A Broader Aspect 

31 March 2006 tbs.pm/2096

Today, there’s plenty of confusion surrounding television aspect ratios. But this has all happened at least once, if not twice, before: Richard Elen and Dave Jeffery investigate.

The very first television images looked nothing like they do today. Instead of modern horizontally scanned, landscape-format images, the picture on a Baird Televisor of the late 1920s and early 30s was in portrait format – taller than it was wide -with an aspect ratio of 3:7 (width:height), designed primarily, it would seem, to accommodate a head and shoulders. The image was made up of just thirty lines scanned vertically, occupying a space equivalent to about that of a postcard, and was in ‘black and orange’ rather than ‘black and white’, being illuminated by a flickering neon lamp.

feb-baird2.jpg

BBC 30-line television caption card

This was very different to the film industry. Here, a landscape-orientated image of aspect ratio 4:3 (four units wide to three high) had been established in the earliest days of the cinema, when William Kennedy-Laurie Dickson, an employee at the Thomas Edison laboratory, invented the Kinescope back in 1889. This employed 35mm film with frames 1 inch wide and 3/4 inches high – a ratio of 1.33:1, or 4:3. Its adoption in 1932 by the Academy of Motion Picture Arts and Sciences led to its becoming known as the ‘Academy Ratio’, and it was the primary (though not exclusive) standard in the cinema until the early 1950s, when additional, wider-screen ratios were introduced (no doubt in reaction to the impact of television), an early example being the Paramount movie Shane, which was shot in 1953 on 35mm film but masked to 1.66:1.

The British Government’s decision to move to so-called ‘high-definition’ television in the 1930s left the old 30-line dimensions behind. Now the image was to be displayed on cathode ray tubes (CRTs), which were for decades almost circular, and Marconi-EMI chose an almost square aspect ratio – 5:4 (1.25:1) – for their 405-line electronically-originated pictures, no doubt to use as much of the CRT screen area as possible.

However, it has been pointed out, by R W Hutchinson in Television Up-To-Date (1935, 1937) and quoted by Alan Pemberton, that while the aspect ratio of Marconi-EMI’s 405-line images was 5:4, the Baird 240-line system which competed with it for the first few months of the BBC Television Service’s operation from late 1936, “…used the slightly wider 4:3… Academy Ratio used by the film industry.”

For at least two reasons, it is entirely reasonable that Bairds would have chosen the same ratio as standard film frames. Firstly, by the time that Baird Television Ltd (BTL) had moved from Long Acre to build an extensive new studio and workshop complex at the Crystal Palace on Sydenham Hill in July 1933, a controlling interest in the company had been bought by a subsidiary of Gaumont-British Film Corporation. One of the applications Isadore Ostrer of Gaumont-British saw for television was as a large-screen display system, which could be used to show live events and centrally-broadcast newsreels in the company’s cinemas: a similar ratio to conventional film would have made sense for both origination and display.

Secondly, Bairds had two different television systems, both based on mechanical scanning. One, the Flying Spot Scanner, was particularly suited to telecine and for a while out-performed Marconi-EMI in this area – at least, when it was working properly. Baird’s German subsidiary developed a system, called the Intermediate Film Technique, which used a film camera followed immediately by an on-line minute-long (and somewhat dangerous) film development process and a flying-spot scanner. The film camera used at the Alexandra Palace BBC Television Station in 1936 was a Vinten 35mm model, the head of which is now in the National Museum of Photography, Film and Television in Bradford. It was modified to take 17.5mm film – 35mm stock split down the middle, with a single row of sprocket-holes.

While we do not have any Intermediate Film shot at AP, we do, thanks to the late Ray Herbert, have a scan of a frame of film from the system used for test transmissions from Baird’s studio facility at the Crystal Palace as well as an image of a test card from the same period. However, this frame of Intermediate Film showing Doris Sonne performing at the Café del Diablo set in Studio One appears to be neither 4:3 (the green box) nor 5:4 (the red box): if anything, the frame is square (yellow box).

baird_aspect-sonne.jpg

Frame of Baird’s Intermediate Film Technique material from Crystal Palace

Also from Ray Herbert, we have a picture of a test card used by Bairds at the Crystal Palace. Although this image is blown up from a rather small part of a picture and may suffer distortion (though the centre circle appears fairly accurate – the blue circle overlaid on the centre fits well), we find that the aspect ratio of the card content is also more or less square.

baird_aspect-testcard.jpg

Baird Television test card used at Crystal Palace, 1935.

It is thus by no means clear if Bairds used a 4:3 ratio when they were running test transmissions from the Crystal Palace between February and June 1935, which is the period during which the above samples were probably generated – in fact a more-or-less square ratio seems more likely. At this time Bairds were running 180 lines, however, not the 240 lines that were to be used in the final system for the BBC at Alexandra Palace. They had only relatively recently adopted electronic display of the received picture and had rapidly ramped up the number of lines in their system’s image: it is quite reasonable that, without any defined standard for high-definition television extant at the time, they may have simply decided to use as much of a circular picture tube’s area as possible.

Baird’s 240-line system began to be tested on-air from the Crystal Palace site from November 1935, with the company demonstrating a dual-standard (240/405-line) receiver suitable for receiving the forthcoming BBC service, the T5, early that month. Edward Pawley in BBC Engineering 1922-1972 reports that 240-line telecine was demonstrated in May 1936, presumably at Alexandra Palace.

240-line test transmissions from the Crystal Palace included a ‘talking picture’ segment, and we have an opening caption from what we presume to be the film section of a Crystal Palace test transmission. Examining this (below), we find that indeed, it conforms to a 4:3 aspect ratio (the green box). This is not surprising, however, as the film segments of the tests could well have been shot conventionally.

baird_aspect-flag.jpg

Courtesy of The MHP

To validate the claim that Bairds used a 4:3 ratio for their Alexandra Palace transmissions in 1936, we require additional evidence. We have photographs of the caption cards announcing both formats of the BBC Television Service from 1936, but it is not possible to see any screen edge on them (they may be taken from the actual cards) and the images we have are both 4:3!

The Baird system was abandoned by the BBC from 8 February 1937 and subsequent broadcasts were made solely with the Marconi-EMI 405-line system. A tuning signal from the late 1930s confirms it to have had a 5:4 ratio.

tuning-signal5-4.jpg

Courtesy of The MHP

When the service returned in 1946 after the War, the first true Test Card was developed. Pawley records:

“…trade test transmissions with electrically generated patterns (artificial bars) and 400-Hz test tone were started on 1 February 1946 and the studio equipment gradually became available for testing in March and April; some production exercises were started in May. The first television test card, test card ‘A’, was produced during the testing period early in 1946, to facilitate the testing of camera channels for frequency response, non-linearity, aspect ratio, streaking and gradation. The frequency response was indicated by the resolution of frequency gratings on the card, from 0.5 to 3 MHz; this method saved a great deal of time compared with the measurement of response, and test patterns came into universal use.”

In 1941, the NTSC (National Television Standards Commission) in the USA had proposed a standard that used the Academy Ratio for television, but it was almost a decade before a change from 5:4 to 4:3 was instituted at the BBC. 5:4 was also the common aspect ratio in Europe during this period, being used in France, Germany, and Italy as well as the UK, according to a television standards guide published in 1943: the only odd one out in Europe was Russia, with an 11:8 ratio. The UK did not change, according to Pawley, until 1950:

“On 3 April 1950 it [the visual aspect ratio of the television picture] was altered from the pre-war standard of 5:4 (width: height) to 4:3, thus bringing it into line with the format of standard cinema film.”

Alan Pemberton notes, interestingly, that he had been unable to unearth any pre-1950 test cards or tuning signals in 5:4 ratio. However, while Alan was not able to locate a 5:4 version of Test Card A, we have, thanks to Arthur Dungate, who provides a frame of 35mm film on his fascinating site Direct Television from Alexandra Palace, which shows exactly such a card. Superimposing this image and the 4:3 version mentioned by Alan and other sources, it is evident that versions of Test Card A existed for both aspect ratios and there are certain differences evident between them.

testcardacompared.jpg

In particular, notice the difference in castellations at the corners, and the differing thickness of the centre circle. Here’s a reconstruction of the 5:4 version, derived from Arthur Dungate’s film frame:

TCA_54.jpg

This is interesting, because according to Pawley, “Test Card C was used from 1947…”, and one might have imagined that once the more comprehensive Test Card C had been developed, with its greyscales and other more highly-developed features, Test Card A would have fallen into disuse. Evidently not: it was worth someone’s while re-drawing it to the new aspect ratio.

The overlap is no doubt due to the fact that while labelled as part of an apparent sequence, these different cards had distinct applications, as hinted at by Pawley, as indicated previously. It seems very likely that Test Card A (and the almost-lost Test Card B) were designed primarily as camera line up charts (later replaced by the similar looking Test Card H), whereas Test Card C was designed to permit alignment of the entire chain from transmitter to home receiver. This probably explains the redrawing of Test Card A into 4:3. Indeed George Hersee, the originator of the famous Test Card F, refers in an Andrew Emmerson interview to Test Cards A and B being used to test cameras to see “basically, how well they turn light into electricity”. This is very different to how we think of, say, Test Card C being used. Hersee also talks of the pitfalls of designing a card for one purpose that can almost be used for several others.

Test Card C, introduced in 1947, also existed in both ratios. Here is a reconstruction of the 5:4 version, and a superimposition of the two versions (4:3 overlaid in magenta) to show how they differ:

tcc_54.jpg
tcc_compared-recon.jpg

Similarly, while Alan Pemberton’s BBC tuning signal with a clock at the centre and greyscale panels either side is definitely 4:3 in aspect ratio, another version of this signal (provided by The MHP, below) is equally evidently 5:4. Again, this existed in both formats.

300bbcts-clockts.jpg

Alan hits the nail on the head when he considers the practical challenges associated with a change of aspect ratio. He comments:

“The original specification for the 405-line standard had a blanking period per field of ‘at least ten lines’, but later it was standardised at 14 lines. If this change was made at the same time as the aspect ratio change it doesn’t account for it entirely. Reducing the number of active lines per frame from 385 to 377 gives a change in aspect ratio of 1.25:1 (5:4) to 1.28:1 rather than 1.33:1 (4:3). The number of active lines on that basis would have to be 361, so it has to be assumed that the camera scan amplitudes were adjusted as well.” No wonder we have two versions of Test Card A.

But what of the impact on the viewer of this change from 5:4 to 4:3? Unlike today’s shift from 4:3 to 16:9, where an error is extremely obvious on-screen (the “I see fat people” effect), in 1950 it was different. “Theoretically, all the domestic receivers would also have needed adjustment,” says Alan, “but in practice the stability of the circuitry of the day meant that the height and width would vary by more than that in an evening’s viewing.”

In other words, on 3 April 1950, nobody at home would have noticed a thing.

Thanks to Alan Pemberton for the invaluable assistance his research provided for this article. Reconstructions by Dave Jeffery.

A Transdiffusion Presentation

Report an error

Author

Richard Elen and Dave Jeffery Contact More by me

Your comment

Enter it below