November 11, 1951…First Video Image From Magnetic Tape
Believe it or not, Bing Crosby Enterprises engineers Jack Mullen and Wayne Johnson were the first to produce an image from video tape.
What follows is a first hand account from BCE engineer Robert R. Phillips. Below that is a link to a much larger article on Crosby’s work in tape of all kinds. Enjoy and share! -Bobby Ellerbee
While there were many interesting experiences working at Bing Crosby Enterprises, none was as challenging as the development of a video recorder for Bing. When Jack told Bing in 1950 that it could be done, Bing responded by setting up the video recorder project to be led by Jack in the Electronic Division.
Mullin reached out and recruited Wayne Johnson to assist him, and by November 11, 1951 they demonstrated the first video picture from magnetic tape. It was a video of an airplane taking off. The image was very snowy and the airplane was a dark blob, but with the commentary by Frank Healey everyone managed to see the takeoff. Frank was an ex-movie producer.
I started with BCE about a week later later and found myself involved in the construction of what I call the Mark II version. The Mark I system used the top plate from one of the portable Ampex 200 audio tape recorders modified to operate at 360 ips (20 MPH) with a modified head that gave it a bandwidth of about 1 MHz. The tape was quarter-inch, and the reels did not have any sides to reduce once-a-round effects. This lack of sides produced large piles of tape if something went wrong. At times I had to go out the front door and down Sunset Blvd. with the tape to get it back on the reel. The drive was an early version of the tight loop.
The Mark II version used the same top plate as that of the Mark I, but it was modified for one inch tape. It still operated at 360 ips with the early version of the tight loop drive. Jack Mullin and Wayne Johnson had decided to build the Mark II system before the November 11 demonstration and were actively engaged in its construction. When I started at BCE they were testing various head stacks and working with William Wetzel at 3M to improve the tape quality for the recorder. It was obvious that main problem was how to increase the recorded bandwidth on the tape. We would find out over the next months that there was a lot that was not known about how the head-tape process works.
A discussion over using a rotary head approach to that of longitudinal had taken place, and there was much doubt that the rotary head would work. It also would have been difficult for the BCE group to implement it because of our lack of the mechanical facilities to build one. The longitudinal approach was selected with multiple tracks. Instead of using a rotary head, the multiple tracks were electronically scanned. The Mark II system had 12 tracks recorded on the 1 inch tape. Ten of the tracks were used for the video, and the other two, one on the top of the tape and one on the bottom, were used for audio, sync and reference signals.
The video signal in the record subsystem was sampled with the rate tied to the video sync signal. These samples or pulses were sequentially recorded across the 10 video tracks. For each set of 10 samples the polarity of the recorded pulse was reversed. This was done to prevent a bias on the tape. The result was a series of pulses recorded on each track on the tape alternating in polarity. The amplitude of the recorded pulse was equal to the amplitude of the video sample, and each channel had a 1 MHz bandwidth.
The playback system had to reverse the polarity of every other pulse and assemble them into a video stream of pulses combined with the video sync signal. The reference signals from the top and bottom tape channels were used to correct for tape speed and skew. This information was used to adjust the sampled video pulses from the tape. The video picture on the monitor consisted of a number analog “pixels” that had to be processed to eliminate the dot structure by averaging between them.
About a month after I started, Dean DeMoss and Chester Shaw were hired to help build and test the new electronics for the Mark II system. Gene Brown provided the mechanical support, and the system was built and operational by early 1952. Wayne Johnson and I were working on the system Friday night, 14 March 1952, and had completed the last adjustment to it. It was ready for its first recording test, and we decided to try it before we went home.
A video program was recorded and played back; it could have been classified as a fair picture. However, it did have the “pixels” since the additional processing had not been added. We jumped up and down and cheered. It turns out that we were the only two to see the picture because Wayne came in the next day and dismantled the recorder to “improve” it. Jack Mullin found out on Monday and was not happy that he missed the first successful test recording session.
During the 1952 and 1953 period the major effort was concentrated on improving the frequency response and reducing the tape speed. Many heads designs were tested using different types of magnetic materials, gap sizes and materials, and head windings. To make the heads, a precision optical lapping capability was installed along with a way of winding the coils on the assembled head segments. After trying many different head configurations, it was realized that there was a limitation in recording high frequencies on magnetic tape. A point was reached where changes in the head configuration and the tape speed made little difference in the upper frequency that could be recorded.
There were a number of other groups that were working on the video recorder. One of these was RCA, and its chairman General David Sarnoff wanted one for his birthday. By 1953 they had demonstrated a system that ran at 360 inches per second like the first BCE machine in 1951. It had better quality using video compression, but only lasted 4 minutes per reel.
The same year Sarnoff of RCA and his board of directors visited BCE to see if they could buy our recorder. The party arrived in a number of black limousines. Sarnoff was in the first dressed in gray and the board in the rest, dressed in black. As Sarnoff marched up the driveway the board fell in two-by-two behind him. They went into our small laboratory for the demonstration. Sarnoff sat in the middle and the board on either side of him.
After the demonstration they went upstairs to the conference room. Sarnoff told Frank Healey that he wanted to buy the recorder, and Frank told him it was not for sale. Sarnoff told Frank that he would just buy them out; to which Frank replied “You do not have enough money!” The RCA group left.
During this period the tight loop drive was perfected, and the correction of wow and flutter effects addressed. Much of this work also was used to improve the recording of the instrumentation telemetry. In the summer of 1953, I met Eugene Sakasegawa who was chief engineer of the USC television station. We were still looking for people who were good engineers with video experience, and I encouraged him to join us. He was hired, and his first assignment was to learn how to make magnetic heads. Gene was a craftsman and took up the challenge. After a couple of months of building new types of heads and not making any progress, Gene produced a head that broke the 1 MHz frequency barrier. Jack and Wayne could not believe it.
When asked what he did to make the head, he said that he made some small cuts inside the head to make it look better. He had been lapping the head halves to create a uniform gap and also the head face. These actions did make a small difference in the performance, but the “beauty treatment” was the key. The head laminations at the head gap were different widths and caused a non-uniform magnetic pattern across the tape (top to bottom). Gene had made a very accurate cut behind the head gap across the laminations. This cut made them the same thickness and caused the magnetic pattern across the tape to be uniform.
A number of different cuts were tested until proper configuration was found. The result of the head “beauty treatment” was to reduced the tape speed of the Mark II to 120 ips and produce a new design for the recording the video information. The new “Mark III” recorder operated at 100 ips with half-inch tape using longitudinal recording and no scanning. It recorded color video and sound with three heads – video, color, and sound/reference. The recorder also employed video compression techniques, and an early version was demonstrated in February 1955.
The recorder was further refined and demonstrated in June 1955. Based on the performance of this new recorder CBS ordered three of them, and the summer of that year we worked building them. In late 1955 Bing asked Jack Mullin to visit Ampex to see their video tape recorder. We knew that they were working on one, but did not know how far they had come.
Jack went to Ampex near the end of the year and came back with the news that it was over, Ampex had a better recorder. Bing sent another $50,000 check to Ampex for the first of their video recorders. It was delivered to his TV station in Washington State. Ampex demonstrated their rotary head machine on 14 April 1956; CBS cancelled their order with BCE and bought the Ampex recorder. The Mark III recorders were then converted to wide band instrumentation recorders and sold.
For MUCH MORE, follow this link.