Early BBC colour tests.

[ukcol]

The following may be of interest and is a reproduction, in full, of a short report from Electronic Engineering Vol 27 No 333. November 1955.


Colour Television Tests.


The BBC have recently begun a series of experimental transmissions in colour from its transmitting station at Alexandra Palace. These transmissions will in no sense be a public service, and are not an indication of any early start with colour television. They will consist mainly of still patterns, demonstration films in colour, and simple studio shots, designed to provide technical information. The transmissions will have no interest as programmes; they will take place outside normal programme hours and are likely to continue for some months. They are being carried out in agreement with the Television Advisory Committee, who have been asked by the Postmaster General to report on the whole field of colour television.

Preliminary work on a colour television system which might prove suitable for a public service in this country has been in progress in the BBC’s Research Laboratories for some years. This work has been done in close collaboration with the radio industry, and it has now reached the stage where it is desirable to study the practical problems of transmission and reception such as would be encountered in a public service.

The system to be used in these first experiments is based on the N.T.S.C. system adopted for public service in the U.S.A., it has been adapted to make it suitable for British standards. Other systems may be tested later on. One of the problems is whether a truly compatible system should be the final objective, but it is first necessary to establish whether such a system is possible using the adapted N.T.S.C. standard. An important feature of the experiments will be to examine whether, if this system is used, there is any appreciable degradation of the picture as received on present-day black and white receivers.

The studio colour camera and the associated equipment to be used in these experiments has been supplied by Marconi’s Wireless Telegraph Company Limited. The film and slide scanning equipment has been developed and built by the BBC Research Department.


xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx


It is interesting to read what was going on at this time and to relate it to the very successful PAL system that was adopted in 1967. The main problems of transmission and reception had already been solved by NTSC, particularly the compatibility problem. PAL solved the problems associated with phase distortion and tweaked NTSC for the 625 line structure. It seems a pity that such a good system will only have been in service about 40 years by the time it is abandoned.

[ppppenguin]

Originally Posted by David_Robinson

NTSC is much better than PAL in this respect, for a similar subcarrier frequency.

When PAL was being invented, I think they tried initially with the subcarrier as an odd multiple of half line rate. Just as for NTSC. This gives a horribly visible pattern. So they went for a quarter line offset which is a vast improvement. Then they added that 25Hz offset in the subcarrier frequency. This breaks up the pattern by making it move. It also gives us the 8 field sequence that has been cursed for many years.

For PAL B/G/I: Fsc= 1135/4 * Fh * Fv/2

For NTSC M: Fsc = 455/2 * Fh

[David_Robinson]

Originally Posted by ppppenguin

For PAL B/G/I: Fsc= 1135/4 * Fh * Fv/2

I think you meant

1135/4 * Fh + Fv/2

I claim my prize!

[ukcol]

Originally Posted by ppppenguin

For PAL B/G/I: Fsc= 1135/4 * Fh * Fv/2

Fsc = (1135/4) * 15625 + (50/2) = 4.43361875 MHz


Now you know where that odd frequency comes from.

I understand that the timebase frequencies are generated by counting down from the sub-carrier frequency, not the other way round. Never having worked at the broadcast end of the chain I don't know from experience. Is that correct Jeffrey?

[ppppenguin]

In early colour work, the usual arrangment was to start with 2fsc for NTSC or 4fsc for PAL. Then divide down. For PAL you fangled the 25Hz offset with a balanced modulator. It was actually more complex than this, especially before modern digital divider chips but this gives you some idea.

There are several modern approaches but the one I use in my own designs (and is used in many encoder chips etc) is a discrete time oscillator (DTO). The following explanation is a bit complicated but I'm not sure I can do any better in a few sentences. The same principles are used in many synthesised signal generators and in all direct digital synthesis (DDS) systems.

As a simple example, if you take an 8 bit adder/accumulator and set the coeffficient input to 3 it will have the output sequence 3,6,9, etc etc. This is the digital representation of a sawtooth wave of frequency 3/256 times the accumulator clock. This can then be put through a sine lookup table to get a digital sine wave. This can then be used in a digital system or converted to analogue. By suitable choice of accumulator length and coefficient size this method can be used to divide any frequency down to any other frequency (Fin > 2Fout). Typically 27MHz is divided down to subcarrier in a modern design.

The actual design is complicated by the fact that the division ratio is not usually exactly correct, whatever coefficient you choose and however large the accumulator. This can be overcome by several methods, the simplest of which in a PAL coder is to reset the accumulator to zero once every 8 fields. Typically I use a 32 bit accumulator and the phase jump at the reset is well under 0.1 degrees. In some applications you don't need to convert the sawtooth to a sinewave and you can just use the MSB of the sawtooth as the output frequency, albeit with some cycles too short and others too long.

[ppppenguin]

The USSR used SECAM for a mixture of political and technical reasons.

The technical reasons are that it it didn't suffer too much over their very long distribution chains. Also the Soviet copies of quadruplex VTRs could not reliably replay PAL or NTSC since they did not have the equivalent to Ampex Colortec which used some prettyhigh technology to stabilse the output.

Politically, I believe the French gave some financial incentives to use SECAM.

While we're on the subject of SECAM, and to bring this thread back on topic, there was a variant of SECAM proposed by a BBC engineer, the late Brian Pethers. It is referenced here: http://www.pembers.freeserve.co.uk/W...Standards.html

[Booby]

Perhaps also a lost opportunity for the huge dual standard colour TVs of the 1960s , in that there could have been colour on 405 lines until the 625 line network was ready?

[paulsherwin]

Originally Posted by Booby

Perhaps also a lost opportunity for the huge dual standard colour TVs of the 1960s , in that there could have been colour on 405 lines until the 625 line network was ready?

I believe it was a policy decision not to launch colour on 405 once the strategic decision had been taken to move to 625 line PAL.

Best regards, Paul

[brianc]

Hi all

Jeff mentioned the late Brian Pethers and the colour system he developed. I knew him quite well and used to see him working away behind the racks in studio H at Lime Grove which was the BBC's later (early '60s) colour test area. When I first saw the equipment, it was very much a breadboard lash-up but it produced excellent pictures. He got an award from the BBC in "recognition of inovative work" but as Jeff said, it was the Russians who were credited with the system. As a matter of fact, I have a "Pethers" decoder built on an Electrokit chassis which, presumably, is the last existing part of his system. It will go to Bradford when I get round to it.

Cheers

Brian