Black and white photography: key considerations

As we’ve discussed in a former post, Black and white photography is now more associated with the fine art medium than is color, and there is no surprise there. It has become a means of representing and transferring one’s own and unique inner vision of the external world to the others through a medium that is limited in characteristics, rather than just portraying things the way they are.

By its nature black and white photography differs in some characteristics from color photography. Some differences are vital while others are more subtle. Today we look at some of the main characteristics of black and white photography, and key considerations to keep in mind when producing black and white imagery.

Tonal compression, brightness, and global contrast

Photographers usually manipulate the tonal compression, brightness and global contrast of their photos to produce more appealing results that better represent their intentions. Generally these 3 characteristics can often be pushed to either extreme more acceptably and freely in black and white photography than in color, while still maintaining the integrity of a photograph.

Tonal compression is the process of limiting the tonal range (the range of tones that make up a photograph from the brightest tone to the darkest) of an image by clipping at either one or both ends of the scale, thus producing an image with less tonal information. This process is usually more acceptable when clipping image information at the shadow end than at the highlights, mainly because of the popularity and acceptance of the silhouette.

Global contrast, as opposed to local contrast, affects the entire image while maintaining the original shadow and highlight end points of the scale. This is generally achieved by applying an S-curve or a reverse S-curve to the image.

Film characteristics

Every film type produces a different black and white result. These characteristics vary in terms of texture, contrast, and grain among other things.

What people refer to as grain is actually a much larger structure than what an individual grain really is, for individual grains are in fact very hard to pinpoint at normal viewing distances.

One of the main advantages of film over digital sensors is that film reacts to light more like the human visual system. This means that film is more adept at handling wider dynamic ranges and preserving really bright highlight details as opposed to the digital sensor.

The digital sensor

Although the majority of digital photography is represented in color, the digital sensor actually responds to amounts of light rather than portions of the visible spectrum (and so does film so to speak). A process of color separation is then conducted to reproduce different colors. This process is generally handled by red, blue and green filters positioned in front of the digital sensor.

The almost universal approach of reproducing color is handled by the Bayer array, which is a mosaic filter of red, blue and green patterns with green patterns having twice the number of the red or blue in an attempt to simulate the human visual system which responds greater to the yellow/green parts of the visible spectrum.

To generate a black and white image later, these colors need to be de-saturated in order to re-produce a grayscale representation of the image. The main reason for introducing color and then subtracting it is mainly for the benefit of greater latitude in post-processing, giving the photographer greater freedom in representing their black and white image by more precise tuning of different specific hues.

Digital sensors are much less forgiving than film in their response to light falling on each photo-site. This means that clipping, either at the right end or the left end of the histogram, is more likely to occur especially in scenes with high contrast between highlight and shadow regions.

Generally you can only restore little lost details before reaching a certain limit where there is not much left to extract anymore, or running into the problem of introducing noise at the shadow parts of your image.

Digitally interpreting light

The digital sensor, as we’ve mentioned differs in its response to light from both film and the human visual system. It calculates light striking each photo-site and generates a signal corresponding to those calculated amounts. This is a very basic and linear way of interpreting a specific scene, and yields unnatural results compared to the way we actually perceive reality. Both film and the human eye have a way of realizing high dynamic ranges and adapting to varying lighting conditions by somehow compressing information which yields to brightening of the shadow areas and gentler tailing-off towards the highlights. Film’s response to increasing exposure slows down. So at the right hand of the histogram, film responds more sluggishly to increasing brightness than it does at the mid-tones.

To reproduce the same results with the digital capture, a Gamma Correction can be applied. If you shoot TIFF or JPEG the camera actually does apply such correction to provide a more natural image by creating a somewhat gentler roll-off in the highlights, and to some extent avoid the sudden break point and banding which occurs in the highlight region when an area shades from bright to pure white.

Digital clipping

Due to the linear nature of the digital capture, highlight clipping (loss of information at the brightest areas) is likely to occur especially in scenes of high dynamic ranges and great contrast. This can be more troublesome in black and white photography than in color where tonal gradation receives more attention due to the lack of color distraction.

Another main advantage of shooting color when the end result would be a monochrome is that red, green and blue channels respond differently to light. This means it’s highly unlikely for all three channels to clip information at the same point. Thus, when attempting to recover image details (which is a popular method of RAW image convertors), lost information of one channel can be restored from the other two. Even lost information of two channels can sometimes be restored from the third channel as long as there is some data to retain.

The following 2 images illustrate how clipped image details can be retained using the recovery slider in Adobe Camera Raw:

Noise

Noise is typically caused by multiple factors. Things such as higher ISO settings, long exposure times, along with sensor design contribute much to this unpleasant phenomena. One of the advantages of black and white photography compared to color in this area is that only luminance noise needs to be concerned about. Color noise, since color is absent is one less thing to worry about, which is something that makes the overall final effect all the while somewhat less disturbing.

Bit-depth & banding

Banding is the non-continuous, non-smooth tonal gradation that might be present in an image. Where there is banding, there appears to be sharp breaks of tones which should shade continuously and subtly.

Banding is a result that is very much dependant on image bit-depth. Often, final result images are presented as 8-bit color. This means that 8 bits of data are used to represent each channel, which only offers 1 of 256 levels to reproduce a given tone in an image.

High end cameras allow for shooting in 12 or even 14-bits per channel (4,026 or 16,104 levels for tone representation), which is much better. However, if you’re shooting in JPEG those 12 or 14 bits will have been mapped down to 8-bits for storage after having passed the camera’s processing stage.

If you’re shooting RAW, on the other hand, the 12 or 14-bits will be converted to 16 bits for further editing in 16-bit image mode with photo-editing software such as Adobe Photoshop (provided, of course, that you open and process the images in 16-bit image mode). This way, the risk of image banding is very much reduced, and sometimes even diminished. Processing 8-bit images, on the other hand, is more likely to introduce banding which is apparent as combing (thin spikes) along the histogram.

While these artifacts might not be disturbing at normal viewing distances, banding is definitely unacceptable in full-blown, high-quality black and white prints especially with the absence of the color factor which would perhaps lessen the obviousness of such effect otherwise.