Light travels in straight lines and when it falls on the camera’s sensor this is known as a linear capture. The more light the stronger the response. Unlike our eyes, the camera’s sensor doesn’t cope well with light. We are able to ‘compress’ light so that when it is really bright the effect isn’t as obvious. The camera has to process the light information in order for us to see the same image that we see first hand. Unprocessed images would appear very dark compared to their processed equivalent.
Simulating a linear image
The above image is a normalised image of a charity function. The camera has processed the image into a JPEG. I chose it for this exercise.
Before this image had been processed it would have looked something like this image (below).
The histogram illustrates that most of the tonal values are squashed up to the left-hand side, with very few pixels at the highlights end. In order to bring this image to what would be considered to be a normalised version and compensate for the high contrast, a gamma correction curve, as shown below, needs to be applied. This involved moving the curve to the left. It took a while to achieve similar tones to the camera’s
This correction produced the below image.
The processed image’s histogram illustrates the affect of gamma correction. The peaks have moved towards the centre away from the shadows and into the mid-tones. The processed image is much lighter and brighter by lifting the dark tones.
I usually use Raw to take photographs because of the ability to adjust different aspects of the image before it’s processed and as a safety net in case I misjudge my camera settings. However this exercise has taught me about what has or hasn’t happened in the camera depending on whether my output is RAW or JPEG. One of the disadvantages of RAW is that it takes up a lot of memory space, where as JPEGs are smaller because of the processing. This is a technique I will try when I take photos in JPEG.
However there is a payoff with this technique. The strong curve needed to lighten the image has meant ‘noise’ has started to appear in the lightened shadow areas and the definition isn’t as sharp.