These cameras are perhaps the easiest to describe and have historically been the highest-performing designs, with some tradeoffs.
There are three elements to pretty much every camera:
- A means of forming an optical image on a surface (usually a lens)
- A means of recording that image (nowadays mostly digital image sensors such as a CCD or CMOS sensor)
- A means of controlling when the image is taken (usually a shutter, either on the sensor or a separate unit)
A sequentially shuttered high-speed camera simply takes this concept further. It groups together multiple camera units, all pointed at the same direction/object, and fires the shutters one after another very quickly. The shutters of these cameras usually end up being very complicated as to increase the framing rate of the camera, one must increase the speed of the shutter (the off-on-off time). These cameras have speeds ranging from tens of thousands to up to 100 billion frames per second in modern versions that use image intensifiers to shutter the image.
There are also varying arrangements of these camera units. Most modern designs (See Specialised Imaging's SIMX Cameras, Stanford Computer Optic's XXRapidFrame, or Cordin's Model 214/222) use a single entrance objective arrangement, which allows for considerable versatility (interchangeable lenses). A beamsplitter or mirror arrangment is used to feed the light from the single entrance lens into the multiple camera units.
In contrast, some earlier cameras such as those used by the British for their atom bomb tests were simply an array of cameras pointed in the same direction. These minimized parallax as atomic bomb tests generally occur quite far away from the camera and long focal length lenses could be used.
While these cameras seem very advantageous given their high temporal resolution, there are numerous disadvantages that limit their use. Since a separate camera unit is necessary for each frame that is taken, any frame count above 16-ish is economically infeasible. One should keep in mind that 1 second of normal motion picture footage takes up 24 frames, so these cameras are rarely used for video analysis.
The beamsplitting arrangement that is commonly used is also not advantageous. The light from the entrance lens must be divided between 8 or more channels, which means the amount of light getting to each individual channel is 1/n the light the entrance lens "puts out", n being the frame count. With modern image intensifiers this is not much of an issue as their gain can be upwards of 10,000, but for the most demanding low-light applications this can still be limiting. Additionally, the mirror pupil beamsplitting arrangement (or separate camera) results in slight parallax error which is highly undesireable for photogrammetric applications.
Shuttering methods
There are a few different ways to shutter these types of cameras, which will be discussed further in other sections. The simplest type is simply a disc with radial slots cut in it that rotates in front of a slit that covers the lens. This is very simple but the exposure time is usually more than a few hundred microseconds, which limits the framing rate. A commonly used shutter for early cameras was the Kerr cell, but it is expensive and difficult to use. Modern cameras use image intensifier units which are expensive but allow for extremely fast shuttering action.
More information and images to be added later!
Contact me at [my first name]h@utexas.edu