The camera’s job
A television camera has a harder job than a film camera. Film and other photographic cameras capture a whole image at once on a chemical emulsion. A television camera must convert light into an electrical signal and then send the result out in a continuous stream, one line at a time. The device that does the conversion is the heart of the camera, and its history is a story of ever more sensitive and stable ways of turning an image into electricity.
Mechanical scanning
The earliest cameras used a mechanical scanner such as a Nipkow disc, a rotating disc with a spiral of holes. As the disc turned, each hole swept across a strip of the scene, and a photocell behind the disc measured the light. The signal was a series of brightness values, one for each point in the scan. Because little light got through the holes, the subject had to be brightly illuminated, and some mechanical systems used a flying-spot arrangement, in which a bright moving spot of light scanned the subject instead. The pictures were low in detail, with tens of lines, and were a dead end for broadcasting beyond the 1930s.
The first electronic tubes
The breakthrough was a family of camera tubes that could scan electronically. Two names dominate the accounts, and credit is argued over. Philo Farnsworth in the United States demonstrated an image dissector in the late 1920s, a tube that focused an electron image on a small aperture and swept it across. It worked, but it was not sensitive, because it captured light only from the point being read at a given instant.
Vladimir Zworykin, working at Westinghouse and then RCA, developed the iconoscope, which was patented in the 1920s and became practical in the early 1930s. It used a mosaic of tiny photosensitive elements that stored charge as light fell on them, and an electron beam scanned the mosaic to read the stored charge. Storage made it much more sensitive than a dissector, because each element accumulated light between scans. In Britain, EMI developed the Emitron camera, based on similar storage principles, and it was used from the BBC’s regular service in 1936. Disputes between the inventors, and between their companies, led to years of patent litigation.
The iconoscope still needed strong light and sometimes produced a shading effect, an uneven background across the picture, that operators had to correct manually.
The image orthicon
During and after the Second World War, a more sensitive tube appeared: the image orthicon, developed at RCA in the 1940s from earlier work. It combined a photocathode that released electrons where light fell, a target that stored the resulting charge pattern, and an electron multiplier that amplified the signal. The design was sensitive enough to work in ordinary studio lighting and, with care, outdoors.
The image orthicon dominated broadcast cameras for roughly the next decade or two, in black and white and later in early color cameras that used several tubes, one for each primary color. Its quirks are well known among those who used them. It could show a halo around very bright objects, stationary images could leave a ghost, and the tube was fragile and expensive.
Vidicon and Plumbicon types
In the 1950s smaller tubes using photoconductive targets arrived. The vidicon was simple and cheap and became widespread in industrial and early portable cameras, though it responded slowly to changing light, causing smearing on moving objects. Later variants improved this. The Plumbicon, introduced by Philips in the 1960s, had low lag and good color performance and became a standard in studio color cameras for many years. Other tubes, such as the Saticon, followed in the 1970s.
Tube cameras were large by modern standards. Studio models were mounted on heavy pedestals, and outside broadcasts relied on large vans and lots of cable. Tubes needed regular alignment, since beams drifted, and operators spent time before a show lining up and balancing the three color channels. Bright lights could damage the targets, a problem sometimes called burn-in.
The CCD
The charge-coupled device was invented at Bell Labs in 1969 by Willard Boyle and George Smith. It was originally proposed as a type of memory, but its ability to shift packets of electric charge across a silicon chip made it a natural image sensor. A grid of light-sensitive cells collects charge from the light that falls on them, and the charge is shifted out and measured to build the picture.
Early CCDs had few pixels and noisy output, but they improved through the 1970s and 1980s. Broadcast CCD cameras began to appear in the early 1980s, and by the 1990s they had replaced tubes in most professional use. The reasons were practical:
- no beam to align, so the picture stayed stable;
- smaller, lighter and more rugged bodies;
- low power use;
- no burn-in or lag from the tube; and
- a fixed, precise geometry.
Boyle and Smith received a share of the Nobel Prize in Physics in 2009 for the invention, according to the Nobel announcement. Later, another sensor type, the CMOS sensor, became widespread and has largely taken over in many cameras because it is cheaper to make and lower in power. The line between television cameras and still cameras blurred, as the same kind of sensor that powers digital photography is found in studio cameras.
From analog to digital output
The move to solid-state sensors also helped the shift in the signal itself. A CCD or CMOS sensor produces data that can be converted straight to digital form, so cameras can output high-definition digital signals without the processing of an analog tube. Modern broadcast cameras shoot HD and beyond, and the same technology that created 4K cameras sits at the end of this lineage. The mechanical disc turned a scene into a few dozen lines; the sensor in a current camera turns it into millions of pixels, many times per second.