The Need

As we moved through into the 1970s TV studios were using three main pieces of equipment. New toys appeared along the way but Cameras and Video Recorders were used in the largest quantities, followed by Telecines. Basically Telecines are used to transfer film to TV. In the late 60’s and through the 70’s they were the only way you saw fresh news from outside the studio. Obviously they were also used for the usual televising of movies. TV cameras were used for preset events such as football and horse racing but unpredictable or remote news came in on 16mm film, in some cases after the programme had started; sometimes the film would not have even been processed. To make it a bit more difficult, the sound was on separate 16mm magnetic tape with sprocket holes. In the UK this was possibly more the case for ITN as the news was quite a long programme and there was time to get the film ready. Film was also used for drama inserts as TV cameras were still a bit bulky for outside use. With the MKVII studio camera in place we at Marconi needed a new Telecine design as soon as possible. Thus, in about 1966, staff from the Studio Design area moved to create a new Telecine Section.

At the time there were only two or three companies working on Telecine and it never got to be more than about four. Rank-Cintel was our main rival.

In the late 60’s a Telecine was often little more than a modified studio camera with a number of fairly standard projectors fed to it via a mirror light switching system. (Rank-Cintel used a different system called a Flying-Spot Scanner). This initial comment is possibly over simplifying it a little as the optics and the video system needed to be somewhat special (details in a moment). Also the projector was only standard, or nearly so, for countries using the 625 line TV (more or less everyone except the USA and Japan who use 525 lines). The projector problem for 525 line areas arises due to the different rates of the TV system and normal projection rates. Cinema film is shot at 24 frames per second (fps) whereas UK TV operates at 25 fps. The US TV operates at 30fps (matches their 60Hz power). If the projection rate is not close to being synchronous with the TV system, horizontal banding on the picture will result. For the UK a 1960’s Telecine ran at 25fps, the movement was 4% percent faster than in the cinema and the sound a little higher pitched. The projector shutter blanks twice per frame to give a flicker rate of 50Hz and the film is moved (pulled down) during every other one of the shutter closings. Clearly running at 30fps would be very noticeably fast and unacceptable. For the USA a system known as 2:3 was used where the film frames are ‘shown’ to the camera first for 2 half TV frames (fields) and the next film frame is shown for 3 fields, and so on. The projector is now blanking at 60Hz and the time for pulldown is significantly reduced. This is not too bad for the relatively small 16mm film but for 35mm it becomes difficult. The film loop is moved so quickly that it sounds like a machine gun going off. It is may be worth noting that the system used by Cintel at that time had no solution for the USA market, but we did.

The New System

The new system was to be of the ‘island’ type where a number of projectors are mounted on pedestals and directed to the camera via a switchable mirror system. The film projectors, a Phillips 35mm and a 16mm from Bauer (mix as required), already existed and only new control and illumination systems needed to be designed. Another change was to fit sync-interlock drive motors. The sync motor section provided the main drive while the interlock section could be connected to another machine. The internet will provide a description of how the interlock section worked; search for Selsyn. The ability to interlock machines allowed the separate magnetic sound tape already mentioned to be run in synchronism with the film. A two newly designed 35mm slide projectors were also added. The optical multiplexer was also a new design for the B3402.

Camera System Changes

As noted above the camera system for Telecine needed certain changes from the studio camera version. These will be briefly described here.

Gamma Correction

In the electronics a new transfer characteristic was needed. The CRT type displays had a 2.2 power law drive voltage to brightness characteristic, i.e. the black end was crushed. Studio cameras were therefore fitted with a gamma corrector with a 0.5 power law, i.e. black stretch. For the dark surround viewing conditions in cinemas a scene to projection characteristic of a 1.5 power law is used; it would look washed out without this. Because of this the Telecine needs stronger gamma correction; about 0.3 gamma is used. This was not that easy to achieve with the analogue circuitry of the day. Sometimes film that was shot specifically for TV was made lower contrast to make things easier.

Optical Characteristics

Studio Cameras are designed to have specific spectral characteristics. Figure 1 shows the ideal red / green / blue responses for the standard CRT phosphor set. Clearly a camera tube cannot provide negative sensitivity as shown for the red channel and this is achieved by taking a little blue and green signal from the red (Note that this was not fitted to studio cameras until the MkVIII). In fact a nine-term linear matrix is needed.

Now for Telecine we have a different situation. The ‘scene’ is now provided by a yellow / cyan /magenta dye set (figure 2) that has been optimised so that the reproduction is good when the image is view by a human, not a Telecine. The human eye sensitivities are not at all like the colour matching curves of figure 1. To keep this brief, it is far better to equip the Telecine with very narrow spectral bandwidth filtering, and thereby obtain a more accurate representation of the initial film exposure.

The Design

Before tackling the video design it was necessary to learn a little about the characteristics of colour film. With the various manufactures using different processes and ‘tricks’ there is a considerable amount of variation between different film stocks. Additionally a number of negative / positive print processes may have been used if the film was a cinema print. While all this was not so important in early days, it needed close attention as we strived to improve quality. With this initial learning done to a sufficient level, the design work for the camera system was duly completed in fairly short order. This mostly concerned the video processing electronics transfer characteristic as noted already but the ability to reproduce negative film was also included; this would only be in black and white. The system was made by taking a studio camera system and replacing the necessary parts and adding the projector system.

System Proving

As noted above, the use of film to provide the outside shots for dramas was common and it was essential that the film and TV camera pictures matched. For example an actor could be seen walking in the garden and then had to appear to have walked straight into a room. This had to be done as seamlessly as possible. A sudden change of colour was not acceptable. The two shootings might be separated by a numbers of days or more so Continuity had to ensure that the actor was dressed with the same clothes etc. Here may be it should be noted that, except for special effect, all TV pictures are displayed as though they were produced in daylight (6500° Kelvin). Daylight has considerably more blue content than tungsten studio lighting; although Quartz-halogen bulbs run about 300 – 400 degrees C hotter than the 1960’s lamps and therefore have a bit more blue. For the launch of the new system it had been decided that the ability to match film and studio would be demonstrated with 35mm film that had been shot some time earlier being inter-cut with live images of the same set and TV presenter; not an easy task at all. First of all the new film was aging daily and the colorimetry just did not seem right. The technical press pointed to the solution; a special colour corrector (masking) was needed. Initially, due to shortage of time, a home built a corrector was used. It was inserted and now we could achieve quite a good match between film and live video. By the formal demonstration a good balance using split screen could be achieved. The corrector design then had to be reworked for production as soon as possible.

The first big sales for the B3402 Telecine were ITN and Yorkshire TV, others followed.

Further Improvement.

The colour correction just mentioned used a linear matrix much like that employed in the studio cameras. For various reasons that are possibly a little beyond the scope of this article, Telecine masking needs to be a multiplicative function. After some discussion it was decided to upgrade the design to achieve this. Analogue video multipliers were not feasible so the process was done by using logarithms. The signals are log converted, added or subtracted as needed to perform the multiplication or division and then sent to an exponential converter. This is very difficult in analogue form and the digital system of the B3410 that followed in 1980 was brilliant by comparison.  At this time (early 70’s) telecines could reproduce black and white negative film but not colour negative. No reason why colour could not be handled was seen and so this was included in the new design goal. This was eventually achieved and at that point the development programme was more or less wound up.

Further information

For further details click here.

Advertisement: