Peter Swinson's History of Telecine

From TIGwiki2
Jump to: navigation, search

History of telecine 

This is a UK perspective and no doubt others can add details of US and other countries developments.


Flying Spot Systems, pre Telecine

In the late 1920’s and early 1930’s John Logie Baird (JLB) was working on a system of sending moving images over the AM radio network. Rather than inventing a TV type camera he concentrated on using a technique that we now refer to as flying spot.  While the CRT had by then been invented it was way too weak to be of any use for scanning or indeed displaying a moving image. 

IMAGE 1: A 1935 studio scanner (click image to view full size)
Most of JLB’s imaging inventions made use of the Nipkow disk. This  effectively generated a raster scan (see Wikipedia drawing) and could be used for capturing a moving image and displaying it. In the capture mode the scene was brightly lit and a lens focused through the spinning Nipkow disk focusing the image onto a photocell. The first flying spot scanner, albeit of an original image rather than a film image. 


If this photocell signal was transmitted together with synchronization information about the spinning Nipkow disk’s speed and a reference pulse for a start position of the disk, a second receiving Nipkow disk could be made to spin in snyc with the transmission. A neon bulb modulated by the signal from the transmitting photocell was placed behind the Nipkow disk and when the observer looked through the disk the transmitted moving image could be seen. This, at least in the UK was the origination of TV in the 1930s. For the technically minded the system was a 30 line transmission, I am not sure what the field rate was but it must have been one disk rotation. The transmissions were made in the evenings on the BBC medium wave after audio transmissions had shut down. 


The challenge for a real TV camera system

Very soon after these experimental transmissions the BBC set a challenge for a plausible live TV system to be developed. JLB and EMI (Electro Mechanical Industries) competed for the “prize”. At this time others were developing genuine electronic tube cameras, Farnsworth in the USA , EMI in the UK and others. 


JLB needed to improve on his camera system as the Nipkow disk and photocell were not good in terms of light gathering speed. So he worked on two new systems, both still relying on Nipkow Disk flying spot scanning.  The first was his flying spot studio. A large Nipkow disk was placed in front of a powerful carbon arc lamp, the resulting very bright flying spot was projected through an aperture into a totally dark room. (Image 1 is a 1935 version, image 2 a 1937 version).

IMAGE 2: A 1937 studio scanner (click image to view full size)

in the room, (Image 3)

IMAGE 3: a "live" 1937 flying spot studio room (click image to view full size)
actors  etc would perform and around the wall in the room where the aperture was projecting the flying spot were placed sensitive photocells that captured the shades and tones reflected off the actors etc as the spot scanned them. This as far as I know was the only time that flying spot was really “live”. Obviously it was not a great success and also the studio area was very constrictive..



The Telecine is born

His second new system was truly the beginning of telecine. He chose not to develop an “electronic camera”, although Farnsworth had sent him one to try out. ( Image 4). No, JLB knew that motion picture film had a very high sensitivity compared to any electronic developments.



IMAGE 4: Farnsworth camera loaned to Baird (click image to view full size)

He set about a system that shot b/w film on a fixed studio set that could be normally lit. He then in the same “box” processed the film and then, using the ubiquitous Nipkow disk, carbon arc light and a photocell, scanned the film while it was still wet.( Image 5 and Image 6)

IMAGE 5: Baird Intermediate System 1937 (click image to view full size)
 
After much experimentation he managed an end to end time of shoot to transmit of under 1 minute. This was largely achieved by using a concentrated cyanide solution in the developer! And yes, I have interviewed the chemist that did the work for Baird and after transmission they did flush all the chemicals down the drains.  Most of you will recognize the back end of this system comprises a telecine. 


Needless to say after some BBC trials of JLB’s system and the EMI image Orthicon camera the JLB system lost. 

IMAGE 6: Early Baird Photomultipler Detector (click image to view full size)
IMAGE 7: A 1937 Telecine system (click image to view full size)
However once TV transmissions started it was clear that there was a great need to transmit theatrical films, news reels and other material based on Nitrate or Acetate film.



JLB of course had already demonstrated the flying spot telecine principle in his BBC challenge. Using the same type of scanner, carbon arc and Nipkow disk, he set about building 35mm telecines for the BBC. (His original camera/telecine system used 35mm split to 17.5mm to save costs, he was after all a Scot!)


Two of Baird’s Flying spot Film scanners (Image 7 )were employed by the BBC for some years before WW II, when all TV in the UK stopped. 


By the end of the war photoconductive camera tubes had evolved to the state where film scanning consisted of a fast pull down projector and a “video camera”, commonly known as a telecine chain. To cater for multiple film sources, 35mm, 16mm and slides, many of these chains comprised a single video camera and an array of projectors that the camera could be switched to via mirror arrangements.


Rank Cintel

Post war the John Logie Baird company was re-named Cinema Television Limited, later purchased by the Rank Organisation, and the name shortened to Cintel. So was born Rank Cintel. In the 1960s and early 70’s Cintel developed both photoconductive telecines and color Flying Spot telecines.


The flying Spot versions now used high resolution high intensity CRTs with phosphors designed by a UK company called Rank Electronic Tubes, which later merged with Brimar. The CRT based units allowed the film to run continuously with no intermittent motion. Early versions, known as Twin Lens telecines overcame the interlace scanning by having a fixed raster projected through two lenses , one above the other. A shutter uncovered one lens for the first field and the second lens for the second field, the lens offset being the distance the film had moved between successive fields. This obviously only worked for scanning film at two fields per frame. If I remember rightly  Bosch at that time also made a CRT based flying spot scanner where the vertical scan was accomplished using a rocking mirror that tracked the continuously moving film.


The Famous Rank Cintel MKIII

In the mid 1970’s Rank Cintel designed the MKIII telecine a radical new concept at the time, where the scan patch comprised of two field rasters on one CRT where one field shifted relative to the other to compensate for the film movement between fields. This system again was only good for 625/50 25fps systems. However Cintel came up with a version for 525/60 systems wherea 3:2 pulldown was required. This was achieved by using 5 different patches on the CRT to compensate for the different position of the film during each scanned field. As some of you who worked on these machines will remember, adjusting the geometry of the scan for each patch was  a nightmare. Misalignment caused a sort of flicker twitch of the image between fields; relatively easy to adjust for two field positions but not for five!


Then in the later 1970’s someone at Cintel had a Eureka moment. Why not scan each film frame progressively then extract the required fields. Some of you will note that this same Eureka moment was occurring at a certain company called Image Transform, in Hollywood!


The first Digital  Stores for Telecine

But the progressive scan would have to be stored and then each field “read out” of the store. This was early days of computers and memory chips were small, maybe 8 Kbytes or less.


Anyway, Cintel incorporated a two frame store into a MKIII and called it the Digiscan, this was way before any digital video hit the world. With the images in the stores, almost any film frame rate and system standard could be catered for, Indeed many believe that the Cintel MKIII Digiscan telecine signed the death knell of the old Projector based film chains which , in the USA had been the bed rock of 3:2 pulldown telecine 


Digital Television, the modern era

The Rank Cintel MKIII went through many iterations, when digital video in the form of REC 601 became the norm,  Cintel and others already had many years experience in internal digital processing. It was therefore relatively easy in the early 1980s to provide full digital Flying Spot Tele

cines. By this time Cintel had developed an early CCD line array scanner te ADS-1, however it had limited functiona

lity compared to Flying Spot systems of the day and it never caught on.


While Cintel continued to develop ever more sophisticated flying spot scanners, the URSA range and al

l that followed, others were developing CCD based scanners, both line and area array with great success, and the re

st as they say is history. 


Kinoscope.


The Kinoscope was mentioned on the original thread  The Ki noscope is the exact opposite of a telecine. A Kinoscope records electronic TV images onto film. </p>


Before the invention and commercial use of the VTR, the only means to preserve any TV programme was t

o record it to film. While in this era many programmes were shot and edited on film, and just required telecines to

transmit, any TV camera material was lost after transmission.


Th Kinoscope usually comprised of a film camera and some form of electronic projection of the tv sign

al into the camera.  In its crudest form the film camera just shot a hig

h quality tv screen, however much more sophisticated systems existed in the 1970’s prior to VTR.


Two such color Kinoscope systems come to mind, although they were developed  after VTRs were available (Ampex Quad), many still wished to preserve the tv images on film). 


In th UK the film lab Colour Film Services (CFS) had a massive device that looked a bit like a battle

ship, probably because they painted it battleship grey! It comprised of a 16mm film camera with a very fast pulldow

n, (needed due to short frame interval between TV frames). Arranged in front of the camera at angles were three very high resolution monochrome CRTs, all based on scanner CRTs. Each one was fed a Red, Green or Blue tv signal. These CRTs were focused at the film camera aperture through RGB dichroic mirrors arranged to combine the RGB images at the film plane. This system, at the time was regarded as the highest quality kinescope available. 


A little later in time Rank Duplication in London acquired from CBS in the USA two laser based kinosc opes. I saw them only twice and my memory is sketchy. But I think they each comprised of RGB lasers firing into a film camera that had an extremely fast pulldown created by a vacuum on the bottom film loop in the camera. When running the noise was incredible, the vacuum pulldown sound like a machine gun


Both these devices ran at real time, I guess because digital stores did not exist to “slow down” the video.


Around this time ITV in the UK, or it may have been ITN, developed a machine that recorded a color tv

signal to B/W film in an almost modern way. They split the tv image ito its B/W (Y) component and its two color co mponents, (Cr & Cb). On the film it recorded the Y signal across half the film width and then each color component across  the adjacent two remaining quarter widths (Yes the image was

squeezed 2:1 horizontally). A b/w film scanner could then recover the images and they could be electronically recombined to form a color tv image.



In the 1990’s Sony in the UK developed a HD video Kinoscope where the image was recorded onto 35mm b/ w film directly by an electron beam, all carried out in a vacuum. 

Known as the Electron Beam Recorder (EBR), each HD tv frame was recorded as three adjacent R,G,B vertical film frames. The resulting film was then sent to a local film lab, Rank Film Labs, now Deluxe, in the UK . The film was loaded into an optical effects printer and each frame was step printed via sequential RGB filters onto co lor film stock. At this time it was the only way to get HD, shot on an HD video camera, back onto film. 


Of course today in the digital world we need to record our digital masters back onto film. The modern

film recorders are indeed sophisticated versions of Kinoscopes. And Scanners are the modern sophisticated versions

of Telecines

Whew, well that wore me out, any questions

Cheers


Peter 


8th October 2010



Rob Lingelbach 00:45, 20 November 2010 (UTC)