International Conference on Light and Color in Nature 31st May – 3rd June 2016 Granada, Spain http://www.baobabeventos.com/#!lightandcolor2016/c1re0
Colour flowers Thomas Bangert thomas.bangert@qmul.ac.uk of http://www.eecs.qmul.ac.uk/~tb300/pub/LCNature2016.pptx
Sunlight Sunlight is mostly spectrally flat over the visible spectrum direct sunlight in the range of 450-700nm varies no more than 5% from the average There is very little UV light there is a notch at 430nm and a sharp decline below 415nm There is plenty of light beyond red (IR), but not flat above 700nm https://en.wikipedia.org/wiki/Sunlight#Spectral_composition_of_sunlight_at_Earth.27s_surface
Origin of the Flower? 200 million years old stems from the age of the dinosaurs dinosaurs (ancestors of birds) were the first animals with really good colour vision Flowering plants use birds and other animals (like insects) to reproduce. it is very important to attract the right kind of animal
Purpose of the Flower? to attract the right kind of animal which means advertising! often a reward is offered many animals (including humans) like sugar many animals need sugar to live sugar is expensive to produce rewards are kept to a minimum – preferably no reward at all which makes it important for the animal to visually discriminate – to see good from bad
Colour of Flowers Flowers do not have colour flowers only reflect or absorb light from the sun plants usually absorb all light they can use for photosynthesis – except a little bit of green Colour is purely the product of the animal’s visual system Colour does not exist as a physical property
Why investigate the Colour of Flowers? measuring the ‘colour’ of flowers is very easy understanding the visual system of animals is very difficult animals and flowers have co-evolved for hundreds of millions of years the way flowers absorb and reflect light comes from the way animals see animals dedicate a large part of their brain to vision – the visual system the way flowers reflect & absorb light can help us understand the visual system
Colour Vision dividing the spectrum Human Bird 4 sensors Equidistant on spectrum How are these sensors used? What information do they code? What does colour mean?
What we know about colour vision we can make sense of monochromatic light monochromatic light is not common in our natural environment flowers are never monochromatic the closer light is to monochromatic the more pure the colour appears
what we know from television b&w images need 1 number per pixel colour should be backwards compatible with b&w we can code colour with 2 extra numbers (channels)
opponent process we cannot see both red and green OR blue and yellow because colour is coded by 2 channels – 2 opponent channels for a total of 3 channels including b&w a positive number codes for one colour and a negative number codes for another you get negative numbers by subtracting
Rough Guide to Colour R-G Y-B we divide the spectrum up into primaries one for each sensor gives us 4 numbers we pair up the primaries we subtract the pairs gives us two numbers we assume colour is monochromatic light allows us to calculate where we are on the spectrum from 2 colour values R-G Y-B
… expressed another way x and y coordinates map to any colour
Measurement Instrument: X-Rite i1Pro spectrometer Range: 350-740nm Resolution: 3.3nm Precision: 3 decimal places designed for monitor calibration and print industry … works equally well measuring flowers, but has limited resolution white wild european daisy (bellis perennis) Spectrum from 376.666667 to 730.000000 nm in 107 steps 4.336, 4.607, 5.511, 7.069, 9.404, 12.475, 16.229, 20.543, 25.107, 29.655, 33.932, 37.739, 40.938, 43.502, 45.470, 47.003, 48.172, 48.941, 49.512, 49.984, 50.304, 50.528, 50.714, 50.864, 50.969, 51.075, 51.147, 51.184, 51.206, 51.219, 51.249, 51.262, 51.294, 51.333, 51.321, 51.329, 51.351, 51.357, 51.354, 51.347, 51.337, 51.348, 51.395, 51.403, 51.370, 51.345, 51.334, 51.340, 51.374, 51.396, 51.396, 51.407, 51.401, 51.340, 51.265, 51.266, 51.365, 51.457, 51.508, 51.526, 51.538, 51.526, 51.523, 51.560, 51.615, 51.631, 51.612, 51.609, 51.624, 51.627, 51.621, 51.618, 51.630, 51.643, 51.637, 51.611, 51.583, 51.597, 51.637, 51.638, 51.591, 51.554, 51.621, 51.781, 51.827, 51.708, 51.590, 51.533, 51.536, 51.543, 51.495, 51.409, 51.315, 51.273, 51.295, 51.313, 51.260, 51.159, 51.074, 51.071, 51.121, 51.170, 51.160, 51.115, 51.069, 51.047, 51.030
White Human Solution: reflect as much light as possible don’t worry about accuracy reflect over 90% of light deviation is up to 10% White is the absence of colour to an animal with colour vision How to present the best white? a human solution is very different from nature:
The Rose Rosa Spinosissima range: 460-740nm max deviation: 0.95% reflectance: 57% average deviation: 0.18% max deviation: 0.95% The Rose
The Rose Rosa Carolina range: 500-740nm max deviation: 0.4% reflectance: 50% average deviation: 0.1% max deviation: 0.4% The Rose
The Rose Rosa (garden hybrid) range: 440-700nm max deviation: 7% reflectance: 53% average deviation: 1.2% max deviation: 7% The Rose
Pelargonium Zonale (cultivated hybrid) range: 440-740nm reflectance: 60% average deviation: 0.07% max deviation: 0.74% Pelargonium
Primula vularis (cultivated hybrid) range: 460-740nm reflectance: 52% average deviation: 0.15% max deviation: 1.5% Primula [black center] range: 460-740nm reflectance: 0.35% average deviation: 0.03% max deviation: 0.14%
misc
Nature’s Solution to White can tell us a lot white flowers do not reflect as much light as possible but rather reflect light precisely white flowers reflect visible light exactly equally area of equality equals the visible spectrum of the animal flowers can precisely control absorption/reflectance spectrum of flowers is by design not accident very few flowers reflect light below 450nm most flowers reflect far beyond human upper limit of red – into IR consensus view that animals like birds see UV but not IR is probably wrong
area of visible spectrum is almost completely flat (equal energy spectrum)
visible spectrum of target animal far into IR No UV
Human Solution: divide spectrum into 3 regions – RGB colours are levels of each region adding light: RGB absorbing light: CMY Colour What is colour? nature has a very different answer than the human solution
HDR Television
Computer Monitors
Human Solution three primaries: Blue Green Red
Nature’s Solution to Colour very different to colour reproduction schemes invented by humans common underlying theme no indication of any primaries
Rose Rosa (garden hybrid) range: 440-700nm angle: varies low reflectance: 0.33% Rose
Pelargonium Zonale (cultivated hybrid) range: 440-700nm angle: high low reflectance: 0.32% Pelargonium
Sunflower you can have any colour you want as long as its yellow Helianthus range: xxx angle: steep low reflectance: 1.1% Sunflower you can have any colour you want as long as its yellow
Poinsettia colour is not great by default Euphorbia Pulcherrima range: xxx angle: low low reflectance: 1.85% Poinsettia colour is not great by default
How Nature Defines Colour one area of absorption – equal and extremely high absorption one area of reflection – same as white colours change by a spectral shift of absorption/reflection boundary range is yellow to red no such thing as green or blue flower shift is circular, with red to purple range can be coded by one number (dimension)
colour by spectral shift reflects like white range is yellow to red absorbs like black
purple indicates circularity
Conclusions precision is due to difference measurement by sensors white when colour sensors exactly negate each other colour is the contrast between precise absorption and precise reflection indicates 2 sensors for each area current colour reproduction technology is woeful in comparison to flowers
comparing white: the orchid with an apple 5K display
Questions? http://www.eecs.qmul.ac.uk/~tb300/ http://www.eecs.qmul.ac.uk/~tb300/pub/LCNature2016.pptx References Poynton, C. A. (1995). “Poynton’s Color FAQ”, electronic preprint. http://www.poynton.com/notes/colour_and_gamma/ColorFAQ.html Goldsmith, Timothy H. (July 2006). “What birds see”. Scientific American: 69–75. http://www.nature.com/scientificamerican/journal/v295/n1/full/scientificamerican0706-68.html Chittka, Lars. FReD floral reflectance database http://www.reflectance.co.uk/ Bangert, Thomas (2008). “TriangleVision: A Toy Visual System”, ICANN 2008. http://link.springer.com/chapter/10.1007/978-3-540-87536-9_96 Hurvich, L. M., & Jameson, D. (1957). An opponent-process theory of color vision. Psychological Review, 64(6), 384-404. Wyszecki, G., & Stiles, W. (1967). Color Science. John Wiley. MacAdam, D. (1985). Color Measurement. Springer. The Colors of Flowers. (1882). Scientific American, 13, 5204-5206.