[PaulM]

The BBC wasn't alone on the 4-tube idea - RCA also thought that it was a neat solution, but they were using a 4.5" IO for luminance with vidicons for RGB which made life tough! The idea was two-fold. Firstly, to avoid the failure of constant luminance and also to minimise the visibility of colour registration errors.

The MkVII has a high-quality, high resolution luminance channel providing the whole of the luminance signal with low-resolution colour 'added' over the top. This was the 'simple' solution and providing the gammas matched it works. Plumbicons have near unity gamma (although provision for correction is provided), so it was just the normal CRT gamma to apply.

The 2001 is much more complicated. The luminance channel only provides the 'highs' of the luminance and the 'lows' are constructed by adding the colouring channels together. The gamma correction becomes much more complex but it was solved and is arguably better than the 'simple' method. It doesn't help any with registration error visibility on the 'lows' but the colourimetry is good - at least for PAL (NTSC coded 2001s came out rather red, so I'm told). Matrix correction can be applied in both designs.

The Marconi MkVIII and MkIX use the standard 'highs out of green' principle whereby the blue and red are low resolution and the green is high resolution. The 'lows' luminance is a simple addition of RGB 'lows' with the 'highs' coming from the high resolution green. With green representing almost all the luminance energy in most scenes, the resulting errors in luminance in the highs are almost invisible. That's as per the PAL/NTSC/SECAM basic principle that the eye/brain colour resolution is quite low. Thus, the separate luminance channel can be dispensed with and that became the standard method for almost all manufacturers in the 1970s and beyond (except Bosch, who liked 3-tube YRB cameras, deriving the green by a matrix technique).

That's all straight out of memory, but I know that there's a lot more detail which is text-book sized!

Incidentally, there were reports of some exported MkVIIs having the luminance channel removed and with some signal re-routing turned into a 3-tube to save money on tubes! The 2001 would have been a tougher proposition to do that without spoiling its colourimetry.

It's all well nigh forgotten now, but there was a huge amount of work went into this all over the world and there were many, many factors to consider such as comet tail colouring, registration stability and optical paths. As an example, the MkVIII and MkIX famously used 'minifiers' in the red and blue to equalise the operating points of the tubes with the green (the loss of resolution wasn't important for red and blue). This led to registration issues (the scan amplitudes were different) and many folk reckoned that this patented feature should have been avoided!

As a bit of light relief, attached is a picture from earlier this year of two MkVIIs in my 'Southern' truck displaying pretty good pictures.

Best regards,

Paul M