Anyone who grew up in the UK will remember Daz, whiter than white advertisements; perhaps they still even persist. White was King when it came to washing powders and detergents. But what is white exactly? Silly question it may seem, but in digital photography it is far from trivial… here, too, white is king.
You have enamel-white, pearl-white, snow-white, ivory … perhaps they’re not strictly speaking white, but if you saw each of them contrasted with black, you’d probably almost swear they’re “white”. But they’re not. Compare side-by-side two pages of white printing paper from two different producers and after careful examination, one will likely be whiter than the other. But they’re both white, right…?
White Balance (WB). What is it, and why do some photos look warm and cozy, while some make our subjects look like a recently thawed Ice-Queen, when the flash was turned on? This concept is one that I think is mostly significant with digital photography versus film, though it still plays a role in film-based development none-the-less. My focus is on digital photography in this post however.
Temperature of light and its effect
If you have a room in your house that has 2 different types of lighting, say one is tungsten while the other is fluorescent, I would like you to try a very quick experiment. Turn off all of one type; let’s start say with the fluorescent. You will notice a distinctly warm, orangey tinge to the room. It’s like candles or a blazing fire… it’s a warmer, cozy, more romantic feeling. Immediately turn all of these off and turn on only the fluorescent. The contrast should be very clear. It’s freezing cold looking. Items and people have a blueish tinge to them and this is how not to set up a romantic date scene. That all said though, whatever way the room goes in terms of colour temperature, your vision is not phased. You can see it all fine and after a few moments, you’ve adjusted and all is right again. You don’t normally notice the difference because your eyes-brain adjust and provide you with a picture that is adapted to the environment. It doesn’t put colour where it isn’t, but if there was a white object in the room under any temperature, you could point it out based on its relative hue (or rather complete lack of it) with the rest of the objects in the room.
So what? Believe it or not, digital sensors cannot make that distinction. They cannot tell you what is white, grey or pure black – that is, any colour that has no intrinsic hue whatsoever. They have to guess what the colour temperature of the room is in order to digitally map out the natural colours it produces for the final image.
For example, if you told the camera that you’re taking a photo in a room that has only fluorescent lights in it (i.e. it’s cold), it will interpret the image it saw accordingly, compensate for the blueish hue and map out the colours in such a way that they reflect the natural colours and items that were white, are in-fact white. Human skin tone is highly sensitive to this colour temperature in photographs and for photos with people in them, it is probably this factor more than anything that will make a picture look natural, or not.
White balance setting in digital cameras
All digital cameras today come with some WB setting feature. At the very basic end of the spectrum, it will just provide you with automatic – the camera takes in a view of the picture to be taken, decides on the temperature of the scene and adjusts its colour as it finds appropriate. This is fine, but auto white balance in general sucks. The ability to accurately determine the appropriate temperature setting varies greatly between models and manufacturers also.
You can usually set the scene for white balance using presets that are depicted using a series of icons for conditions you’re shooting under. There will usually be a picture of a cloud, the sun, lightning bolt (flash), shade, a fluorescent tube, and an tungsten bulb. These all point to an preset approximation for the temperature of a given scene and simply perform a pre-specified colour shift to compensate and help to ensure the whites are white.
One experiment to try to bring this white balance concept a little clearer into focus and see how drastically it can affect your final photos, is to manually set your camera’s white balance to the “opposite” of the conditions you’re taking under. So, if it’s a tungsten-lit room, set the WB preset to ‘fluorescent’. You’re telling the camera the colours out there are cold, but in fact they’re very warm. The resultant photo will likely have their colours shifted very noticeably towards red and orange because the fluorescent filter would be compensating as you told it to. Then take the same photo with the more accurate WB setting to compare the difference. And for the sake of testing your camera’s auto-WB function, see how it determines the scene and pick which is most appropriate.
The point here is simply that the WB setting will significantly impact the colour of your photos. It can make a potentially great shot look pathetic because afterall, colour is what it’s all about and if you get it wrong it ruins the final result. This is most apparent with human skin tone… even slightly off, it makes people look a little unhealthy and unnatural, which is clearly not a desired result for say, portrait photography.
Manual white balance setting
Depending on your camera model, you will have the option to not only set automatic and use preset white balance settings, but also specify the temperature exactly. The temperature is measured in kelvins and oddly enough, the higher the temperature value, the cooler the colour. So that 5000K+ would be cooler green-blue, while 2000~3000K would be warmer.
By using a piece of white/gray card such that it reflects the light source of the room to completely fill the fram, using the WB function the camera will calibrate and adjust colour temperature accordingly. A quick search online will reveal much more details about how to set this up, but it isn’t difficult to do and you will find your photo’s colour reproduction improve dramatically.
Camera RAW and white balance
If you’ve read my previous article on Camera RAW, you will know that whichever setting you make for your white balance when you’re taking photos in RAW mode, wont actually make a difference to the end result. Why? Because with RAW you’re storing the data that hit the sensor and nothing more; this does not include any post-processing and WB adjustments fall under this category. So is this an advantage? Definitely! It means that though your previews from the photos that you take in RAW mode might be slightly “off” in terms of true colour reproduction, since it’s RAW it doesn’t matter – you can set the best white balance you think fits in your post-processing workflow with no detrimental effects on the original photo. This would be very challenging to do accurately with a JPEG image.
An example
Below I have included a photograph that will highlight the difference in a photo’s white balance adjustment. The original was taken using RAW and the only adjustment in any picture was the temperature. On the far left is how I remember the scene, in the middle is the “daylight” preset, and on the right is the “cloudy” preset.
- Left: my interpretation; Middle: “daylight” preset; Right: “cloudy” preset
You should hopefully be able to see the significance that getting the right colour temperature setting can make to the natural look of a picture. I believe daylight makes it just a little bit too cold, while cloudy over compensates and there is just too much orange/red. Of course, as with any setting, you can make an artistic statement with exaggerated use of the white balance setting to create very warm scenes, and conversely very cold, barren landscapes for example.
There is no right or wrong…
{ 0 comments… add one now }