The sky RGB values in this image are roughly 38,39,60. Clearly Blue is the dominant component, but Red and Green are not insignificant. To test this further I created a powerpoint slide with three blocks of colour Blue, Yellow and Red. The Blue was created with RGB = 0,0,255, Red with RGB = 255,0,0, and finally Yellow with RGB = 255,255,0.
From a computer monitors point of view, these are the pure colours! With a properly calibrated screen and printer I could now print these and get a reference for further tests. However, what would be the point, pure colours such as these are the creation of a machine and not natural light.
All of this is a preamble to my set of primary colours, none of which remotely match the colours above, but to my eye at the time of creating the image were pretty close to what I would see as a base colour
In each case I have shot at a -1/2, 0, and +1/2 stop variation from the exposure suggested by the camera. The blue is from the side of a container, the red was a tarpaulin, and the yellow is from a post box. To improve this exercise I should have established a neutral white balance and exposure at time of shooting using a grey card. In the images in this project and the next I have used the same white balance in Lightroom (cloudy) as all shots were taken in shade under a cloudy sky.
In the case of the blue and red the middle exposure has the best rendition of the colour I experienced at the time. In the Yellow sample the camera has darkened the colour as it tried to push the brighter yellow towards a neutral grey. In this case the +1/2 stop shot is the best, although I think +1, or even +3/2 would have been better. I have done a few weddings for friends and found from experience that bright colours, especially white need to be over-exposed or heavily adjusted in Lightroom, thanks goodness for RAW.
The best primary colours I have achieved recently has been when shooting the autumn foliage, sadly I did not take equivalent over/under shots, now it is too late, the leaves have fallen:
No comments:
Post a Comment