1956 Ampex Quadraplex using 2" tape reels: The first method of video recording that proved to be commercially feasible.
Others had demonstrated TV recording, as early as 1951 by John Mullin (the same person that introduced the audio tape recorder to AMPEX, you see AMPEX did not invent either audio or video recording). RCA's method was to simply move ordinary 1/4" stereo audio tape [using one track for monochrome video] at a horrendously fast speed. Not very inventive or practical, a thirty minute TV program would consume 10 MILES of recording tape. On the plus side, it was the first method that could record color, using 4 tracks (audio plus RGB video components) on 1/2" audio tape and was so demonstrated in 1953. TV needs 2.4 MHz of bandwidth to achieve acceptable results. Audio tape recording at 15ips could reach upwards of 100 KHz, therefore boosting speed 24x to 30 feet per second did the trick.
A more novel and promising method using stationary head technology was under development at DuMont Labs. No stranger to practical needs, the Allen B. DuMont Labs was the first to produce a picture tube that did not burn up in 25 hours of use, which paved the way for commercial television to become a reality in the first place. What DuMont did was, instead of one track for the entire video signal, split the bandwidth into multiple tracks, ie using 24 tracks on 2" tape. Imagine though that this plan was not easily adapted to color. And when Paramount trashed the DuMont TV network in 1956, the system lost its only user.
Quadraplex spun 4 rotating heads so fast that they were tracking essentially transverse to the tape direction., with each swipe across the tape recording only a few TV scan lines. In order to spin that fast, the headweel was floated on air bearings. The first system (Ampex VR1000) was monochrome video, monaural audio and moved at 15ips. In 1956 this was no problem as color program distribution was still in its infancy, audio was mono, and the standard program increment was 15 minutes (1200 feet of tape). CBS jumped at the chance to get serial number 1.
A truly durable system though must adapt with the times, and Quad was a survivor for 40 years. There never was a standard for stereo audio on quad tape, because when TV finally shifted to stereo in the mid 80s, use of quad was in great decline nationally. But KVCR in San Bernardino did adapt one machine with a special replacement sound head that split the linear sound track in two. By recording the live PBS DATE (Digital audio encoded in the video vertical blanking interval) network feed on this machine, there would be stereo audio available on replay, but only for simulcast on KVCR-FM as KVCR TV itself has never made the conversion to stereo television audio.
Another carry over from audio tape recording was razor edits. The VTR was a huge mechanical contraption that was slow to get started and synchronized but it could stop on a dime. Part of the synchronizing method required recording pulses on a linear track to mark each video frame. (This is called control track on later systems and facilitates accurate lockup of tape head position and speed) By placing a magnetic dye on the tape, these edit pulses could be seen at 1/2 inch intervals (15ips) showing exactly where to cut an even frame. If you were skillful and lucky, one could even cut and paste a decent looking program together this way, although not very practical. And no one would ever want to re-use a tape that had splices in it.
Being slow to start meant that it had to be up and rolling for about 10-15 seconds before its picture could be used for air. To have it ready at the right second meant pre-rolling the machine 15 seconds before it was actually needed. As an aid to cuing the tape, pre-recorded programs would include a countdown timer (picture of a clock second hand, or in later years electronically generated numbers) so the VTR could be stopped by the operator the required pre-roll position. Actually this was not a new concept, as film had been used from the very beginning of the TV broadcast era and it too required a short startup run to get to speed. Everyone has seen the SMPTE or Academy film leader with second hand and numbers flashing.
This all is great when using a pre-made program, but what about starting a clip in mid-reel? No big problem with film, just look for the first frame that's needed and count back or measure off the equivalent distance (3 feet is 120 frames of 16mm film). With quad there is a viewable picture only when the tape is upto speed. Finding the right "frame" could take several viewing passes. If stopped at the right mark, measuring back 15 seconds means hand pulling out 18.75 feet of tape and then carefully manually winding back onto the reel. A better way was found. A 1.25" diameter roller along the tape path would make a complete turn every 4 frames (15 ips), so AMPEX geared this to a numbered turns counter that was calibrated in units of time (frames-seconds-minutes). So reset the counter at the "in" point and rewind tape until it read the required seconds without ever unthreading it. The counter also made it easier for operators to quickly get exact total program timing on pre-recorded tapes.
This turns counter made cuing tape so "easy" that it enabled electronic editing to be possible. This is the so-called linear method of playback deck to recorder deck editing. With pre-roll timing perfomed by the counter, one merely had to start both decks simultaneously and push the "edit" button at the appropriate time. Clumsy as this sounds, it was a giant leap in quality compared to razor edits, and original tapes were not sacrificed by this method.
But what really kept quad viable for so many years was it worked also for color, but not in the way color recording was originally envisioned. In an old monochrome studio, the cameras and other sources supplied video only, no sync pulses. The sync information was added as a last step prior to sending out for broadcast, which insured that critical sync quality could be controlled at all times. It took a lot of tubes and calibration time to generate a legal sync signal, so doing this for every camera and device in the studio was not wise. The exception was the relayed network signal would simply pass the master control switcher uneffected on the (good) assumption that network signals were always perfect. Originally color studio design called for the signals to be kept separate and switched as 3 monochrome (RGB) video until sync was added and at that point the 3.58 MHz color signal was also generated to make NTSC composite. The studio VTR was expected to be component also. The master control switch still just passed network, and often was the old B/W unit that just happened to have sufficient bandwidth to handle full NTCS composite. The term "passes color" was a popular sales gimmick for disposing of old B/W design equipment stocks during the transition years.
KVCR became a color television station by accident. Its "network" was KCET TV28 which was picked up off air and retransmitted on channel 24. When the otherwise B/W KVCR received a color program from KCET, it got retransmitted in color also. The antique monochrome transmitter at KVCR actually passed color (technically illegal - but who cared?).
But by now transistors were replacing tubes in studio equipment, and 2nd generation design (Ampex VR1200) VTRs were also transistor. The idea of 3 component color did not work well at all in actual practice. For example, a difference of only a few inches in wire lengths between channels was sufficient to corrupt the color at the NTSC composite generator. The superior method was to create the composite signal at each originating device, with sync, so that downstream switches could pass the signal unchanged. Now with solid state circuits and especially using the RTL miniture digital logic chips that had just come to market, this was feasible as well. It was discovered that when fed a full color composite signal to record, the VR1000 could actually play it back, it just needed to have a color sync circuit (called ColorTec by Ampex) added. A competitor, RCA, helped Ampex out because its TV network (NBC) needed color videotape. As a thanks, Ampex licensed RCA to manufacture its own versions of the color quad machine.
Everyone of course now wanted the color capable VR1200 model, so Ampex donated an unsalable VR1000 to KVCR when the station first started operation in 1962. The last Quad VTR at KVCR was removed from service in 1995. In case you were thinking of perhaps restoring one of these for display in your house, you might want to consider its size (washing machine with companion dryer class) that it will require a dedicated 220 volt power outlet, plus a large air compressor!! The first station in Casper Wyoming, KTWO, started in 1957 and they also had one VR1000 which later became a "parts bin" when replaced by two VR1200s. That is where I was mentored in broadcast engineering by Tom Norman, and they were still using quad when I left in 1986. The air compressor makes so much noise that it was placed in an outside shed. Moisture in the air lines was always a problem and in Wyoming , as Tom recalls, the line from the shed would freeze up occasionally in the wintertime. Tom finally fixed that problem about a year before the machines were taken out of service by running the line into the ventillation duct that came from the nuclear bomb shelter that was under the station. Yes Conelrad stations built in the 50's were hardened against nuclear attack, remember the 2 little triangles that used to be on every AM radio dial? While the quad is gone, most stations have kept the air compressor because it is so handy to have air available for cleaning dust out of "old" electronic equipment.
next on the history of videotape, 1962 Helical Scan using 2" tape reels