The possibility to alter images has been around for ever, initially in drawing and painting and certainly in photography. I am sure that the earliest cave painters must have added the occasional extra mammoth or two for dramatic effect .. However, what do we mean by alteration and what is my reaction to it?
At the most fundamental level a digital image straight out of the camera needs some level of manipulation to look its best even if the objective is to make the photograph look as close to the original scene as possible. I am talking here of images shot in my preferential RAW format, as it is obviously possible to set up modern DSLRs to shoot in jpeg and use the camera settings to adjust white balance, exposure etc. and get the camera output to look as like the subject being shot as possible A RAW image file will require some measure of correction to apply the lens correction profile and adjust white balance (I always shoot in AWB), maybe tint, exposure, contrast, clarity (mid-tone contrast) and sharpness as well as colour related parameters such as saturation, vibrance and tone. Personally, I do not regard this as anything other than the routine workflow required as part of the optimisation of the appearance of the image, although clearly it is technically an alteration to the photograph as shot.
To me, alteration/manipulation of an image starts at the point where the photographer starts to remove or add pixels. Minor amendments such as removal of sensor dust spots are acceptable in all competitions and by all agencies that I am aware of and there can be few who would regard this as outside normal practice. The next step of course is the removal of small irritations that are not in any way part of the main subject of the image, so passing aircraft in a sky, wires, ‘fly away’ hairs in portraits etc. I would regard as fair game for removal. They have no impact on the photograph, and in fact detract from it, so why should they not be obliterated? I would never consider removing anything from the main subject of the photograph that altered its context or meaning, although that does not mean that I feel I have justification for criticism of those who do as long as they are not trying to pass an image off as something else by altering its context or meaning without declaring so. There are many well-known examples where major subtraction or addition has taken place in photographs, going back to the early days of film, although the advent of digital imagery and the ease with which photo-manipulation software can now be accessed and used has released a massive wave of fakery. Now just about anyone can remove things from photographs and add them at will, although I am always surprised at how clumsily this is often done by allegedly ‘professional’ bodies who you would think had good reasons to make a better job of covering their tracks.
I have written so far about the alteration of images that are purported to have some documentary or reporting validity where changes can influence public opinion and inflame debate and where alterations are an insidious part of the political connivance undertaken by governments and companies. I don’t think that Di Tzeitung was too worried about white balance and vibrance when Hilary Clinton was removed from the images of the US government party watching the assassination of Bin Laden .. The other side of image manipulation is where photographs have been altered, combined and distorted in ways purporting to support some artistic endeavour, and there seems to be an increasing prevalence of these.
There are an increasing number of websites which devote themselves to the exposure of photographic fakes, and it seems to take them only a little time to establish falsehood such that many of the well documented fakes often have short lived celebrity standing. Interestingly, some of the older film fakes took rather longer to debunk, such as the Cottingley fairies and the Loch Ness monster; there will be more on these later.
So, what about photo fakery then? In my view a fake is an image where there is deliberate intention to mislead the viewer by subtraction or addition, not where a few things have been removed to tidy up an image. One of the earliest recorded photographic fakes arose at the birth of photography. Although only using a misleading caption and posed image, the supposed suicide of M. Bayard, who actually created this photograph himself, and who claimed to have been the original inventor of the photographic process before Daguerre, produced the construct to make a point.
A simple search reveals that many photographic fakes were generated long before the advent of digital photography and the use of image manipulation software (see references), but one endeavour that reputedly made the perpetrator some money was the ‘spirit photographs’ of William Mumler which first appeared around 1861 with a self-portrait that had the appearance of a shadowy figure in the background. Clearly a processing error of some description, but Mumler was able to make capital out of the first image of a ghost (the figure was likened to a dead cousin by Mumler’s associates). Error or not, by 1871 Mumler produced a work of Abraham Lincoln’s widow which appeared to show the ghostly shape of the president behind her.
The first of the fakes to make it into the mainstream of public debate was The Cottingley Fairies, constructed by two teenage girls, which convinced many people of the existence of fairies, especially as its provenance was backed up by Conan Doyle. Remarkably, it was not until the 1970s that the relevant research was performed to finally dismiss this image as a hoax.
The Loch Ness monster and UFOs have certainly been subjects of many photographic hoaxes in the past and there are numerous examples of fakes from the field of conflict, starting with the American Civil War onwards. However, the main thrust of debate in the hoax world has progressed since the advent of digital imagery and the availability of Photoshop etc., so I intend to focus on that area for the remainder of this short overview.
For a fake to go viral it has to have a reason to capture the public imagination, and probably none have done this more than the Accidental Tourist, supposedly taken moments before the horrors of 9/11 unfolded. It is an episode in history which has had more web searches than any other and the appearance of this photograph initiated unprecedented interest. The image has been discredited on many fronts.
There is a plethora of photo-forensics websites which analyse the provenance of images and the quote below is from www.snopes.com which details the many pieces of evidence confirming the image to be a fake. These analyses exist for a multitude of fakes and ‘photo-forensics’ is a rapidly expanding area, although nobody should be surprised by that I don’t think, as even online policing by self-appointed photo vigilantes is better than letting cheats get away with it. The FotoForensics website is worth a look and try putting one of your own images in there for it to analyse – I was staggered by the amount of information that is not readily viewable in Photoshop that the program displayed from the image I chose.
…. aside from all the digital imperfections in the image (e.g., shadows of different objects don’t correspond to the same light source, the date-time stamp is the wrong type of font), a number of logistical errors proved it to be suspect beyond credibility:
- September 11 was a warm and sunny late summer day, not the type of weather in which a tourist would have been decked out in a winter coat and hat.
- The airliner shown in the picture is approaching from the north and would therefore have been the plane that hit the north tower of the World Trade Center (WTC1), but WTC1 did not have an outdoor observation deck. The south tower (WTC2) included an indoor observation deck on the 107th floor and an outdoor deck above the 110th floor, but WTC1 housed only Windows on the World, an indoor restaurant with a magnificent view of the city but no outdoor deck.
- The operating hours in September for the WTC2 observatories were 9:30 A.M. to 9:30 P.M., meaning they opened too late for a tourist to have been present on one of them on September 11 before the first plane hit the WTC at 8:49 A.M.
- The aircraft shown is a Boeing 757 bearing American Airline markings, but Flight 11, the only American flight to crash into the World Trade Center, was a 767. (The 767 is a wide-body jet; the 757 is a smaller, standard-body craft. This photograph on Airliners.net, from which the image of the plane used here was probably taken, is of a Boeing 757.)
Although the photo itself was quickly debunked as a digital manipulation, one looming mystery remained: Who was the tourist whose likeness appeared in the image? A Hungarian man named Peter staked a much better claim to the “Tourist Guy“ crown by supplying the original photograph of himself (and other similar snapshots from the same session) and revealing that he himself had created the infamous “Accidental Tourist” image.
The Denver Post is an excellent source of information and analysis of faked digital images. The death shot of Bin Laden, a simple and not very good combination of two images, heads their collection of manipulated shots of political figures, celebrities and events that have been sold to the world as genuine in an attempt to sway public opinion, steer political debate or pander to religious or ideological beliefs. The removal of ‘powerful’ women from the image showing the US government watching the Bin Laden operation by the orthodox Jewish newspaper Di Tzeitung focussed recent public debate on photographic forgery and highlighted it in the public domain. I recall press coverage at the time varied between those vilifying reprehensible and transparent forgery and those claiming it to be an acceptable approach to fit a belief system. I can feel myself edging towards my soapbox here, so time to move on.
The World Press Photo of the Year, Gaza Burial by Paul Hansen, has been subject of much debate as to what level of alteration was imposed upon it. The forensics experts point out that there have been significant levels of dodging to lighten a number of parts of the image (which seems very clear to me – there would appear to be just too many perfectly lit faces) and there is suggestion that the published image is a composite of three or four sequential photographs. The ensuing debate surrounds whether the alleged actions of the photographer actually contravene the rules of the World Press Photo competition and it is clear from the whole discussion that there is no real level of agreement around the acceptable level of fakery.
Finally, It interests me how often I now hear people who are looking at photographs quickly start to debate whether an image is real or fake and how an element of suspicion immediately falls on anything that is unusual, of outstanding interest, or indeed, even just technically brilliant. “I wonder how they did that ….?” Well, maybe sometimes they just pressed the shutter button …
Assignment 4; early thoughts
I gave consideration to a number of options for an image for Assignment 4. The OCA study guide claims that the exercises and projects in this section should “help clarify your opinion on the potential for altering content and viewers perception in an image, and helped you to define your own stance. The purpose of this Assignment is for you to demonstrate this stance, and the means involve completing a task which lies in the middle ground of the real versus fake argument.” The objective is to produce an image to illustrate an imaginary book (or present an alternative cover for an existing book) or magazine cover, giving due consideration to the cover as a sales vehicle which is potentially open to wide interpretation. The OCA notes claim that this is potentially ambiguous moral ground and asks that the student gives their ethical justification for the approach and final outcome. I may revise my opinion here, but I’m not sure I will have an ethical justification for something I would never be interested in doing through lack of reason or interest, but I’ll see how that thought develops as I go along.
I want to produce an image that leaves the viewer undecided whether the photograph before them is real or fake, and have considered a number of options. I wanted to remain with the railway theme that I have followed for much of DPP and a little thought around appropriate material conjured up ideas like The Railway Children as a book cover or Brief Encounter as a film poster as possible alternatives in the re-creation line. I also had some thoughts around a cover for a railway photography book and the notion of including myself in the image.
My starting background image is of an onrushing steam train and the plan is to place myself in front of it such that it casts doubt as to whether the image is real or fake. The train image was taken from almost head on and selected to give the impression of it speeding towards where I wanted to place a figure. The figure was shot separately using a remote triggered setup and was taken against a contrasting background, or maybe I was wearing contrasting clothes, to make selection and isolation easier. My vision was to have the figure stepping over the tracks, so images were taken of the figure walking in such a way that the gait would be appropriate for what I had in mind for the final image. Both were shot under similarly overcast conditions to reduce the amount of highlight and shadow work required in the final composite. The angle of the walker and the train were important as I wanted to make it clear that the figure was neither seeing nor hearing the train, so the angle of the head and eyes was critical to the intended fake.
The initial images were optimised in Lightroom 5.2 in the way I have described previously and both exported into Photoshop CS6 and layer copies made to preserve the initial background layers. The figure was selected using the Quick Selection tool and + or – the Alt key used to make fine adjustments; tablet pen pressure set the brush size and I used 100% hardness. Refine Edge was the next step and I used the Smart Radius tool which set an edge detection of 2.3 pixels which seemed visually fine. The other settings for smooth, feather, contrast and shift edge that I selected are shown in the snip below.
The selection of the figure thus made was sized using the Transform tool and copied onto the image of the approaching engine on a separate new layer. It took some while to get the proportion of the figure as I wanted as it needed to be positioned some way in front of the train, so I used the known length of the boot and the known dimensions of train track (yes, I know I’m a nerd) to get the appropriate scaling. Then using the Move tool, the figure was positioned in the foreground such that the illusion of stepping over the railway tracks could be finalised.
The longest part of the process followed and this involved use of the Clone Stamp and Healing Brush tool and with Content Aware fill selected to tidy up around the feet and remove stray blades of grass carried over from the original. This was again done on a new layer copy.
I still needed to consider the words at this stage, both in terms of font and structure, to make it look like a book cover, and the following is obviously just a draft I created to check out the feel of the image. I’ll develop this further after giving it a few days refection to ensure this is really the way I want to go. My last choice is whether to go with black and white as I have done for the rest of the railway images, or to use colour. My first thought is to use colour as it seems to have more impact in this instance, but I’ve put both here while I think about it.
I’m happy with the image as a colour version as that would seem to be a more likely editorial choice if this really were the cover of a novel, so I finalised my decision on that. At this stage then, I need to work on a real title and anything else I plan to add.
I decided that I needed a better background for the title and Clone Stamp was again used on a new layer to expand the area covered by smoke and hide the pine branches at the top of the original shot – I used 75% opacity to leave just the feeling of there being something there behind the smoke plume.
E Nesbitt’s The Railway Children is a little over one hundred years old and was what initiated my fascination with steam, so it seemed a reasonable ploy to go with this as a title for my composite image. Working with separate layers in CS6 I added the text fields and used the colour sampling eyedropper to select a green colour from the trees for the book title and from the red buffer beam of the engine for the text at the bottom. I did have a brief dalliance with blurred and smudged fonts, but in the end decided to keep it simple and I used a basic curved font structure to align with the curve of the engine boiler.
The final job was to flatten the image and save it as a 16bit tiff file for archiving and as jpegs for screen for the blog (1400 pixel length and 72dpi in sRGB) and for print (not resized and 300dpi in AdobeRGB).
Demonstration of technical and visual skills
I was pleased with the range of technical skills in CS6 and LR5.2 that initially crafted the two separate images and then blended them to generate the final composite. I think there was the need for significant visual awareness in making a photograph of the figure that was walking with the correct gait to appear natural when it was placed over the tracks and this took over a hundred attempts before I felt I had got the stride and the angle just how I wanted them. There was also the issue of sizing the figure, which was done, as I said earlier, by scaling against the dimensions of two known objects.
Quality of outcome
The book cover worked well as a vehicle for the fake and required the alignment of photographic, compositional and Photoshop skills and I think I presented the initial review of photographic fakery and my own subsequent project in a cohesive manner that had a logical flow. Bearing in mind that a book cover for a novel is likely to be seen at considerably less than A4 size, the visual impression wou ld be expected to be convincing.
Demonstration of creativity
The development of a personal voice is a topic that often features in OCA-speak and although the type of photographs of any specific subject I take are often quite similar in nature and structure, I’m really not sure that I have developed a photographic voice as such across a disparate range of subjects. I think perhaps a voice takes a long time to emerge and after 18 months perhaps I’m being unreasonable to expect such, as in my wildlife work I have often been told that “I knew it was one of yours from the thumbnail before I opened it“ so perhaps in that area a voice has developed .. or maybe that’s just a style? At some point I need to give greater consideration to ‘voice’ and blog some thoughts as to what I think it really is.
Assignment 4 took me into the realms of photographic fakes and photo-forensics, both areas to which I have given little thought in the past other than as a casual observer of well-documented fakes. I researched quite deeply into some of the areas and was surprised at the extent to which even the earliest film photographers used fake images. Maybe I should not have been, as the history of painting is rich with pictures which had been altered, over-painted or were the products of more than one artist. Why therefore should photography have been any different after its invention? In the days of film it took expertise of at least an experienced amateur hand to fake photographs, but since the advent of digital and the possibilities opened up by Photoshop and the like, anyone with a little practice can falsify images that will at least pass cursory on screen examination.
I enjoyed the technical challenge of getting just the right image of the figure and of blending the two images, and was pleased with the outcome, but I don’t think I have a future in compositional art or photographic fakery. Either the photograph is there to be taken, or it’s not.
A number of times I considered introducing more elements into the image, although in the end I decided this was just replicating the challenge so elected not to expand the assignment.