Colour Management in the RAW!
Monday 29th April saw a fascinating evening with two presentations from Professor Terry Hewitt. His work specialism is scientific computing and supercomputers but he is also passionate about photography and it was in this field he visited the club to inform everyone about colour management from camera through to printer in his first talk and then about RAW images and processing in the second.
Terry started the first talk by comparing a projected digital image with the corresponding print and noting the differences as well as pointing out the various places where the difference could have occurred. He explained we all see colours differently and displayed 6 colour blocks, one by one, asking what colour we would call them. They were all red…… but all different! We see colours differently and the light we view them in can change perception of colour. An international committee attempted to define all the colours in terms of their component colours (how much each of red , green and blue). There are various standards e.g. sRGB and AdobeRGB and they are all different in the variety of colours they can reproduce. sRGB cannot produce as many colours as AdobeRGB but the range of colours printers can produce is worse than sRGB! In order to get anywhere near a decent print of a colour image there has to be an ICC (International Colour Consortium) profile which helps to convert a colour that can’t be produced on a printer to one that can. By this means an sRGB or AdobeRGB image can be printed but there are bound to be differences to the colours captured at source. Even then the ICC profile conversion can be done in different ways called the “rendering intent” – perceptual, relative colormetric etc. Terry showed how perceptual intent changes ALL the colours yet most people recommend using it.
How then do make the best of these problems? Terry showed that a logical approach can help. He said you should start by calibrating the monitor with a good calibration device. There are free ones but he never found them successful so he uses devices such as the XRITE COLORMONKI and the Datacolour Spyder and he then demonstrated using the former. He stressed that you need to calibrate regularly as devices change over time. If you leave it too long there will be a big change and then all your old work will look rather odd! You should also then calibrate the camera, although not many people do, and then calibrate the printer itself. All the reasons and pitfalls were fully explained by Terry and there were lots of questions. It was clear that he could have spent the whole evening on colour management and the audience would have been very happy for him to do it.
In the second half Terry concentrated on RAW files. He explained that RAW captures the “image” before anyone has seen it (in reality the RAW file is not actually an image but a set of numbers and we only see a processed version of it). The camera sensor has red, green and blue pixels (in fact twice as many green pixels as any other individual colour as human eyes are more sensitive to green light and the sensor is engineered to mimic this) and the camera’s processor establishes the individual pixel values (how strong was the blue light which fell on a blue pixel). It then “de-mosaics” them (or interpolates them) to get a red, green and blue image. This can then be further processed into a JPEG image. In order to “show” the image to us the processor will always show a JPEG on the screen, even if you only shot in RAW.
When the RAW file is imported onto a computer with software for processing RAW files (such as Adobe Camera Raw) the software creates an intermediate file to display on the screen so you can see what you are doing. As you make adjustments using the software the displayed image changes appropriately BUT the RAW file is not changed at all. Instead an XMP file (or sidecar file) is created logging all the changes. This means it is a non-destructive edit maintaining the integrity of the original RAW file. One consequence is that if you delete the XMP file or move the RAW file by itself to another location then you will lose all your changes. When you open the adjusted RAW file in (say) Photoshop then all the changes you applied are used to create the Photoshop file leaving the RAW and XMP files as they were. With the Photoshop file you cannot go back to the original RAW state; you would have to open the RAW file again.
Trevor then took us through some examples of RAW processing in Lightroom which has a RAW file processor built in. Just like with Adobe Camera Raw and Photoshop, Lightroom only makes the changes permanent when the file/image is exported. Given that software companies can come and go Trevor converts all his RAW file to DNG format as this is almost universal and stands a better chance of being able to be processed by future software.
As you can probably tell, this was an evening full of fascinating information which may sound quite technical but Trevor’s enthusiasm, humour and clear explanations made it a thoroughly entertaining evening as well. His talk generated lots of questions and his answers were only limited by time. It would have been quite easy to have the full evening on either of the subjects and everyone would have still wanted more! The enthusiastic applause testified to how much everyone had enjoyed it.