Table of Contents
color / colour in photography
see also:
Introduction
- we perceive various wavelengths of light as colour - see electromagnetic radiation
- some people cannot perceive all wavelengths - see color blindness / color vision deficiency
- how a camera or RAW conversion software “develops” a RAW file into an RGB image we can view requires:
- a selection of color space for the RGB model (eg. sRGB, Adobe RGB or ProPhotoRGB)
- a color temperature variable which is used to adjust the “white balance” to neutralise or enhance the color of light hitting our subject
- a camera picture profile which adds control to the tonality of the image
- if you are planning on printing your photo, when you decide to edit your images, ideally you should ensure your computer screen and your editing software is calibrated to your printer, printer paper and ink, so that what you are are seeing on screen more closely represents what your print will look like - see color calibration
Digital representation of color values
- a given colour can be digitally represented by several systems:
- HSL and HSV
- these models were designed by computer graphics researchers in 1978 and derive from the analog HSL system invented in 1938 to allow color signals to be sent and be compatible with B&W TV's which would only read the L channel and hence formed the basis of all TV broadcast standards as well as composite video
- HSL = hue from 0 to 360deg on the color wheel, saturation from 0-100%, lightness from 0 to 100% (black to white and a fully saturated, bright hue having lightness of 50%)
- to ascertain a mean value of a range of hues, one must use mean of circular quantities
- HSV is same as HSL except L is replaced by Value (or “brightness”) where 0% is black and 100% is brightest tone of the hue (NOT white unless saturation is 0%)
- note:
- luminance is a measured value such as one would get from a light meter
- brightness is how we perceive luminance and in image editing it is the darkness axis of a hue and has a percentage value with 0% being black and 100% being the brightest tone of that hue (this would become white if saturation was also 0%)
- lightness or luminosity is the perceived brightness of an object on a scale of black to white when compared to a purely white object and is what is used in luminosity histograms and in luminosity masking
- chroma is the colorfulness relative to a similarly illuminated purely white object
- saturation is the colorfulness relative to the brightness
- in the HSL and HSV models, to avoid values of chroma which fall outside the gamut, the term saturation is used which is the chroma divided by the maximum chroma for that lightness
- see wikipedia
- RGB device-dependent additive model
- used in most image editing software as well as computer screens, and image files
- this is based upon the human eye's 3 types of retinal photoreceptor cone cells - red, green and blue
- 24 bit color has 8 bit value range of 0 to 255 for each of red, green and blue channels
- 100% of each color should give pure white
- in image editing software, if you blend images using “screen mode” this will give the additive effect and produce white if you end up with 100% of each color
- see wikipedia
- CYMK device-dependent subtractive model
- used in printing and is derived from historical color printing processes, long before computers were invented
- 100% of each color should give pure black as each ink is absorbing colors of light and preventing them from being reflected
- cyan, yellow, magenta and black (black is added to give better blacks as CYM inks alone produce a muddy result that takes longer to dry at 100% each and is represented as “K” to stand for the “key alignment plate” used in early printing and to avoid using B as this was already used for blue)
- Y'CbCr
- a bandwidth-reduced digital video transmission model based upon the analog YUV system
- Y′ is the luma component and CB and CR are the blue-difference and red-difference chroma components
- Y′ (with prime) is distinguished from Y, which is luminance, meaning that light intensity is nonlinearly encoded based on gamma corrected RGB primaries
- see also Wikipedia YCbCr
- CIELCH
- perceptually uniform equivalent of HSL
- the CIEXYZ is a color space to model human vision based on the 3 types of retinal cones to give X,Y,Z values
- to achieve uniformity one must transform XYZ to LUV (Lightness and two chromaticities u,v) but this like LAB is not intuitive, hence the u,v are re-mapped to CIE LCH (Lightness, Chroma, and Hue)
- see also:
| HSL cylinder | HSV cylinder |
|---|---|
 |  |
| image editing color pickers use a 2D HSV square section |
Tinting, shading and toning
- the base hue has a saturation of 100% and a brightness of 100%
- lightness or perceived luminosity is determined by BOTH saturation and brightness
- adding white to a hue is known as tinting and reduces saturation without reducing brightness
- adding black to a hue is known as shading and reduces brightness without reducing saturation
- adding gray to a hue is known as toning and reduces saturation and brightness
- a hue can also be de-saturated by mixing it with a hue from the complimentary hue on the color wheel
RGB color spaces
- when processing camera RAW files, one needs to select which color space to convert into
- the gamut is a term used for the entire range of colors available within a color space
- RGB model is device-dependent and this means that it is referenced to a defined color space, the common ones used are:
- standard RGB or sRGB
- this is the most common color space and is the one we are usually using when viewing images on computer screens and was defined in 1996 by Microsoft and HP but does not fully cover the CYMK gamut but provides a “lowest common denominator” capability so that even old computers can display these images
- Adobe RGB
- this was designed by Adobe in 1998 and has a wider gamut than sRGB allowing display of far more colour saturation tints and tones, particularly greens and blue-greens and is better for printing as it covers the entire CYMK gamut but will not display well on sRGB devices
- ProPhoto RGB
- has an even wider gamut than Adobe RGB and was created by Kodak with the 255 Green and 255 Blue being outside human eye gamut
- most professional photographers will use the following workflow system:
- camera set to sRGB for viewing jpegs in camera or distributing the jpegs (this does NOT affect RAW files)
- do RAW file conversion using Adobe RGB as this will allow wider gamut and better saturated greens and reds
- many also use a color profile too, such as XRite to ensure your RAW images from different camera systems and lighting set ups generate similar colors.
- ensure your RAW editing system is optimised for Adobe RGB (eg. your monitor which also should be colour calibrated)
- when saving files as TIFF or PSD, save in Adobe RGB
- when saving files for web such as in JPEG, ensure you tick BOTH:
- convert to sRGB, and,
- embed color profile
Which color space to use?
- if you are shooting RAW files with your camera then your camera color space setting is irrelevant as this is ONLY used to create the in-camera jpeg images and given these are largely used for viewing on the web, you may as well leave it at sRGB
- RAW conversion software is a different matter though!
- if you are going to do a lot of editing, use ProPhoto RGB during the editing stages and then output using the “Perceptual” option to sRGB jpeg if computer display only, or Adobe RGB TIFF if printing (no current printer or computer screen supports ProPhotoRGB but modern printers with wider gamut thanks to extended range of inks will support Adobe RGB's gamut)
- if you are not doing a lot of editing and just want to print to highest print rendition, then use Adobe RGB but for best results you need to use a monitor which supports at least 99% of the Adobe RGB gamut
- otherwise, if you are only creating images for web, computer screen or smartphone display then use sRGB