Raw Image Basics

All digital cameras shoot RAW images. Normally RAW images are not seen and have to be reconstructed for realistic viewing on the camera's inbuilt screen or on an external computer monitor.



Above left represents a theoretical 26 by 38 pixel camera sensor. Each element responds only to the intensity of light and therefore only produces a black and white image.

To enable the sensor to record colour, each pixel is covered by a coloured filter of either red, green or blue (RGB). Therefore, for each point in a subject to be photographed the camera sees either an amount of red, blue or green light. There are twice as many green sensors because our eyes respond mostly to green light.

When you fire your camera's shutter each pixel, which contains a light sensitive circuit known as a "photosite", is opened and collects and stores the amount of light falling upon it. Once the exposure finishes, the camera closes each photosite, and assesses how much light fell into each. The relative quantities of light (number of photons) collected by each photosite is then converted into a number.

Therefore, the following photographic subject would appear to the camera as illustrated.




The image seen by the camera is known as RAW because it has not been processed or 'cooked' in any way. And because each pixel only records a third of the available light, it is much duller. Try squinting and the RAW image looks more like a dull version of the original.

Once exposure has finished, the camera uses a in-built, program to reconstruct the original image as a JPEG file. It does this by looking at the surrounding pixels for each pixel. Most compact cameras only produce JPEG images – the RAW image is discarded after processing. However, more and more compact cameras and all DSLR cameras retain the RAW image for processing outside of the camera.



RAW image processing reconstructs an image by looking at the state of adjacent pixels for each individual pixel. e.g. If a red filtered pixel has recorded a level of light and is surrounded by green filtered pixels which have also recorded levels of light, but adjacent blue filtered pixels have seen no light, the red filtered pixel is concluded to have seen yellow light.

The more adjacent pixels taken into account, the more accurate the reconstruction becomes, but the longer it takes.

And this is why computers, which are more powerful than cameras, can produce faster and more accurate reconstructions and why, especially with low end cameras, you have to wait before you can take another photograph, because the camera is still busy reconstructing the last image.

One of the most important benefits of working with RAW images is that unlike processed images such as JPEGs, RAW reconstruction will become more realistic as processing software such as included in Adobe Photoshop continues to improve.


Leave a Reply

Your email address will not be published. Required fields are marked *