Caponigro on Colour
Colour Analysis
What is colour theory? What is its vocabulary? And what are its applications?
In his essay, “Colour Theory”, artist and author John Paul Caponigro answers these questions. And in doing so, he shows how — by properly understanding colour theory — professionals who work with colour can sharpen their perception of colour and describe colour more accurately.
To better see color within an image it can help to abstract it. By de-emphasising an image’s representational qualities you can more easily direct your attention to the colour relationships within it. In short, you can see them better because you’re not distracted by other concerns.
Using Photoshop, there are many ways to modify an image in order to better reveal it’s colour structure.
You can blur an image. Blurring an image reduces detail so that you can more easily see the basic composition and the colour relationships within it, without getting hung up on the details. (Duplicate the background layer and apply the filter Gaussian Blur.)
Blur an image substantially and you can reduce an image to a field of colour. Compositional elements are significantly downplayed, leaving pure colour. (Duplicate the background layer and apply the filter Gaussian Blur with a stronger setting.)
Pixelate an image and you can reduce an image to blocks of colour. Composition is eliminated while contrast between colours is more pronounced than in a flat field of colour. (Try Filter>Pixelate>Mosaic. This works best for lower resolution files or copies of files.)
Average an image and you can reduce all the colours in an image to one. (Duplicate the Background Layer and apply the filter Average — found under Blur.) This often confirms the dominant colour in a composition. Well balanced colour photographs containing a variety of hues tend to average towards gray.)
While accents and other important colours can also be used, the dominant colour is an excellent choice to further analyse colour relationships in an image using Blend Modes. With this technique you can see the variety found in the separate components of colour within an image — Luminosity, Saturation, Colour and Hue. (Change the Blend Mode of the averaged layer to the desired colour component.)
With a blend mode of Hue, all values in an image will be driven to the same hue. This will help you see variety in luminosity and saturation more clearly.
Original image.
Blurred to reduce detail.
Blurred to create a colour field.
Pixelated to create a field of colour with greater separation.
Averaged to find the dominant colour.
Using the averaged image and a blend mode of hue.
Using the averaged image and a blend mode of saturation.
Using the averaged image and a blend mode of colour.
Using the averaged image and a blend mode of luminosity.
With a blend mode of saturation, all values in an image will be driven to the same saturation. This will help you see variety in luminosity and hue more clearly.
With a blend mode of colour, all values in an image will be driven to the same hue and saturation. This will help you see variety in luminosity (the tonal structure) more clearly. (This variant is often the most useful as it is the easiest to interpret. With repeated analysis of many different images, you’re likely to note that images with less variety in hue and saturation — particularly neutral ones — will require more luminosity contrast to have impact. By the same token, images with a great deal of variety in hue and saturation will often appear overly harsh with excessive contrast.)
With a blend mode of luminosity, all values in an image will be driven to the same luminosity. You’ll eliminate contrasts in value which will help you see variety in hue and saturation more clearly.
This type of analysis will better reveal the colour relationships at work within an image. You can use the information you’ve gathered by analysing colour relationships in an image not only to better understand it but also to make predictions about how you might improve them.
Increasing contrast in one or more of the elements of colour (hue, saturation, luminosity) will increase separation in a composition; conversely, decreasing contrast will create greater unity.
Typically, well structured images use a large amount of contrast in one colour component, a medium amount of contrast in a second component and a small amount of contrast in a third component.
