A Camera Profile

August 27, 2010

You don’t hear much about colour management these days, thank goodness, largely because computer operating systems and software applications manage to take care of it semi-automatically. The major point for photographers is coordinating the input and output equipment and media so that the general balance of colours and tones stays the same throughout. When you open a digital image file in an application like Photoshop or Lightroom or Aperture, you want to have the confidence of knowing that what you see on the screen will reproduce that way in a print, or in a brochure, or wherever.

By and large, this all runs smoothly most of the time, with few demands on our attention. But once you begin to look at the finer points of colour balance, you’ll notice a large set of variables that sits apart from the computer-monitor-printer group – the cameras themselves. The sensors, and the way the information they capture is processed inside the camera, naturally vary. In fact, the way in which the camera manufacturer chooses to process the data from the sensor has a considerable effect. Three things have to happen, and each of them is open to interpretation. In this case, the manufacturer’s interpretation. They are:-

  • The colour filtering of the captured mage by the Bayer screen in front of the sensor has to be decoded. Each pixel in the sensor is covered with one colour – red, blue or green – so a full colour interpretation needs to be made.
  • A tone curve has to be applied to give the linear, very dark, captured image (which we normally never see). This makes it look as the eye expects it – again, a matter of interpretation
  • The recorded colour need to be interpreted so that they too, look ‘normal and expected’.

Because there is room for interpretation, images of similar scenes from different cameras tend to have their own ‘look’, identifiable by most people, but at the same time not so easy to describe. In fact, however and wherever you look at a digital image, whether it’s a JPEG, TIFF or Raw, and whether you’re looking at the display screen on the back of the camera or on your desktop monitor, you are looking at a processed image. Now, if you would prefer to have a slightly different (or even very different) ‘look’ to your images, this is fairly simple to do. It’s a way of over-riding the manufacturer’s ideas on the subject, and means that you can have a tailor-made look for your camera and to accommodate your particular likes and dislikes. Also, say that you have two different models or makes of camera, yet want them to deliver the same kind of ‘look’. That’s possible, too.

The tool that’s used is called a profile, which is a small digital file that instructs software (like Photoshop) to shift certain colours in certain ways. The shifts may be small – a red slightly towards orange, a green slightly less saturated, and so on – but the overall effect creates the ‘look’. Now, if you shoot raw, and I think most of us do for the obvious advantages of having more control and more image information, the profile doesn’t have to be applied in the camera, but instead can be called up when you start to process images using a raw converter. Personally, I use Photoshop, which uses Adobe’s Camera Raw, and the list of available profiles is in the Camera Calibration window, in the form of a drop-down list, as shown here…

untitled-22

The differences are not major, but for anyone committed to get colour exactly right, they are significant…

untitled-33

In the Camera Raw window above, you’ll see that there’s one profile called Sony DSLR a900_cloudybright, and this is one I made for this particular camera. The tool for doing this is smart, simple, and free. Called the DNG Profile Editor, it’s available for download and use. You can get it from the Adobe site, HERE.

You’ll need an Adobe ID (you can create one on the spot) and log in first.

The user guide and other documentation from Adobe Labs is HERE

… and note that it says that this tool is ‘intended for advanced users’. Don ‘t let this put you off, but it is a reminder that we’re looking at things more meticulously than would a regular amateur. One simple way to begin is to use one of the default profiles in the processing software, such as ACR 4.4 in Photoshop or Lightroom. This is called a Base Profile, and if you would simply like to tweak it to make certain colours more or less saturated, or brighter, darker, or shifted, it’s straightforward. Read through Tutorial 1: Getting Started and then go to Tutorial 3: Using Base Profiles. Open up the base profile and adjust it as you like and as described.

A more accurate way is to use a standard colour target. In fact, the standard colour target is the GretagMacbeth ColorChecker. It’s unfortunately a little expensive – around £50 – because the colour patches are very carefully printed, but it is extremely accurate. The camera profiling is discussed mainly in Tutorial 5, and the instructions there are clear and straightforward, so there’s no point in my repeating them here. You photograph the chart in even lighting, and the DNG Profile Editor will analyse it. Because the color patches are measured and known by the software, the profile is created by measuring the shifts that are needed to bring the camera’s rendering of the chart into line. Very simple, very effective.

untitled-41

untitled-51

There remains the question of what kind of lighting in which to photograph the chart. Strictly speaking, you should have a profile for every specific lighting condition, and indeed, that’s what many professionals do. If you were working in an artificially lit space where there were doubts about the kind of lamps, you could shoot the chart and use a profile from that to balance the colour for the rest of the shoot. For general purposes, however, I simply shot the chart for each of my cameras under the one fairly predictable kind of outdoor lighting – cloudy bright. This gives me a known starting point.

Tags: , , ,

Comments are closed.