I've had a few messages from various individuals querying white and I fear people are getting the wrong end of the stick. I am
not suggesting an off-white model paint here.
Pure white does not exist in reality. Human eyes are wrong 100% of the time. They receive light and your brain tries to assemble it all into an image.
Pure white and pure black are theoretical. Pure white being 255, 255, 255 in the RGB model and L100 a0 b0 in CIELAB, whilst pure black is 0, 0, 0 and L0 a0 b0 respectively. They simply do not exist in reality. You may
think you see them, but you don't. It's all juxta position and reflection of ambient light. I realise it's an uncomfortable notion to consider that absolutely everything you have perceived and relied upon from your eyes is distorted, but it is. There are many well understood biological reasons for this. I am not a biologist though but a degree educated Engineer. I understand the optical measurement aspects, but will kindly direct the reader to Wikipedia et al for the biology lesson on why your eyes are not to be trusted as being more accurate than calibrated scientific instruments.
Here is but one example. RAL 9010 Pure White. It's as white as any real-life white paint tends to be. Is it actually white? Of course not. Rendered on a screen it looks deceptively creamy and no doubt your eyes are not telling your brain the screen is wrong. It probably is, but so are your eyes. Using a calibrated light source and peer reviewed and validated scientific measurement, RAL 9010 has published colourspace values as follows:
RGB: 240, 237, 225
CIELAB: L93.61 a-0.48 b6.08
Light Reflectance Value: 84%
Look it up for yourself here complete with rendered swatch - infact I implore the reader to do so:
http://www.e-paint.co.uk/Lab_values.asp?cRange=RAL%20Classic&cRef=RAL%209010&cDescription=Pure%20whiteHere's another which you may think is a computer generated image. This is a pair of identical 3D face masks. One is coloured gold, the other coated in a rather extraordinary substance called Vantablack - which I believe is the least reflective colour and coating known to mankind at present:
Attachment:
vantablack.jpg
It's close to theoretical black but still falls slightly short. I am sure we can all picture this mask in regular black paint and immediately appreciate that normal black paint isn't black at all in the measurable sense.
Why is all this relevant? It's relevant because if we cannot get the idea of pure white actually not being pure white, particularly when comparing to images and photographs, whether on screen or off screen into our heads, we will never reach the correct answer. The subjects of the photographs those of us who build model ships look at are taken outdoors in natural environments. The lighting conditions change continuously and the colour perceived, whether by observer, artist or camera, is coloured (pun intended) by the ambient light. That is precisely why instrumented colour measurement requires the exclusion of ambient light and exclusive use of calibrated light sources.
If we cannot get the idea of black or white not actually being black or white into our minds and accepted as simple fact then we will forever be seeking contrast levels in photographs which simply aren't there. That does not mean we should use off-white on a model - that is to miss the point entirely - "pure white" paint IS off-white. That is the point.
As the risk of labouring the point, the following is the colour image of Prince of Wales with a rectangle of theoretical black and theoretical white added on top, straddling the black fore-funnel top and aft white upper funnel respectively. The "problem" and reason for my suggesting "realistic white" with LRV in the 75-80% range becomes obvious. The black paint recorded on the image measures out at 19% LRV - that's almost as light as the Snyder & Short AP507B paint. The brightest pixel on the white funnel top appears to be glaring and still measures out at 4% duller than pure theoretical white, whereas a darker pixel sampled from the white area measures at 65% LRV.
Attachment:
HMS Prince of Wales1.jpg
Again, that does not mean that a model should be painted in off-white. I say this because when we sample the shades on these images looking to prove or disprove relationships based on contrast, we absolutely must get it in to our heads that we are not actually looking for pure white or pure black in any of them. The percentage wrongness is large. Having understood that white paint in real life measured out at 75-80% LRV in 1942 (as written down at the time) which to be generous we will call 80%, and that MS4A (for example) measured 55% LRV, we are looking for a 25% contrast. If we cannot accept that white isn't actually white, we will be looking for a 45% contrast and that will never be found. Worse, looking for such a huge contrast will lead to spectacularly wrong assumptions to base interpretation from - e.g. by assuming the lightest shade was 20% darker than it really was to try to prise open the gap between white and anything else. You'll come a cropper at the dark end of course when one runs out of graduated shades and/or overlaps with real-life black - but that's how discussions end up going round in circles forever without ever reaching a conclusion.
The moral of the story? Eyes are not to be trusted. Trust instruments instead.
EDIT! There is now a blacker Vantablack that seems to be unmeasurable. Here's a link to a clip showing how something that's virtually theoretical black behaves under laser light.
http://www.iflscience.com/technology/heres-what-happens-when-you-shine-a-laser-on-the-blackest-material-ever-made/