Thursday, January 2, 2014
Colour is colour!
It all started days ago with the subject of color management in digital imaging workflows. After many years of miserably failed efforts, I finally took the daring step to fundamentally and correctly grasp the darn process, once and for all, and henceforth apply it properly in my own photography, mainly for reasons of improved productivity, predictability and consistency of output.
In an earlier post I have explained some of my recent colour management experiences, rather extensively I’d dare say. However, from early on, there have been quite a few bits that bothered me in the way science explained human vision. Things like, for instance, the ‘known fact’ about us being ‘trichromatic’. Why precisely trichromatic? Do our eyes work like DSLR CCDs then, and our cortex generates visual images just like those a computer graphics card sends to the monitor? And, if our colour vision had to be RGB based, why has evolution decided to work with precisely those three peak absorption frequencies that we associate with the colours Red, Green and Blue? What’s so 'special' about these frequencies, please tell me, at least in terms of effectiveness and efficiency of the 'human vision' process? Since it is my strong belief that nature itself behaves in most ingenious ways in deciding how to execute its many processes, there must have been a genuine rationale why those three R,G and B were selected and became the ‘primary’ colours, not? Ever crossed your mind, what would have been our colour experience of the space and objects surrounding us if evolution had yielded a different set of primary colours? Would we be ‘better’ off? Or ‘worse’? What makes a colour ‘good’ or ‘bad’ after all? Saturation or brightness? Have you ever thought about that? What if we suddenly experienced deep green skies at noon and purple seas with black foliage on cyan tree stems? Pink elephants, somebody? BTW, do your reds look ’n feel just like mine? Do we experience in other words, among ourselves that is, ‘identical’ colour sensations? All the time? I could go on and on and on... Understanding human vision managed to be not my usual cup-of-tea, I bet.
Also in my recent readings, I came across this generic definition about sensation that proved quite an eyeopener. At least to me it was. «Perception is information about our surrounding, and the way this materializes is via ‘energy’ out there that is captured by our sensor cells and becomes ‘information’ inside our nervous system.» If you heard about Einstein’s formula of transformation of energy into mass (or vice versa), well, in sensation, ‘energy’ is also captured but this time it is transformed into neural ‘information’ and the result of the transformation we call ‘perception’. How cool is that ?! Yes sir, never heard that before framed this way. But it’s so cool and also seems so true! (Found in the Blackwell Handbook of Sensation and Perception edited by E.Bruce Goldstein, Chapter 2, Abstract, by Michael Levine). One’s never too old to learn, indeed.
Hence, it has been so far scientifically ‘accepted’ that human vision is trichromatic. However, don’t imagine even for a second that our brain (our visual cortex for that matter) works like a computer monitor with either LCD cells or CRT phosphors. It's not quite like a lil’bit of this (Red), a lil’bit of that (Green), and why not an itchy-bitchy this too (Blue), and voila! We got a Prussian Blue hue on that cunning pixel! The actual biological process of colour perception is a lot more esoteric than any common laptop monitor will ever be. I am not even sure whether all experts entirely agree about the exact pathways of our nervous system enabling us to ‘see’ colour (or better said, to experience the ‘sensation’ of it). Of course, as scientific research gallops rapidly ahead, we kind of «gettin’ there», and one day I am sure, hopefully not too far in the distant future, we’ll get all the answers we need! We always do, as a matter of fact.
For what is worth, there’s on one hand the RGB additive colour theory about human vision, where the final colour sensation eventually gets composed by 'accumulating' red, green and blue sensations together (like with computer or TV monitors), but there’s also the ‘opponent colours’ theory, where the perceived differences among certain 'primary' colours become in fact the key to our colour vision. Not too intuitive an idea if you ask me, however, not without some merit notwithstanding, as differences often become better discriminants than larger absolute values. It’s like the EBT (earnings before tax). EBT is always a relatively small number compared to its two constituent larger quantities of which it is the difference, one for the revenue and a second for business expenses. Small percent fluctuations in the two large numbers can have devastating impact on that wee tiny difference, the result often spreading all over the map (red to black and back to red again)! In a likewise reasoning, it’s probably ‘easier’ for the visual cortex to distinguish colours in terms of how different they are from red vs. green or yellow vs. blue than to actually translate some absolute values of energy absorption into a real world color sensation! Sounds amazing! We see colours in terms of differences of somehow ‘opponent’ primary colours! Jeez! God must have been real smart when He thought of that!
In my further ‘quest for the truth’ I was deeply disturbed and intrigued by the ‘signalling processes’ of retina light sensitive cells, the so-called cones and rods. Especially the part that said it were mere proteins, called ‘opsins’ (being firmly fixed in place by 'drilling' thru cone and rod cell membranes with their seven alpha-helix ‘screws’ each), which are responsible for ‘tuning’ a key molecule, a derivative of vitamin-A called ‘retinal’, into absorbing photon energy of predefined wavelengths. Indeed, energy absorption occurs in the form of bell-shaped spectra spanning along large parts of the entire region of visible wavelengths, and it does not happen by any means at only in a few single individual frequencies. There goes the monitor RGB colour reproduction model! There are of course maxima in each spectrum, correspondingly at three ‘special’ wavelengths of 564 (red), 534 (green) and 420 (blue) nanometers, but there’s also energy absorption in each of the spectra along many more frequencies around the peak. In the absence of light, retinals are typically bound to their ‘hosting’ opsins, and wait for photons to ‘pay’ them a visit. 11-Cis-Retinals (that’s how subject matter experts call the specific initial retinal conformations) basically resemble grizzly bears waiting with open-wide gobs to pluck red salmons mid air as the latter jump-swim upstream... Got the picture? When a retinal ‘plucks’ and 'sucks off' energy from an entering photon, one of the molecule’s carbon bindings, the one at the eleventh retinal carbon atom position, breaks up and part of the retinal body twists to bind again in a new spacial position; it isomerizes in other words into the so-called trans-all-Retinal. In that state, it subsequently breaks away from the hosting opsin, which in turn triggers a complex but lightning fast signalling pathway (some chemical events happen in pico secs indeed) in order to send a ‘signal’ to the optical nerve fibers with which the cone/rod cells are synapsed. The pathway of signalling events is kind of reminiscent of a tile layout in a Guinness Book of Records domino tile falling competition. A causes B that in turn forces C to happen and, because of that, D kicks off E... until significant electric potential is built upon the cone/rod cell membrane (a phenomenon known as transduction). This eventually signals the initial trigger from the ‘plucked’ salmon (sorry photon) towards an attached ganglion (a nerve cell) in the optical nerve. Gazillion times a day, for each of the seven and a half billion human inhabitants of planet Earth. Not to mention the remaining living creatures large and small...
Curiosity killed the cat, they say, but I ain’t no cat whatsoever, so, there was still something I didn’t quite get and wanted to find out. It really bothered me to gradually learn how nature worked for colour vision and I hadn’t yet a clue! The theory goes like this: Depending on the type of cone, in other words, red-, green- or blue-sensitive, absorption spectra are positioned around laterally displaced maxima (see graph in the following paragraph). In reality, a single 11-Cis-Retinal molecule without bounds to opsins will be normally excited by photons in the ultra violet region of the electromagnetic spectrum!!! In other words, a 11-Cis-Retinal needs real hot Kelvin stuff to work-out; it needs to swallow high energy UV photons indeed. In the UV region there are photons of considerable energy that normally get filtered away, largely in the atmosphere and also in our eyes before they reach the retina. Therefore, our own retinals in our cones and rods won’t be typically isomerized by invisible UV light, not ever, unless you’re crazy enough to look straight into the sun without sunglasses. Now, then, retinals sitting entangled among the alpha-helices of their ‘hosting’ opsins in any of the three cone types, and waiting for photons to come by, were eventually found to isomerize by photons of lesser than UV energy, coming this time from the visible part of the electromagnetic spectrum. In fact they appear to be happy with mere photons of blue, green and red visible light, right? How could this be achieved, then? What is that those three opsins have that they become capable of tuning their retinals into isomerization by sucking less than needed photon energy? C'm on, I almost gave away the answer...
Now, like you all dudes know, opsins are proteins, and as such they are macromolecules that appear like long chains of a 'backbone' upon which, few atoms apart, amino acids are suspended, one at the time. Most of them buggers are electrically neutral and hate water (hydrophobic), but some are just the opposite, they love to flirt with water all the time (hydrophilic, with OH+ bits hanging out). If you come to think about it, it’s all about electric energy states forming in the neighbourhood of the connected retinals, them waiting for light to shine, get excited and get the hell-outta there, real quick. These electric potential states must then logically be the result of which amino-acids precisely participate in the retinal ‘entourage’! In other words, how many hydrophilic among them are suspended really close to our target 11-Cis? The more electrical energy present the lesser incident energy will be needed to break the Gordian knot of Carbon nr 11, right? Scientists found that opsins with green and red light absorption sensitivity are structurally a lot more alike with only 4% different amino acids hanging from their backbone. I bet you, this difference will be more outspoken in the neighbourhood of the 11-Cis-Retinal attachments than in the remainder of the protein. This 4% difference is just enough (evolution has managed) to make the green and red absorption spectra (L and M) look very much the same, but nevertheless spatially apart and having red and green (564 and 534 nanometers) located maxima respectively. Also, they seem to be cell-manufactured based on genes located in our X chromosomes, meaning that female humans (with two X’s mainly unless you are called Jamie-Lee Curtis) are less likely to suffer from red and green cone malfunctions or possible absence of such cones altogether, or in colloquial tongue, suffer from red-green colour-blindness. Amazing how nature works, innit?