations where a still camera,
in a single shot, did not capture enough information about a subject in
motion and where a motion picture camera record was unsuitable for
being reproduced on the printed page.Andrew Davidhazy
Imaging and Photographic Technology
School of Photographic Arts and Sciences
Rochester Institute of Technology
Photographers use two types of stroboscopes. The first, and by far the most popular, is nothing more than a flashing light source operated in a dark environment. This is the kind that Doc Edgerton popularized. There is also a cheaper, simpler, mechanical alternative. This consists simply of a rotating disk with a slot cut into it. With every rotation of the disc we are able to glimpse the position of the subject at the time the slot passes in front of our eyes. In either case, a stroboscope essentially allows us to view a subject on a periodic basis. This is called time-sampled image recording.
Traditional Stroboscopic Technique
Photographers have generally used a flashing light, a stroboscopic light source, to illuminate a moving subject in order to track the subject position over time. This is accomplished by setting up an action situation if front of the camera, generally firmly attached to a tripod, opening the shutter of the camera while the subject is moving and the stroboscope is flashing and after a while closing the shutter and terminating the exposure.
During the time the shutter is open the moving subject is illuminated by several flashes of light. These leave a superimposed sequence of images of the subject and it is often possible to gain much valuable subject motion evaluation information from such a record or to simply connect with and understand the graceful (or not so graceful) flow of motion of a given subject.
When engaged in stroboscopic photography generally one is interested in photographing relatively long duration events (even though one may only be talking about a second or two!) because it is really pointless to make a stroboscopic record while only recording the subject in only one, two or three positions over time!
For example, to make a record of a golf swing or something similar, we would be looking to make a record over a time period of a second or so. During that time we might want to record our subject in maybe 20 to 100 different positions. This, of course would require a strobe flashing at a frequency of 20 to 100 flashes per second if we kept the shutter open for a second. An exposure time of 1 second is easy to accomplish with a regular camera but many digital cameras have a limited maximum exposure time.
The opaque disk with slot rotating in front of a camera’s lens is a much simpler and cheaper alternative to the flashing light stroboscopes and it is with just such a device that the photograph shown here was made using an exposure time of 1 second. The subject was placed against a large, black, velvet background. The lighting level was adjusted so that the results obtained were of an acceptable quality by making a few preliminary tests and judging the quality of the images on the LCD display screen of the camera. This is very convenient and effective and made possible by digital image capture technology.
Figure 1. Digital camera with rotating disk stroboscope attached.
Figure 2. Multiple exposure stroboscopic photograph made with rotating disk stroboscope.
From this example made over a period of 2 seconds, the camera recorded about 30 separate images of the action. From this it can be determined that the mechanical stroboscope's disc was turning at about 15 revolutions per second. The exposure time (per image) that the disk delivers is simply a function of the slot size divided by 360 degrees and multiplied by the time it takes the disk to turn once.
Note that, as with traditional cameras, the background (which should have reproduced very dark since it was black velvet) and those parts of the subject’s body that remained essentially in the same position appear significantly overexposed because they reflected light to the same location on the camera’s CCD image sensor. On the other hand the moving racquet was exposed in different positions on the CCD with every pass of the stroboscope disc. The moving racquet, therefore, is exposed only once on any given area on the CCD and reproduces with less exposure than the stationary parts of the scene. One could improve slightly on the tonal range of the image by making the subject wear dark clothing while painting the moving racquet with highly reflective or white paint.
Extended Time Digital Stroboscopic Photography
If one leaves the shutter for too long the sequential images will start to overlap until a point is reached that they all blend into a uniform "blur" and the specific position of the subject with any given flash discharge can no longer be perceived. If one is interested in motion detail then, in a film-type camera, it is possible to simply put the film in motion while the stroboscopic light flashes or the mechanical disk rapidly rotates in front of the camera's lens and the subject performs in front of the camera.
Using a standard 35mm camera the film is simply advanced one frame at a time, with the lens covered, into the take-up chamber without exposing it. Then, setting the camera's shutter to "B" and locking it open during the rewinding process causes the moving film to record the subject's image to ultimately appear displaced "time wise" in the final record. It appears as a sequential set of images showing the progression of motion of the subject over time.
In the digital realm generally there is no possibility for moving the image receptor material as there is in the film realm. The sensor in the digital camera is fixed in place and so while standard, superimposed stroboscopic records can be obtained much the same as with film cameras (as shown above) the visualization of motion of a subject over time by introducing motion of the sensor is not possible.
However, there is a method to deal with this problem if one is willing to give up a certain amount of resolution in the digital records captured by the camera. The approach is to put the image in motion in such a manner that it traverses the field of view of the camera during the time that the shutter remains open.
This can be accomplished several ways. One practical way to do solve the problem is to realize that the image can be moved across the focal plane by rotating the camera. The subject can remain in approximately the same general location performing some action and the camera is aimed in such a manner that their image appears at one side of the camera's field of view. Then, after the shutter has been opened, the camera is slowly turned in such a manner that the subject's image moves to the other side of the viewfinder before the exposure is terminated.
Clearly then with each flash of light from the stroboscope the subject's image gets recorded on a different location of the camera's sensor and the motion of the subject can be relatively easily tracked over a significant time period.
A major drawback of this technique is that the field of view of the camera (over time) is now very wide and the flash, stands and studio equipment may reflect some of the light from the stroboscope along with the light from the subject and these times will appear to "bleed through" the subject's image.
To deal with this difficulty the camera is surrounded by a curtain of black velvet set up in such a manner that there is a small open slot located between the camera, rotating on a tripod, and the subject. This prevents the camera from seeing and recording anything but that which is available to it through the gap or slot in the curtain. This provides a "ghost" free final record although superimposition of certain parts of the moving subject is still a possibility. This often just adds to the fluidity of the final motion record.
Figure 3. Studio set-up for digital time displaced stroboscopic record photography.
Figure 4. Result of panning a digital camera to capture action as shown in Fig. 4
Since one is limited in the size of the camera's sensor the total number of images that one can clearly record depends on how small one is prepared to allow the subject to be within the frame of the camera. The smaller the image size one can tolerate the larger the time over which one can "track" the subject or the number separate images of the subject that can be placed from one side of the frame to the other.
Digital Image Comparison/Storage
A further alternative to the stroboscopic light source is a technique only available in the digital realm and this is a non-integrating, periodically updated, image output device driven by a video signal input from a video source such as a camera. A typical example of this technology is the Colorado Video Frame Store device. It is an interface between a video camera and a video monitor that compares the signal level information associated with each point in a video frame with that of the next one. The device can be set so that an increase (or a decrease) over some base signal level refreshes the display at that point of the image being displayed. If the levels do not change in the desired direction then a steady and unchanging image is displayed on the monitor. The frequency of update of the display can be selected from the maximum capture rate of the camera to longer times.
Since a video camera is a time sampling device operating at 60 frames per second the instantaneous records of a moving subject across a dark background can be displayed at 1/60th second intervals and the composite display looks much like it would look if the record had been made with a standard camera. The big difference is that in this case the exposure or level of subject areas that do not move would not introduce unwanted “noise” (due to overexposure) in the final image.
Fig. 5. Digital stroboscopic photograph based on comparison sampling technique.
Digital Stroboscopic Assembly
Finally a stroboscopic image can be assembled after-the-fact using digital techniques that in the past would have required much too elaborate and time-consuming photographic process. The basis for a stroboscopic composite is that one starts with a digital video record of an action captured at some given framing rate and with the camera fixed in place. First a reference image of the scene is made while no action takes place in front of the camera. Then the action is recorded against that background.
Then, using digital technology one isolates from each frame those areas where the image information is different to the reference, static, view of the background. Finally the isolated images of the action are layered one on top of the other in their proper spatial and temporal locations in relationship to the original, or new, background view. The final display is a time-sampled record of the position and location of a subject against a fixed background at several instances in time.
Figure 1. Digital camera with rotating disk stroboscope attached.
Figure 2. Multiple exposure stroboscopic photograph made with rotating disk stroboscope.
Figure 3. Studio set-up for digital time displaced stroboscopic record photography.
Figure 4. Result of panning a digital camera to capture action s shown in Fig. 4
Figure. 5. Digital stroboscopic photograph based on comparison sampling technique.
Details on the making of a rotating disc or mechanical stroboscope and the making of digital time displaced records with a standard DSLR can be found elsewhere on the articles page.
For questions or suggestions or corrections of this material contact me at andpph@davidhazy.org