GRAYnomad Nature Photography :: Essays :: Digital Manipulation

Digital Manipulation

NOTE: This essay was published in Better Digital magazine in 2005. There's been a lot of water under the digital-photography bridge in that time so it probably needs a makeover. I still hold to the majority of the arguments but it's pretty much open slather now, everyone (including me) cleans things up as a matter of course these days.

Example images at the bottom of the essay. →

Is it time we just accepted "Photoshopped" images as the new straight photograph? To some extent I think we should, and I will present some arguments to that end in this article.

When I first thought of writing about this subject it was as much to clarify my own thoughts on the issue of digital manipulation as for any other reason. About four years ago I started seriously dealing with digital images, at that time it was scanned film, but whether scans or digital, the conundrum is the same; how much should I manipulate an image? Indeed how much can I manipulate an image and still call it a straight photo?

Let's get one thing clear; there is no such thing as an un-manipulated photo, every photo is manipulated in some way. Everything we do in the process of making a photograph is manipulation, the choice of framing, lens, cropping, juxtapositions, exposure, colour, motion blur, selective focus, perspective…I could go on.

This article presents some of the issues as they relate to nature photography, which is really a form of reportage and, as such, strives to have the viewer believe that the events in the photograph actually happened.

What is the fundamental issue?
If you photograph for your own enjoyment, and are not selling or otherwise exposing your work to the public, there is no issue. But, if you are presenting your work in any way, and especially if you are selling images, there is an issue, and that issue is one of deception. Can a viewer "believe" a photograph?

Photographs are supposed to depict events or scenes that actually happened, that's the whole reason people are blown away by a fantastic image. We assume that there has been no hank-panky on the part of the photographer; if the viewer were there at the time, he or she would have seen the same thing.

Photography is in danger of becoming disregarded as a medium that depicts real scenes and decisive moments, and lumped in with painting, ironically something many have wanted since the first wet plate was exposed.

Is this a new "problem"?
Many feel that the issue of manipulation is a problem that surfaced with Photoshop and the digital camera. Not so. The earliest example I know of is a photo entitled In a Piegan Lodge taken by Edward S Curtis in 1910. The photo is of two American Indians sitting inside their lodge. Between them on the floor is an alarm clock, but it appears that Curtis didn't think the clock was "indian" enough, so he retouched it from the print.

Move on a few decades. Ansel Adams' Mt McKinley and Wonder Lake photo was printed in 1949 as a washed-out insipid landscape. Years later we see a version with stunning blacks and a fantastic tonal range, a classic Adams print. The image has totally changed between 1949 and 1978. There was more than a little manipulation at play there.

Still with Ansel Adams, what about the retouching of the "LP" graffiti from the hillside in Winter Sunrise, another of his better-known images.

In both the Adams examples I cite, manipulation was at play, but everyone knows about it (or at least us photographers do) so there is no deception; anything is OK as long as the viewer is not being deceived.

Who is the viewer?
But there are viewers and there are viewers. Just who, or what, is the viewer? A photographer can be deceived in different ways to a layperson. Take the example of a macro photo that has had the depth of field enhanced by combining several shots with slightly differing focus points.

The layperson is not deceived because, in essence, the photo faithfully depicts the scene before the photographer. However a fellow nature photographer may look at the shot and marvel that the insect is sharp from head to tail, "How on earth did he get such fantastic depth of field?" the photographer may ask. He would get the impression that the author of the photo managed in some way to have the subject stay perfectly still for a 3-second exposure at f64. He may be deceived into thinking that the author had a different skill set than was in fact the case. Rather than being able to mesmerize animals, the author was simply proficient on the computer.

Image manipulation is a skill and should be recognized as such, but it is a different skill to photography.

Why do we care what the viewer thinks?
I must admit that, when viewing a good image, I now find myself wondering if it is real, or has it been manipulated. Now I'm well versed in this field and can often spot a manipulated photo, but if it's done well there is nothing to spot.

For a while no harm will be done, the average person won't know that images are not what they appear, and won't care. But eventually people will get burned. They may buy a classic shot of an eagle soaring above a snow-capped mountain, only to find that the eagle was on a leash at a wildlife park, and the mountain was located on another continent.

How likely is that person to purchase another nature photo?

If you are selling or displaying images you should care what the viewer thinks.

Some other issues
And now for something completely different. I'm moving out of the photography arena here and into some environmental issues. When is it more ethical to manipulate than not to manipulate?

Take for example the rotten stick that ruins your foreground. Most photographers would consider it reasonable to move the stick and thereby improve the photo. But what about the thousands of termites who call the stick home? The earthworms that will not survive if exposed to the hot sun. The cockroaches that will be picked off by birds if left in the open.

We're not talking the Exxon Valdez disaster here, but there are other issues at play than just photography ethics. In this example it is surely more ethical to remove the stick in the lightroom.

What if you see the distracting stick in the field but decide not to remove it because it's easier to clone it out on the computer. Are you being eco-friendly, or just lazy? What if you didn't see it at the time? From a straight-photo point of view, what's the difference between physically removing the stick and cloning it from the photo? None. Once you interfere with a scene it is no longer a "found view", it is a manipulated view.

Of course if you remove the stick in the field no one will ever know and I wonder how many famous "straight" photos had a stick moved to tidy up the foreground, or a branch broken off to clear the view? I suspect that even some of the most iconic images have had this treatment, but there is no way of knowing.

Therefore I argue that removing the stick with a computer is more ethical and still results in a straight photo, and the photographer has had less impact on the environment that he or she loves.

In-camera VS in-computer
For years some photographers have espoused the purity of an image that is displayed entirely as captured by the camera, to the point of including the negative frame in the print to prove that the image is uncropped. Such images are, some will argue, the purest form of photography. I have no problem with that, as a genre it is as valid as any other. But let's not impose those restrictions on other photographic styles, and as I've stated, even these images are manipulated, it's just that the manipulation occurred before the shutter was fired.

What's so sacrosanct about doing the manipulation in-camera? For over a hundred years photographers with large-format view cameras have been "distorting" subjects, most commonly raising and/or lowering the camera's standards to correct leaning verticals. How is that different from creating exactly the same affect with Photoshop's perspective tool?

While talking about view cameras, the other common trick used is to tilt the standards. This in turn tilts the plane of focus and gives the illusion of incredible depth of field. How is this different from combining similar images with different focus point to obtain the same affect?

What about panoramas, why is a photographer with pockets deep enough to buy a Linhof Technorama any better than one who creates a panoramic image from several consecutive exposures and a stitching program?

Is the use of a ND graduating filter any more ethical than the application of a graduated adjustment layer?

In short I say no, all the above techniques are equivalent. Of course when these adjustments are performed post-exposure, the photographer has a lot more room to experiment, the pressure to get it right in the field is off. So what? New technology has made life easier, if you don't like it, buy some eggs, some glass plates, a lightproof tent, three pack horses, and travel back in time to do some albumen printing.

In fact I see the raw digital image in much the same way as I see a negative, it's the raw material, from which something beautiful can be created. If I do that by waving my hand over an easel, or a mouse mat, it doesn't matter.

Many photographers, especially in the fine art genre, sidestep the issue by saying "This print represents what I felt when I was there". This is a cop out; we can all say that we "feel" whatever we like. After hours of pushing through scrub without finding a single decent image I "feel like" a cheeseburger and a beer, but that's got nothing to do with the photograph. And who's to say the "feeling" wasn't conjured up six months later when the photographer was pressed for an artist's statement by a gallery.

One of photography's appeals is that it depicts an actual event in a manner that was not possible before the camera's invention. If you see a photo of a leopard sitting on a rock with the moon in the background, then you have the right to expect that that scene actually occurred. The photographer can adjust the lighting at the scene by electing to use a flash or not, similarly he can adjust the mood by increasing or decreasing the exposure (either in-camera or in-computer). But he cannot say, "Well it felt like a balmy moonlit night" and add a full moon from three days before.

What do the experts say?
In an attempt to ascertain what the experts think on this subject I did some web surfing, starting with some high-profile competitions.

The British Gas (now Shell) Wildlife Photographer of the Year competition is arguably the most coveted prize in the nature photography genre. They only began accepting digital images recently, and do not have a section for manipulated images. The rules quote…

"Digital manipulation is only acceptable if limited to cleaning work, levels, curves, colour, saturation and contrast work, applied to the image as a whole."

Moving across the Atlantic to the USA we have the Nature's Best competition.

"…all photographs should accurately reflect the subject matter and the scene as it appeared in the viewfinder. Nothing should be added to an image, and, aside from minor dust spots, nothing should be taken away. Cropping and minor adjustments to colour and contrast are acceptable."

And just to spread my quotes more evenly around the world, here is what the ANZANG (Australia, New Zealand, Antarctica and New Guinea) nature photography competition rules say…

"Images must not be adjusted beyond a level that would be applied in conventional optical printing techniques."

Let's think about that last quote for a minute. Only changes made that could be done with conventional optical printing techniques…well I was a pretty dab hand in a darkroom, and there are plenty of people who are absolute wizards. Jerry Uelsmann's images are created entirely in a darkroom using "conventional optical printing techniques", they no more depict reality than Star Trek does, but under this criterion, his images could be entered in the ANZANG competition.

It seems that everything is open to interpretation.

As you would expect photographers opinions cover the entire range, here are some examples from three of the world's top nature photographers.

Richard du Toit
"I really dislike the digital manipulation of wildlife images. For me it destroys the authenticity of the image, and is damaging the public's perception of nature photography…I want my images to be an accurate record of the behaviour of animals as captured by the camera in the natural environment. I believe that's what the public want also."

Joe & Mary Anne McDonald
"To me, a digitally enhanced or computer manipulated image should imply creating an image with elements not originally present in a scene. Adding penguins to an iceberg, for example, or flying birds over a landscape. Just as photo 'sandwiches' … should be labeled, these computer images should be, too. But if a photographer is simply using a computer to truly capture what your eye can see, using the computer as a sophisticated and highly advanced ND filter, I'd argue that that is not digital manipulation. At least not in its purest form."

Steve Bloom
"I believe that manipulation is acceptable so long as the basic integrity of the original picture is maintained. I wouldn't, for example, change a picture so that an animal is shown in the wrong environment or doing something that was completely out of character."

Let's take the case of Steve Bloom's Polar bear cubs playing on their mother.

My initial response was "what a great shot", but then I looked closer. There was something not quite right about the image, and eventually I realised what it was. The light source is different for one of the cubs than for it's sibling and mother. It appears to be a fabricated image; the animals are not "in the wrong environment" or "doing something completely out of character", and I'm sure that the cubs were indeed playing on their mother at the time, but this scene was almost certainly never captured by the photographer in one decisive moment.

Steve Bloom is without doubt one of the world's finest nature photographers, but I now find myself wondering about his images. I emailed Steve and asked him to comment, but as yet have received no reply. Rightly or wrongly my confidence in his images has now been eroded.

Do you want your viewer to lose their confidence in you images?

Do we need a labeling standard to indicate that an image is as it appears? Maybe, but it's been tried before.

Previous image-labeling standards
To my knowledge there have been two serious attempts to provide a standard by which photographers can tag their images as being un-manipulated, Foundview and Trustimage.

Foundview was created in 1997 as a method of indicating to a viewer that an image was not manipulated.

According to their, now defunct, website…

"The FoundView checkmark is their [the viewer's] guarantee that post-shutter manipulation, if any, was limited to tonal variations (contrast, brightness, intensity, hue) and that no one involved in producing a FoundView photograph moved, added, deleted (except by cropping), or otherwise altered any forms or shapes in that photograph after the shutter was clicked."

The Foundview logo could be placed on any photo that satisfied the criteria. As those trying to hoodwink the public will not accede to any form of disclosure, it seemed reasonable to provide a method by which those who don't unduly manipulate can indicate that fact.

Foundview never took off.

Trustimage has grown from the Foundview standard and is aimed more at the reportage arena. Trustimage has a large procedure that you can follow to determine whether an image qualifies or not. It's way too complicated to repeat here, however there are two basic rules.

Qualification #1. The photograph was made from a single uninterrupted exposure, and all forms and shapes within the photograph remain exactly as they were recorded during that exposure (with exceptions made ONLY for cropping, sharpening, and changes to tones and colours).

Qualification #2. The photograph and its presentation meet newspaper reportage standards for non-deception and non-misrepresentation. This applies to both the appearance (including tones and colours) and the circumstances of the original scene as it was while the shutter was open.

Because Trustimage is aimed at reportage they don't appear too worried about some technical issues that would ordinarily disqualify an image as being un-manipulated. For example, they allow the removal of flare from the photograph. This of course has no bearing on the believability of the image when viewed by the average newspaper reader. However flare in a photo is normally a fundamental error; surely this should have been noticed by the photographer and dealt with pre-exposure.

FoundView does allow for "similars" with different exposures to be combined, whereas Trustimage states that the image must be "from a single uninterrupted exposure".

Neither allows for stitched panoramas, something that I believe should be considered straight.

Coming from an engineering background, I realise that standards are very important, so much so that I've decided to create my own.

My "standard"
The problem with labeling systems as describe above is that they only distinguish between "genuine absolutely untouched photos" and "everything else". This lumps the "removed a blade of grass" image in with "added a sky, duplicated several impalas, and dropped a mountain in the background" photo. Not fair in my view.

I propose a four-level scale, ranging from absolutely untouched, to a totally fictitious montage.

Level 1 – absolutely untouched, except to overcome the most fundamental limitations of the medium, (equivalent to typical procedures when shooting with slide film).

When I moved from using 5x4" sheet film in darkslides to pre-packaged Readyloads I thought my dust problems were over, and then I discovered digital. Until the manufacturers solve the problem, dust is a fact of life for digital photographers. It's only fair to allow dust spots to be removed post exposure. Similarly, sensors often seem to produce a raw image that is pretty insipid and doesn't accurately depict the scene at all in terms of tonal range and colour. Therefore the photographer should be allowed to apply global changes to contrast, density, colour, and saturation.

Moderate sharpening is allowed as this is also a limitation of the medium, digital cameras have a low-pass filter placed before the sensor to reduce moire patterns, this slightly blurs the image, and the photographer should be allowed to rectify this.

Panoramas are allowed, the fact that the image is a composition of several exposures produced within seconds of each other is irrelevant. As long as the resultant stitched image has no undue manipulation performed on it, the image qualifies for Level 1.

A Level 1 photograph is the truest test of the photographer's art in the sense that we have known it until recently. Certainly the ability to adjust exposure, colour balance etc. after the event gives the digital photographer an advantage over his slide-film counterpart, but this is about as close as we can get to shooting transparencies.

With the exception of panoramas, photos at this level can be entered in all major competitions.

Level 1 photos can be considered a true likeness of the scene as it appeared in front of the camera.

Level 2 – some minor annoyances removed, selective adjustment of tones, similars combined, (equivalent to using typical darkroom procedures to make a print).

Some grass poking up in front of the subject has been removed, an overly bright rock has been dulled down a bit, the shadows have been lifted, and contrast increased in the bottom right-hand corner. All these manipulations can be done in a darkroom, and therefore, should be allowed in the lightroom, with the results still being accepted as a straight photo. The only difference is that these changes are now easier to do.

Similars with different exposures or focus points, which have been combined into a single image with a larger tonal range or depth of field, are allowed. This is the digital equivalent of various masking techniques in the darkroom.

There is no trickery here, just craft.

With levels 1 and 2 the finished result still depicts a scene or event that actually happened, and which is in no significant way different to the original scene.

When I sell a wet-process B&W print I don't say, "well of course I darkened that rock, lightened some of the shadows, and retouched a white pebble on the shore". The photo is accepted by the purchaser as being un-manipulated, it should be the same with Level 2 images.

Level 2 photos do not deceive viewers, and should also be considered "straight".

Level 3 – major work done, important items removed and/or added.

At this level, trees, buildings, or even whole towns can be removed. Now we are getting into the realm of deception, if a pristine rolling hill is only such because you cloned out a rubbish tip, then you are deceiving the viewer if this fact is not disclosed.

You can also add items at this level. Most commonly this would be used to improve a bland sky by dropping in some nice clouds, or fill in an empty position in an otherwise perfect row of trees.

However, a Level 3 photo may have, for example, had a small herd of animals cloned to produce a mass migration, or a lone canoeist dropped onto a pristine lake. At this level the event or scene was not present when the photographer made the exposure, the image is no longer a photo, but a "photo illustration", indistinguishable from a straight photo maybe, but not real.

Of course you can argue semantics here, for example, if you cloned a green field over a rubbish tip, did you add grass, or remove rubbish? For this reason, although I originally differentiated between adding and removing items, I now believe that both actions are essentially the same.

Level 4 – open slather, what you see is what I wanted you to see.

This level is probably superfluous because, with any image having this much work it will be quite obvious to the viewer that they are not looking at a straight photograph.

Of course there are many grey areas here. What about deliberately darkening the whole image to a point where some detail is lost, thereby effectively removing it? What if the details were lost as a by-product of poor reproduction in a magazine? In either case a Level 1 photo suddenly turns into a Level 2 or 3.

Just exactly how many blades of grass can be removed before a Level 2 photo becomes a Level 3?

I can't answer these questions; fortunately I don't have to.

When I started shooting digital I spent ages on just about every image, lightening a shadow here, tweaking the contrast there, selectively sharpening somewhere else, etcetera etcetera.

These days I just do the minimum required to produce a pleasing image; I try not to overly manipulate for two reasons. Firstly, I aim to produce high quality "straight" photographs of scenes and events. Secondly, it takes a lot of time to manipulate an image; I simply don't have that much time. I can easily shoot 2, 3, 4, or even 500 photos in a day; even without making many adjustments it will take me all night and half the next day to import, adjust, catalogue, and backup those photos.

I have presented some arguments here in the hope that it will cause you to think about the issues. I will soon start to implement the above four-level standard to my own work, but I do not expect anyone else will use it. Once again, I just hope it will provide you with some food for thought.

Manipulation is here to stay and there is absolutely no chance of imposing standards or forcing disclosure. In the end it boils down to trust; the viewer must be able to trust the photographer. Photography's unique position as an art form that closely depicts reality should not be discarded lightly.

Example images

Figure 1:
Lake Albina, Kosciuszko National Park, NSW

Albina Storm

Here we see a classic large-format, Ansel Adams-style black & white landscape. It's a pure un-manipulated image in the classic tradition right? Well yes and no. It is a traditional large-format black & white landscape made on sheet film with a 4x5" field camera, but it's also heavily manipulated. Firstly a red filter was used to manipulate the tonal values, this dramatised the sky over and above what it was in real life, and lifted the reflections in the lake surface. The print is manipulated to blazers, including, but not limited to, split contrast dodging and burning, masking, and selective ferricyanide bleaching.

Figure 2:
Pied heron, Marlgu Billabong, near Wyndham, WA

Pied heron with fish

Traditionally a photo should depict, verbatim, an event that occurred. If the viewer were standing beside the photographer he or she would have seen exactly the same thing. This photo of a Pied Heron is such an image. While it is tempting to remove a slightly distracting water lily, I have left the Photoshop tools in the shed so the image can keep its Level 1 status, ie. a totally unmanipulated image.

Figure 3:
Wedge-tailed eagle, near Cadney Park roadhouse, SA

Wedgetailed eagle

Was it laziness or lack of attention to detail that caused me to have a dead branch poking into the top left of this image?

Neither, in fact it's there deliberately. If I changed my position to exclude this branch, a much larger one moved into the frame and imposed itself in front of the eagle. I chose the lesser of the two evils, knowing that it would be much easier to remove this branch because it was surrounded by plain sky, and not the detail of the bird's feathers as the other branch would have been. With my classification system this is a Level 2 image, meaning that the same manipulation could have been performed in a darkroom, although I confess I would not have had the skill to do so.

Figure 4:
Freshwater crocodile, Windjana Gorge, Kimberley, WA

Johnstons crocodile

Does enhancing the depth of field by combining images with different focus points constitute "manipulation"? I argue that it doesn't, this is simply a new tool that we, as photographers, have available. This version of the photo is as it would have appeared in my folio a few years ago, it is focused on the eye according to rule #1 in the wildlife photographer's handbook.

But with a 640mm (equiv 35mm focal length) lens there is no way to get the mouth in focus as well. By shooting two more frames, each focused a little further down the nose, I was able to merge the three images and create an image that actually depicts reality more than a single exposure does.

Figure 5:
Nulla Nullas, West MacDonnell Ranges, NT

Green pussytails

A shot of this type will typically require some serious burning in of the sky when printing in a darkroom. This it usually achieved by cutting a piece of cardboard to make a mask that roughly matches the shape of the horizon, and using the mask to "burn the sky in" by adding exposure to that area of the print while shading the foreground. In the lightroom we can easily create an adjustment layer to achieve the same affect. No trickery, just a new tool.

Figure 6:
Bells Gorge, Kimberley, WA

View of the waterfall

A scene of this nature, with white water in bright sunlight adjacent to deep shadows, is almost impossible to capture in a single exposure with slide film or digital. By making two exposures, one for the majority of the scene, and another for the water, I was able to combine them and produce a result that closely resembles what the human eye saw at the time. This is not manipulation; it's a way of overcoming the limitations of the media.

Figure 7:
Little Mertens Falls, Mitchell Plateau, Kimberley, WA

Pandanus at the base of the falls

This is a five-image panorama with a field of view of 194 degrees. Before the days of stitching software it was all but impossible to record a scene like this unless you could afford a special panoramic camera with a rotating lens. If taken with such a camera this photo would be eligible for entry in any competition, but because it was taken as five separate exposures it is not.

This ruling should change, this is not a manipulated photo, it depicts exactly what I saw at the time, and what you would have seen had you stood next to me. Stitched panoramas should be considered straight photos.

Figure 8:
Black Kite and moon, Georgina River, QLD

Black kite and moon

Even though this is how the eye would see the scene it's actually not possible to capture with a camera in a single frame. Why? Because the shot requires a very long lens and with the bird at a distance of maybe 10 meters and the moon at 240,000 miles the the depth of field doesn't quite cover that.

So I took two frames and combined them.

Unlike the example in the article (the leopard and balmy summer night) this actually actually was a genuine found image, I just had to overcome a limitation of the media.

Figure 9:
Seal, Jurien WA & Mudflats, Parry's Creek, WA

A Level 4 image like this should need no label to inform the viewer that it is a manipulated photo. Nobody would really think that a seal would be photographed emerging from a hole in a dry mud flat, as though it was an ice flow.


No comments yet

Post a comment




Enter code then