Tag Archives: hdr

What a Photograph Is and What It Ain’t

Every so often I post something lengthy in some photography forum or another, and sometimes I want to get as much mileage out of it as a can… so I share it here. Recently there was a discussion about exposure blending and HDR and related stuff in one such forum and people were trying to decide whether HDR is a good, bad, useful, or indifferent thing. I posted a few times in that thread, but here is the final thing I added.

A poster read and quoted the following:

With our knowledge of post-processing techniques, are we involuntarily pre-disposed to see what could have been rather than what is? Does that limit our ability to appreciate the “what is”?

And then responded this way:

It definitely didn’t seem to limit Ansel’s appreciation of what is. You can see quotes throughout all of his books for many varying scenes on how beautiful it was. But then he will also say that he envisioned the final print as ‘stronger’ and did what was necessary to achieve his vision of the scene. Unless you aren’t talking about a live scene but rather a photo – a ‘plain’ photo that tried to capture ‘what is’. I don’t photograph to try and recreate what is. I would find that a waste of time and boring and leaving little in the way of artistic interpretation of the scene. I try to create a photograph using whatever tools necessary to achieve my vision of a given scene and hopefully with a somewhat unique outcome. But I will never limit myself to trying to replicate ‘reality’ as my eye saw it. I still appreciate what is, just not in my photos.

After that I offered up:

This brings up an interesting subject and one that seems to afflict landscape photography discussions more than it does discussions of other types of photography, namely this notion that a photograph “captures” what is “real” and that this can and should be its goal – and, by extension, anything that “manipulates” that “real” thing is somehow wrong and should be called out.

There is very little support anywhere for that idea, at least in the pure form that some seem to think it might have. Virtually every landscape photographer has said or will tell you today and shows through his or her own work that the idea of a photograph as an objective record of “what was there” is both impossible and undesirable. “Recording” the objective, physical nature of the subject – whatever the heck that even is – is almost completely missing the point.

First, it is impossible.

If we assume that the landscape that we see when in its actual presence at the time of the exposure is an objective and real thing, it is obvious that the camera cannot accurately capture that thing. There is a whole list of reasons for this to be the case, and it could include the following and more:

  1. The reality of the place is a continuum of light and seasons and atmosphere and more, yet the photograph only “captures” a tiny slice of the continuum that defines that subject.
  2. The camera cannot record all of the elements that define the nature of that subject – not the movement of air, the smell, the warmth of the sun, the exertion required (or not) to be there, and much more.
  3. The camera cannot “see” the scene the same way that our visual system does – which is the primary subject of this thread. I’ll just point out that bright clouds don’t blow out and shadows are not blocked and leaves don’t blur in the wind when we use our visual system to view them directly.
  4. The photographer’s most basic choices “edit” and transform the reality of the scene in important ways: where to place the camera, when to click the shutter, what to include/exclude from the scene, focal length, whether aperture choices make everything in focus or are selective, what the shutter speed does to moving elements of the scene, and much more.
  5. Other things that would make this list too long for this thread… ;-)

Second, even if it were possible it would be undesirable.

Let’s use Adams as an example. What moves many about his photographs is not the extent to which they are objectively “real” – fundamentally, they are not real. (The last time I checked, the world was not black and white.) What sets his work apart is the way that he used the tools at hand to interpret (not literally reproduce) the subjects of his photographs and the resulting personality and point of view that are expressed in his work. In other words, the literal subjects were, arguably, primarily a means for Adams to share his point of view and his passions through his photographs. In the end, the photographs tell us more about Adams than they tell us about his subjects. (I used Adams here because he is most likely to be known to all reading the thread, but virtually any other “landscape” photographer’s work would serve as well.)

To loop back to the thread, virtually all serious landscape (and other) photographers understand that it is an essentially unquestioned truth that photographs do not and cannot “accurately” portray the real subject, that they inherently (and aren’t we glad!) express a point of view, and that the notion of a pure “unmanipulated” “capture” is a strange and impossible concept. (Yet, for reasons that I won’t explore here, it seems to persist…)

This means that things are complicated. There is no “right” mode of expression, no “right” or wrong techniques, and no “right” type or amount of modification of a photograph in post. It is all relative and subjective. Some who like to imagine that a world of absolutes would simplify things find this difficult to understand and accept. Wouldn’t it be simpler if we could just declare that HDR or exposure blending or adding saturation or using curves or cloning out a spot were “wrong” because they were manipulations of the original “truth” of the scene and dismiss them as being objectively wrong or even dishonest, unethical, or immoral? But we can’t, if for no other reason than once you start down that absolutist road you would have to exclude most or arguably even all photography.

In the end it is about judgment and taste and the power of the photographer’s personal expression – and not simply an accounting of which techniques were used. Perhaps the least important thing about a photograph is how it was made.

Exposure Blending – A Quick Overview

(This is a slightly modified version of something I wrote as part of a discussion on Google+)

The author of the post that I read before writing this mentions two alternatives for dealing with dynamic range issues in landscape scenes – the use of graduated neutral density (GND) filters or the use of a similarly-named post process technique in Lightroom. (Similar processes are available in other software including Photoshop/ACR.) Another alternative is to use HDR (High Dynamic Range) techniques that allow computer algorithms to combine source images in ways that are not the same as what I describe in this post. I almost always use a different approach, exposure blending, and I’d like to share a few ideas on this topic.

When the scene contains an extremely large dynamic range – a common situation in landscape photography – it is possible to end up with shadow areas that are nearly black and devoid of detail, along with bright areas that are blown out and also completely lack detail. (The latter is especially an issue with digital capture in which overexposed areas can simply become pure white. Film failed more gracefully in this situation.)

One traditional method for dealing with such scenes is to attach various types of graduated neutral density (GND) filters to the front of the lens. These filters include a clear area and an area darkened by a few stops. A typical use might be to position the dark portion over the sky and the clear section over a darker foreground, effectively reducing the dynamic range of the light reaching the sensor by reducing only the brightness of the sky. With film, especially positive transparency film, this was just about the only realistic way to handle the situation in many cases.

A second method is to make a single exposure and use features in Photoshop/ACR, Lightroom, or other software to simulate the effect of the GND, lightening (or darkening) a portion of the photograph. This can often help quite a bit when it comes to balancing out dark and light areas in the image, and the fact is that most savvy landscape photographers do this sort of thing and more in post. But it has its limits. For example, darkening a bright area cannot put detail back in blown out areas, and lightening underexposed shadows can lead to problems with noise and posterization. The bottom line is that this technique works well when the overall range 0f brightness in the scene does not exceed the dynamic range of the camera – here it may be used to adjust the relative levels of the areas – but it does not work well when the dynamic range is actually too great for a single normal exposure.

There is a third method, sometimes called exposure blending, that can often be used when making the exposure. The idea is to make more than one exposure, with each exposure optimized for different parts of the scene by varying shutter speed, and then selectively blend the two exposures during post-production.* Two exposures are sufficient in almost all cases, with one for the shadow detail and the other for highlight detail – though in very complex or wide dynamic range situations more exposures can be used. Continue reading Exposure Blending – A Quick Overview