Tuesday, March 08, 2016

A review of Colour Reproduction in Electronic Imaging Systems: Photography, Television, Cinematography by Michael S. Tooms

All of my colourimetry comes from doing it; my theoretical knowledge isn't that great as I didn't study it at university and the colour training at the Beeb, although entirely appropriate to what a TV engineer needs was just that; TV-centric.
Over the last ten years I would credit the following industry colleagues for widening my understanding and appreciation of colour reproduction;
  1. Martin Euredjian of eCinema Systems was a colourist and telecine engineer before he started making reference-grade LCD monitors - Root6 sold them for a while (around a decade ago) and Martin was very generous in his product training. He opened my eyes to the problems of metameristic failure between different display technologies.
  2. Martin Parsons of Image Eyes who used to be the head of colour at MPC; like me he's now a colour-gun-for-hire but I always enjoy getting his take on things and since he 's done an awful lot of calibration he knows when it's appropriate to maybe go slightly off standard to match two displays and keep the colourist happy.
  3. Steve Shaw is the proprietor of LightIllusion where I have bought probes, LUT boxes and of course LightSpaceCMS which really is the best calibration package out there. In terms of speed and accuracy nothing comes close.  Steve is very gracious with his knowledge and runs a good forum on his website as well as being a prolific presence on Lift, Gamma, Gain.
I recently was recommended Michael S. Toon's book Colour Reproduction in Electronic Imaging Systems: Photography, Television, Cinematography.  I picked this up at Amazon to read on the iPad Kindle app and like most academic volumes it comes in around sixty quid.

So - as mentioned, my background is the CIE1931 chromaticity chart and how that then relates to Rec.601, 709, the DCI-P3 colour space and Rec.2020 (the proposed colour space for UHD TV which is as yet no available on any current-model monitor).

Any practical skills I have with colour come from;
  1. Calibrating and matching TV cameras ("racking" or "shading" in the USA) for studio shoots.
  2. Calibrating and matching TV monitors - firstly in tape edit suites (it was always common practise to have two grade-1 monitors; the record monitor and the preview monitor).
So I think I have a good eye for quickly spotting how two displays differ - I can match two monitors or cameras pretty quickly but like most people I don't have perfect colour memory (and don't let any colourists tell you that they do!).


The first things the book set me straight on was the history of colour research before the CIE1931 diagram. Before cinematography and early colour TV experiments the concerns of industrial colour science was mixing of paints and pigments. An important consideration in colour mixing is being able to predict new colours from the volumetric mix of two pigments and the early colour models (Munsell and Guild & Wright) tried to reflect this with a choice of primaries at 460, 530 and 650nm giving rise to constant lines of "just noticeable difference" - JND - on their chromaticity charts.



However - the negative lobes associated with the gamma (green) and rho (red) responses mean that an all-encompassing colour chart should place the primaries outside the spectral-locus which gives rise to non-consistent lines of JND and a grey-scale definition in video colour spaces that seems to be at odds with the equal-energy point in the centre of the CIE1931 chart that gives us our white mix.

one of the worksheets that are available for download with all the empirical data used in the book tabulated and plotted

So I have no hesitation about recommending this book; it is readable and has enough maths so that you never think there is an slight-of-hand going on. For me it has filled in the holes that bothered me in my knowledge of colourimetry. If all the changes that are going on currently in the industry (principally Rec.2020 and the XYZ colour space) are something you have to deal with then this provides and excellent bedrock of colour knowledge.

No comments: