Minding the Store
Video storage is one of the least considered aspects of videography, but that’s now changing.
Here’s a trick question: Which came first, the video recorder or video? The answer is: It depends.
The subject of last month’s New York Section meeting of the Society of Motion Picture and Television Engineers (SMPTE) was “Tapeless Camcorder Technologies.” It featured representatives from Ikegami and JVC (who propose different forms of magnetic-disk recording), Hitachi and Sony (who propose different forms of optical-disk recording), and Panasonic (proposing solid-state memory). All say their systems are better than videotape.
Most people familiar with the technology of videography know that the first videotape recorder appeared in 1956. Those who have gone deeper into our background know that the first magnetic-disk-based video recorder was used by CBS for instant replay in 1965. Those familiar with the consumer end of videography might cite the LaserDisc, first sold in 1978, as the original optical videodisc system. As for a solid-state image recorder, some might give the first moving-image credit to Panasonic’s recent SD (Secure Digital) memory camcorders and the first still-picture credit to NEC’s 1973 frame synchronizer. And all would be wrong.
Consider solid-state video recording. Panasonic’s SD-based consumer camcorders impressively record motion sequences onto tiny memory cards, but chip-based moving-image video recorders date back at least to the 1980s, when quite a few companies introduced larger (non-mobile) professional versions. Similarly, NEC’s frame synchronizer was a breakthrough in the technology of videography, but it wasn’t exactly the first no-moving-parts video memory.
UNESCO, the United Nations Educational, Scientific, and Cultural Organization, has a Memory of the World archive project to preserve significant artifacts. One item is Kalman Tihanyi’s 1926 patent application for a “Radioskop.”
The Hungarian scientist first presented his breakthrough idea in a 1925 paper, “About the Electrical Television.” Tihanyi was neither the first person to conceive of videography nor even the first to propose an all-electronic version.
Americans sometimes credit Philo Farnsworth as the inventor of television — and he may very well have been the first to demonstrate an all-electronic system. But there was a problem.
Farnsworth’s picture sensor, an image dissector tube, captured signals from only the light falling on its surface at the instant of scanning. Initially, his televised pictures came only from photographic slides and later from motion-picture film. Real-world scenes — even in broad daylight — were just too dim for the insensitive apparatus.
By late 1930, Farnsworth had increased sensitivity to the point where live images could be captured — if the person being scanned could tolerate the conditions. It was said that a 1500-watt incandescent bulb a foot from a face was just about sufficient lighting. Farnsworth recognized the problem and proposed the use of more-photosensitive surfaces.
Tihanyi had a different idea. Instead of dealing with only the light falling on the surface at the very instant of scanning, why not accumulate a charge based on all the light that fell since the last image scan? In short, he proposed storing the electronic image until it was ready to be read. The effect was like that of reducing the shutter speed of a camera by several factors of magnitude.
Image storage was incorporated into the Iconoscope camera tube, considered the first practical electronic video image sensor. Vladimir Zworykin first applied for a patent on it in 1923, but the application was modified through 1931, and several of the original claims were removed based on Tihanyi’s work.
It had not been Tihanyi’s or Zworykin’s intention that solid-state image storage in a camera tube be used to capture pictures for later viewing. Other storage mechanisms conceived around the same time, however, were intended to preserve images for longer periods.
Consider, for example, the optical videodisc. What was eventually called the LaserDisc was probably the first such disk sold, and it formed the basis for the CD and DVD that followed. But it was by no means the first optical videodisc demonstrated, let alone conceived.
In 1975, visitors to an exhibition at New York’s Library and Museum of the Performing Arts could see a videodisc made of flexible, transparent film. In 1957, Antonio Rubbiani demonstrated another optical videodisc at an international technology exhibition in Turin. And in 1929, Reginald Friebus applied for a series of patents (later granted) covering color, optical videodiscs.
Magnetic disk storage? MVR’s 1965 VDR-210CF was probably the first system to be used in a television broadcast, but it was not the first video recorder based on a magnetic disk. Siemens demonstrated a sort of video floppy disk in 1961, and Philips demonstrated video recorded magnetically along the rim of a disk at an international photography convention in 1958.
Those who have studied the history of videodiscs know that inventor John Logie Baird came up with a system for recording crude (mechanical) television images on ordinary gramophone (phonograph) disks by 1927. Television signals in those days were no different in bandwidth from sound signals. The old Phonovision recordings have lasted into the Internet Age and may be downloaded today. But that wasn’t Baird’s only video recording proposal.
He also suggested recording video onto magnetic disks at one rate and reading them off at another (faster) rate, something considered an important feature of the latest camcorders. Baird applied for a patent on the idea in 1927.
That could be the earliest patent application for video recording on magnetic disk, but it’s definitely not the first for magnetic video recording. On June 27, 1922, Boris Rtcheouloff filed an application in Russia for a videotape recorder.
Unfortunately, although the technology had been worked on at least since 1877, video did not yet exist. Baird is generally credited with capturing the first recognizable human face via video in 1925.
There was also the small problem that magnetically coated tape did not yet exist (see sidebar “Cigarette?”). And, unlike Baird’s mechanical disk recorder, which was able to create synchronized scanning lines on playback, Rtcheouloff’s patent application lacked any mechanism for sync.
So it may be said that the video recorder came before video — if an unworkable patent application is taken as evidence of a video recorder. But Rtcheouloff should still be honored if his was the first patent for a magnetic recorder. It wasn’t.
Tape, as we know it, was still in the future, but magnetic audio recorders had been sold for years before Rtcheouloff’s proposal. They used either metallic ribbon or wire. In fact, a magnetic wire recorder was demonstrated at the same 1900 Paris event where the word television was coined. And proposals for magnetic recording of electrical signals date back at least to 1887.
So magnetic tape recording has had a long run. Even commercial videotape will celebrate its 50th birthday in the spring of 2006. Tape-based camcorders have existed for more than 20 years. Is it really worth changing to something else?
In a word, yes. Today, videocassettes from camcorders — from HDCAM to MicroMV — need to be played at real-time speed into other recording mechanisms for non-linear editing, because tape is linear (see sidebar “Really Reel”). Director and former Videography regular Scott Billups was once so frustrated by that extra loading step that he strapped a computer to his body for what was probably the first human-carried direct videodisc recording.
The only reason the change from tape hasn’t happened already is the demands of field video recording. Camcorders need to be extremely rugged. Media need to be extremely reliable. Quality needs to be acceptable. And then there are other factors.
When Panasonic and RCA introduced the first camcorder format, Type M, in 1981, even though it recorded high-quality component-color signals for the professional market, it used ordinary VHS cassettes. When Sony brought out Betacam the following year, it used that company’s Betamax consumer cassettes. Why?
A news crew never knows how long a news event will last. Will the hostages be released now or in ten more hours? Will the war be over today or in two more years?
By basing their professional camcorders on readily available consumer-recording media, the companies made it possible for crews to drop into supermarkets or drugstores to pick up more cassettes — for a pittance. Unfortunately, it’s virtually impossible to find a blank Betamax cassette anywhere these days, and Type M has long since been supplanted by newer, non-consumer-media-based camcorder formats.
Still, media cost remains a factor. Long after Sony introduced metal-tape-based Betacam SP camcorders, many users still put lower-quality oxide-tape-based Betacam cassettes in them for lower cost.
Then there’s archiving. In the days before non-linear editing, the same format tape would be used for shooting, editing, and long-term storage. And videotapes dating back to the 1950s can still be played — if one can find the appropriate playback machine.
Disks and solid-state memories are great for editing, and they now appear to stand up to the rigors of field use, too. But will they last as long as tape?
Archivists often debate the advantages and disadvantages of different moving-image media. There seems to be general agreement that one such medium, also first used for video recording in the 1920s, is the most reliable for long-term storage. In fact, it has been so stable a format that even modern playback devices have no difficulty dealing with decades-old recordings.
What’s the longest-lasting video-storage medium? It’s film.
In 1956, the year the first videotape recorder was sold, an Emmy was awarded for its development. More than a quarter-century later, in 1983, another Emmy was awarded for the other product that made videotape recording possible — videotape.
Videotape was based on audiotape. Audiotape, in turn, was based on cigarettes.
It was fashionable in the 1920s to smoke “gold”-tipped cigarettes (actually ones dipped in a bronze powder). Unfortunately, the powder came off on smokers’ lips.
Fritz Pfleumer applied the same technology developed to keep cigarettes from losing their “gold” to attaching iron particles to paper. In fact Pfleumer’s first recording tape was comprised of a string of cigarette wrappers glued together and coated with a magnetic layer. Although others had applied for patents on similar ideas as early as 1915, Pfleumer’s “sounding paper” patent of 1928 was the one that led directly to the tape recorder.
See? Smoking isn’t all bad.
If tape dies out, it will be because other technologies are seen to offer benefits that tape can’t. Or can it?
One reason Hitachi, Ikegami, JVC, Panasonic, and Sony (among others) are pushing tapeless camcorders is so their recordings may be inserted directly into random-access non-linear editors. Traditional camcorder cassettes can’t hold a candle to disk or solid-state memory in that regard; it takes minutes to shuttle from one end of a tape to the other. But what if it took only a few seconds?
Is there some new tape technology that can bring this to pass? No. But there was an old tape technology that brought it to the past.
It was called longitudinal video recording (LVR) and was promoted by BASF (the first manufacturer of magnetic tape) and Toshiba in the 1970s for potential home video recorders. Longitudinal recording is what takes place in an ordinary audiocassette recorder, where the tape moves at less than two inches per second (ips).
The first successful videotape recorder, in 1956, used moving heads to effectively increase the “writing speed” of the tape. Earlier attempts at video recording moved the tape past fixed heads at speeds as high as 360 inches per second (allowing just four minutes to be recorded on a 17-inch-diameter reel).
Toshiba’s LVR would have moved its tape at more than 200 ips. To ensure sufficient capacity on a small cassette for a movie, it divided the tape width into 300 tracks. That meant the average access time to any point in a two-hour tape was about 15 seconds. With more tracks and, perhaps, an even faster non-play mode, access time could be brought so low that its lack of instantaneity would be insignificant.
It was also at least theoretically possible to use multitrack heads to duplicate tapes in one pass — 25 seconds for a two-hour tape in Toshiba’s LVR, a figure that compares very favorably to DVD burners. And LVR, unlike moving-head versions of videotape recording, allows perpendicular recording, a technique that can increase capacity dramatically — to perhaps hundreds, or even thousands of hours on a single cassette.
Unfortunately, another technology wiped out BASF’s and Toshiba’s hopes to make LVR the dominant video recording technology. Was it holographic recording? Atomic-crystal-lattice restructuring?
No, it was a much more powerful technology. It was VHS.