Photographs |
he following photographs were taken with a
35mm focal plane camera at an exposure of 1/30s, which results in
an annoying bright diagonal band across the screen. The prints are
on 13x9cm rather brown and vinegary-smelling "silk-finished" paper,
so instead I've scanned the original negatives, which are now
rather dusty and scratched, though the results are somewhat
clearer. The trapezoid distortion is a result of my having to shoot
from above the axis of the tube in order to avoid reflections from
a bright window behind the camera. Some 35mm colour reversal slides
were kindly scanned for me by John A Butler, and are much clearer,
being taken (on-axis) from the screen of the new Forgestone set.
The pages transmitted at Christmas 1978 and afterwards were
photographed on panchromatic negative stock and I've taken the
liberty of colourising them using, as far as I can remember, the
original colour scheme.
Note the old-fashioned way of labelling sub-pages. The third of four pages would these days be marked "3/4". Then, it would have been "A B C D", with the "C" flashing slowly. Ceefax had a maximum of four sub-pages, but Oracle sometimes had most of the alphabet strung across a page. It was Ceefax who instigated the change when they introduced new equipment that could transmit more than four sub-pages.
The pages on the Wireless World decoder were selected by thumbwheels on the front panel, so it was not necessary to show the selected page in the top left corner. I opted not to freeze the header display once the page had been captured, so many of these shots don't show the correct number in the page header.
1976 |
![]() |
|
|
|
|
|
|
The handful of glyphs which were introduced with the Tifax ROM together with a few others, such as the "£", "$" and "@" signs were specific to the UK Character Set. Later specifications allowed for a further seven international variants so that accented characters and currency symbols could be accomodated. The set to be used is determined by three control bits in the header row (Packet 0) and can be different for each [sub]page. Later specifications still of what was now known as "Level One Teletext" allowed for individual glyphs to be defined by information transmitted in row numbers higher than 24 and the default transmitted character in any given position on the screen could be overwritten by a suitably equipped decoder to give the correct glyph. |
1977 |
|
Flushed with success at constructing the WW decoder, I designed a new board to display these new modes. It seems to have taken me about a month, just after Easter 1977 to install. Note that the colour change is specified to take place at the right-hand edge of a character, though the WW decoder (because all colour changes originally took place invisibly during a black space) had the change occur mid-way through a character. You can see that the changes from blue to yellow, prior to the word "HOLD" are half a character early. I corrected this later. |
|
|
|
|
..and this one demonstrates background colour. |
|
|
|
|
|
|
|
|
(Note that ORACLE did not carry advertising for the first few years of its life.) |
1978 |
|
You will note that I have added a blue background to the clock at top right, which is permanently cut into the tv picture. Other peculiar "bespoke" effects were to follow. |
|
|
|
ver the spring and summer of 1978 I
constructed a Forgestone colour receiver. This was a state-of-the
art design in kit form, and the huge plywood cabinet I built for it
allowed me to mount the Wireless World teletext decoder internally.
It had a 22" screen which I underscanned to produce the full
rectangular picture, and the teletext display was adjusted to be
the full width of the screen. I ran off a roll of colour slide film
as a record of the initial picture quality, and this included
several teletext shots.
This was the run-up to the famous "Winter of Discontent", and the news stories reflect the climate of the time. ITV was just emerging from a long technicians' strike, and programmes and ORACLE were returning to the screen, though more strife was to come.
|
|
This index page and weather map from CEEFAX on 1 show that the design hadn't changed from the early days. |
|
The very first CEEFAX magazine was transmitted on one channel only, then it was duplicated on the other, with BBC2 using an updated set of "control codes". Originally both had page numbers in the 100s on one day and 200s the next. Eventually, a different service was transmitted on each channel, with BBC1 using Magazine 1 and BBC2 magazine 2, and the old control codes were abandoned. The two services continued like that for many years, until in the late eighties CEEFAX was relaunched with several magazines on each channel. Many pages were common to both, many were different, and some were unavailable, or switched channels, at various times of day and on different days of the week. The Great British Public were not ready for such a complex service and soon it changed to what it is today, with most pages duplicated on both channels and available all day and all week. |
|
|
|
The Tifax character generator ROM contained glyphs for the characters in the range 0-31 and it was an easy matter to have the decoder display them, thus identifying the control characters and showing how some of the pages were composed. The second shot shows the same test page with control characters displayed, and the animated detail shows the glyphs for some of the control characters more clearly. The clockcracker test is a special arrangement of two characters that ensures that the longest possible sequences of 1s or 0s are transmitted consecutively in order to test or "crack" the frequency of the clock oscillator. The only way to ensure that the clock generator - the heartbeat of the decoder - is in step with the transmitter is by comparing it with the pulse that occurs during a 1-to-0 or 0-to-1 transition - when consecutive ones or zeroes are transmitted there's no pulse. If transitions are too far apart it's possible for the decoder to miss a beat, and subsequent characters along the line will be corrupted. In this BBC test page two rows of clockcracker characters are sent alternately. On even rows the ÷ sign and white block give a sequence of .../01111111/11111110/... while on odd rows the Red Text/NUL sequence gives .../10000000/00000001/... The trick was to adjust the frequency of the clock oscillator carefully whilst observing the clockcracker display for minimum errors. |
|
If the signal from London failed, the ITV regional Data Bridges inserted a simple test magazine 1, comprising 99 pages of the caption shown here, and a clockcracker on p199. Although the ITV company was identified, the date and time were not generated dynamically. There was also a generic ORACLE caption occasionally transmitted from London, and it was that that was seen during the prolonged ITV-wide strike of 1978-79 when both programmes and text were off the air. |
|
|
|
ITN provided the news service for ORACLE, and had its own magazine on the 200s. Note the headlines in upper case - ITN newsflashes were also all in upper case white text, which made them look more like agency tickertape. ORACLE has always used most of the eight possible magazines, originally transmitting them sequentially, in numerical order, at the rate of four pages per second. When the number of data lines permitted was doubled to four per field, the IBA launched "High-Speed ORACLE" and transmitted four magazines, one on each data line, which unfortunately is perceived by the viewer as only two pages per second (the rolling numers of the other six, in different magazines, are not seen). ORACLE quickly had a rethink. These days, with more than twelve or more lines available, the situation is more complex, and it's possible to put two or more magazines sequentially in each stream (group of lines) which makes access seem faster by increasing the speed of the rolling numbers in the Page Headers, while having parallel streams means the viewer doesn't see all the magazines rolling by while waiting for the selected page. |
|
|
|
|
|
Because the Wireless World decoder was a set-top box with buttons and switches (the remote control came later) it was annoying to have to get up to clear each newsflash manually, so I put in a timer that erased each one after about fifteen seconds. I can't understand why that's never been a standard feature of decoders. I wonder how many viewers habitually select the newsflash page these days (or even know it's there)? I tend to watch the subtitles instead, but I well remember the Iranian Embassy siege story starting to unfold by newsflash while watching the Crucible Snooker Championships, so I was probably less annoyed or surprised than most when live coverage switched from one Embassy to another. |
|
By 1978 it had been reduced simply to a menu of some of the latest pages, before disappearing completely. |
|
|
|
Later, it was realised that a more appropriate use would be to identify each sub-page of a rotating page by adding the appropriate time code - eg 1980012 for sub-page 12/15 of page 198, so you could select just one page of a set. There was also another ingenious application. A commercial enterprise advertising on ORACLE invited you to phone a premium rate number and be allocated your own "sub-page". Having done so, you would then select page 777xxxx and see a menu of features. You would then key the appropriate number on your phone, and your requested page would be transmitted immediately. If you just keyed in 777 with no sub-page, you could see everyone's selected pages fleetingly, but they were immediately erased. Sadly, the alarm clock feature is no longer used, though there's no reason why - it would not interfere with the use of the codes to identify sub-pages. Perhaps we should campaign for the "World Clock" pages to have their time-codes reinstated. |
|
|
|
|
|
This information page was on p150 - the same as the CEEFAX newsflash page - hence the redirection to ORACLE's p250 Newsflash page. The bottom four rows were boxed and would have appeared as a newsflash/subtitle cut into the picture on the WW decoder, though a present day set would either display the whole page or just the boxed text (depending on the information set in the page header) - turning to "TEXT" as directed wouldn't switch between the two as suggested here. |
|
The test card was received from an out-of-area transmitter, hence the snowy picture. Such poor reception often results in incorrect teletext characters, though I think on this occasion the odd spelling of "subtitles" may be due to an ironic sense of humour on the part of the Editor of the Day. One Christmas CEEFAX transmitted an animated graphical subtitle which depicted the Three Wise Men travelling to Bethlehem. It was shown regularly throughout Christmas Day. CEEFAX used to push the boat out at Christmas, with a fortnight of special features, stories, quizzes and puzzles. Nowadays - nothing special. I suppose it's just a job ruled by bean counters like any other, now. Subtitles on CEEFAX were demonstrated as early as 1975, but as far as I remember the earliest subtitled programmes (around 1977-79) were as follows: Carl Orff: Carmina BuranaThis stage production of this Latin and Old German language work was subtitled in English by CEEFAX. The subtitles were single height, pre-typed, and inserted manually (by several pairs of hands, in fact!) in the way such captions are these days sometimes inserted in to live scripted programmes.Malice Aforethought, starring Hywell BennettThis four-part serial had "précis subtitles". At the end of each scene a newsflash-style box appeared explaining to befuddled deaf viewers what had just happened. As far as I know the experiment was never repeated, though the idea has been resurrected as a feature of the new digital so-called 'interactive' service, whereby comments and critiques have been available as captions in Promenade concerts and Shakespeare plays, for example. Ironically, although the programme itself may be subtitled, the 'interactive' version never is.CEEFAX is Here!This short promotional film about CEEFAX which had been shown on several occasions had a special screening one lunchtime complete with closed CEEFAX subtitles. This was probably the first showing of a programme that had been subtitled in the way we now expect - that is double height, coloured, and synchronised to the videotape time code.Quietly In SwitzerlandAccording to the BBC Annual Report of 1981:"The year 1979 will be remembered as the year in which, for the first time, a television programme was broadcast on the network with accompanying subtitles transmitted on CEEFAX. The programme, Quietly in Switzerland, was of particular interest to deaf viewers since it dealt with the visit of a party of deaf children to Switzerland. Deaf viewers with teletext receivers could enjoy the programme by displaying the CEEFAX subtitles, while other viewers were quite unaware that subtitles were being transmitted. During the year several other programmes were subtitled in this way, including the Queen's Christmas message." Blankety BlankBlankety Blank was a game show involving very little dialogue, but plenty of on-screen text, and so was deemed an ideal choice for the first light entertainment series to be subtitled, in 1979. When ORACLE subtitles began, they had a slightly different style from CEEFAX's in that subtitles were often built up row-by-row as the dialogue progressed instead of the whole screenful of three or four rows appearing straight away. This proved the better format for quiz shows - for a long time CEEFAX used to present the answer at the same time as the question in fast-fire quizzes.The final subtitle shot shows one of the earliest regular subtitled programmes, from about 1980. It was a Sunday evening BBC1 series about a small charter airline business (not to be confused with the later one starring Roy Marsden made by Yorkshire Television). The subtitle actually belongs to the previous scene - I "held" it while I was taking the photograph. The white rectangle at top left is where the page number selected by the recently added remote control appears. It was completely unlike the present day page selection - in fact almost entirely opposite. On pressing "page" on the handset, page acquisition was inhibited (and so served as a hold-page function) and the white rectangle appeared, boxed, showing the selected page number, in black, as the next three keys were pressed. When the page was acquired, the number appeared white-on-black as normal, and the box disappeared from the tv picture. All the teletext functions operated in the tv, as well as the text, mode, so it was possible to select pages in the background, and switch to them once they had been captured. At this stage, I had modified the character generator to give a one-pixel gap between letters instead of three. Comparison with the earlier shots shows how much more legible this was. |
|
|
|
|
|
Back in 1978 though, despite the gloom and despondency prevalent at the end of Sunny Jim's reign, CEEFAX managed a Christmas and New Year magazine packed with puzzles, quizzes, stories, humour, competitions and, yes, satire, before it was all outlawed during the subsequent regime of Queen Maggie. |
|
1979 |
|
|
Meanwhile, I modified the Wireless World decoder to display a red background on the centre of the Header Row, as well as making the whole row double height. The page numbers did not "freeze" when the page had been captured, unlike more modern decoders. This news page would have been in the 200s, and the fact that 703's header is displayed means that the magazines were still being transmitted serially, rather than in parallel with each magazine on a different tv line as is the case today (when you only see the headers from the selected magazine). |
1991 |
|
Finally, an oddity from 1991, captured on my BBC Microcomputer with a Morley teletext adaptor. The file was transferred to my Archimedes, and thence via a DOS disc to my Power Mac, where it was converted to a GIF file. Despite the rather circuitous route, I can assure you it hasn't been edited along the way. My faithful "Beeb" died in the early nineties, and with it went the teletext adaptor. I don't seem to have captured any other teletext pages, and this is the only one I transferred to the Archimedes. |
|
Teletext Website Bookmarks
|
Compiled by Alan Pemberton
Sheffield, South Yorkshire, England
Email me