Photography is framing

This feature is about what I think photography does and how it does it in terms of technique. I originally wrote a forum post elsewhere on the subject, inspired by camera features such as auto modes, and how they help in photography, or don't help.

Photography is about selective attention.

In a photo all you can do is to select what will be seen. In many types of photography, such as landscape and architectural, you have little choice in setting up the subject except maybe, in waiting for the right amount of light. In people photography you can control settings, looks, and light. But unlike drawing or sculpture where you can add the unseen and unreal right from the start, in photography you still have to work by selecting what amount of of a real life scene will be seen, what is necessary to be seen in your photography in order to tell your story.

To select is to frame

Photography is therefore about framing. As a consequence photography is not about showing reality of course. It is most important that the photograph not contain "everything" that was there - it should only contain the restricted set of things that you have chosen for your particular attention - the facts you have now framed, the only ones you want the viewer to pay attention to.

The main dimensions of the framing process in photography as i see them are as follows.

Framing by geometry

With "geometry" I mean, selecting the scene. You have two choices - what is included in the picture in terms of motive, and in what kind of perspective it is included. For the first, the viewfinder is the main tool. For the second, choosing a different lens will change the background and geometric perspective if the main motive is kept constant size. A wide angle up close will include more background and more depth of field than a tele lens framing the same scene from farther away. So, the choice of lens is important because not all framing issues can be solved by stepping back or forward while using a single, standard lens.

Framing by light

What will be within the bracket of the limited dynamic range of the medium, and what will be in the center of that bracket ("correctly exposed"), is determined by lighting and exposure. If it is not possible to include or exclude items of interest, you can use filters or lighting devices. But dynamic range tends to be a fairly tight constraint, you can't allow too much blowout or blocking in the picture. As a subset of framing by light, contrast control allows to shape where and how the motives appear between the extremes of minimum and maximum brightness. The main motive of interest can be set anywhere within the dynamic range bracket - this is the essence of Ansel Adams' zone system, now all done in post processing by contrast curves or HDR techniques.

Framing by focus

Focus point and depth of field allow you to select your desired sharpness bracket within the scene, and set the point that is in true focus. This must be done in real time when you take the picture, through standard photography technique: aperture for depth of field, and shutter speed for frozen motion or motion blur (can also be done by freezing with strobe light). Post processing options exist but are limited, you can't recover out of focus areas and while you can blur parts of the image, this does not work well for all motives. Shift/tilt lenses and view camera movements allow unusual positioning of the plane of sharp focus, though in digital very few photographers have access to this kind of equipment. For maximum flexibility you need fast lenses to allow occasional shallow depth of field.

Framing by color

This can be done by setting the scene, by lighting, by filters, by white balance, and by a host of manipulations in post processing. The real physical means in preparing the set and lighting are the most focused because specific elements can easily be highlighted in different light colors. In post processing you can add white balance changes, tint, saturation, changes in the color palette etc. but all this will affect the whole image. Brushing and masking can add local effects but these techniques can be quite demanding to get right. Setting a "wrong" white balance can also simulate lens filters.

Color is so rich and flexible as a tool that few photographers actually seem to use it as a deliberate tool for shaping content in terms of using color as a framing tool. Often color is just there as an esthetically pleasing side effect. That also explains why many photographs "work" as well or better when converted to black and white. And even here, how exactly the colors are represented during black and white conversion can have dramatically different results. Color adds many levels of complexity, from the well known qualities that apply to a single color, hue, saturation, lightness, to the effects of color contrasts, color families, complementarity, proportions of equal perceived lightness etc. A good primer here is a book from the Bauhaus epoch: Johannes Itten's "The elements of color". And it is not even about photography.

Ironically, in the mature age of color photography color is seldom fully used. It tends to just be there. I include myself very specifically in this criticism. Every time I see that one of my photographs looks better and more interesting in black and white I feel like I must have missed an opportunity to use the medium to my advantage.

Conclusion

What a photographer does is to take decisions on framing reality in every sense of the word. He or she controls what is in the picture. How many dimensions of framing he controls determines his or her options as an artist, and how well he controls them, his or her effectiveness.

There is a popular, purely representationalist idea that says that photography is about showing reality. Nothing could be further from the truth - it is not "reality" but your vision of it that you are showing. In order to do this you should control most of the framing options. And that implies that you must take the decisions on what to include and exclude in the photo, what aspect of it to make important - yourself. Technology should not do this. But technology can do something else - it can help you control these dimensions better.

So then now the question turns to technology. How can technology help? Post processing does help enormously but many decisions have to be taken at the moment of shooting. Then it becomes a question of camera technology. One can approach the situation from two possible logics of operation of a camera. You could want to

1- learn and know the technical means necessary to achieve an effect, say aperture for depth of field, and get the effect by setting the technical means right. This works, is unequivocal, but requires lots of knowledge and experience: traditional technique.

2- tell the camera you want a specific result. The camera then chooses the means to achieve it, automatically. Primitive autofocus or shiftable program modes are one example, but they still require knowledge input. A better way and perfectly feasible nowadays, would be to have dials for desired depth of field, included dynamic range, and middle contrast point, instead of aperture, shutter, and autolevels in post processing. You control directly by effect, not by means. The camera finds a way how.

Option 2 is perfectly valid and straightforward as an approach, but for whatever curious reason no camera maker goes all the way in this alternate philosophy of the logic of operation for a camera. There are hybrids of this logic - landscape modes, portrait modes etc. but these tend to multiply impossibly because there are too many possible combinations of brackets of focus, light, contrast, color, sharpness. If each of these dimensions had just 3 possible values you'd have 3x3x3x3x3 = 243 film modes already, by combinatorial complexity. And color should really have many subdivisions of options. So camera makers aggregate choices, say they force you to use low depth of field with low contrast and low saturation in portrait etc. But that's no good either, you might miss out on a combo you wanted. It would be much simpler to just have sliders of desired set points to mix and match at will: grey point for exposure; black and white points, which also implies contrast; focus point; depth of field near and far points; color/white balance. This could be done say, by thumbwheels on the camera. "Average" could be set by a middle detent. I would think this as a much better way to achieve complex mixtures of intent than "film modes" or "exposure modes".

In the meantime  we are left with aperture, time, focus middle point, and a lot of experience and post processing to get the result right - to have framed the subject appropriately.