When image manipulation applications were first invented, there was little discussion about the rights and wrongs and the limits of what can or could not be manipulated. The reason for this was, in large measure, a simple economic one. Only top agencies or highest-paid advertising photographers could afford to use the dedicated hardware that was required. As we know, this changed in the late 1980s when Adobe Photoshop was introduced to the Macintosh platform. In all this time, we have struggled to achieve some kind of clarity. I am not sure that perfect clarity is possible. But I believe that clarity in establishing the parameters that are agreed to suit a given genre of photography will help to enhance debate. I offer here a way to frame critical assessment of image manipulation within agreed contexts.
Discussions of what level of image manipulation is or is not acceptable usually make sweeping generalisations and imprecise references to dark-room techniques. The lack of precision in the debate is egregious, particularly when experienced in the context of news or documentary photography or photojournalism. Yet that’s where clarity matters most. Appeals to maintaining ‘high standards’ have little value in the absence of a shared understanding of norms that distinguish right from wrong, unacceptable from acceptable.
When deciding what level of image manipulation is acceptable, discussions often make appeal to analogue dark-room techniques to set the normative markers. Apart from its imprecise boundaries, this analogy limits the scope of discussions to the knowledge base of the participants. Increasingly, today’s photographers have no idea what went on in the dark-room. Nor do many realise that inside the dark-room, a great deal of manipulation went on whenever the available budget, skill and demand came together. Retouching prints, negatives and even colour transparencies, multiple printing, masking, re-photography, shadow and highlight recovery, using toners for subtle through to strong split-toning effects are all dark-room techniques which were able to and did lead to substantial changes. These changes were not limited to the general appearance of the image but also could substantially alter content through the addition or removal of parts of the image – often both. And because some of this work required demanded skills and therefore cost a lot meant it did not happen in the mainstream, but for some it was ‘standard’ dark-room manipulation.
Notwithstanding the effect of a possible knowledge gap, undefined references to analogies lead to confusion and error because the lack of precision allows different parties to appear to agree when, in detail, they hold positions which are substantially different. Worse, the protagonists may appear to occupy distant positions when, in fact, basic notions are actually shared but have been calibrated in different ways.
Perhaps worse still, differences in the assessment of the image may become apparent only when a specific image or narrow range of issues is under the microscope. This can lead to embarrassing discoveries about how e.g. competition rules lacked sufficient precision or were open to being mis=applied.
Our debates need a more elaborate, precisely-defined, notion of image manipulation (and its nemetic partner, the set-up up shot as well as related captioning or text support, to be discussed in another paper).
I propose that we use a scoring system based on scales of visible differences to give a measure to our notions of what is allowable in terms of manipulation.
Value of a metric
I hope this can help to locate points of difference as well as common ground when different judges are evaluating photographs. Applying the measure does still call for subjective assessments, but by placing assessments such as ‘just visible’, ‘clearly visible’ on a ordinal perceptual scale in which there can be broad agreement will undoubtedly help debates to agree on a starting position. In particular, this can be achieved relatively easily where the original capture and the submitted image can both be examined at the same time.
For example, in photojournalism, tonal changes which normalise a non-normal scene are generally accepted. For example, an early morning scene in low ambient light with soft, diffused lighting will capture as very low in mid-tone contrast.
We generally accept, for pictorial use, when a photographer boosts mid-tone contrast to bring out details, as this emphasises shadows and brighten brighter areas without overall increase in dynamic range. The strength or effect of the adjustment can vary from the nearly invisible to the clearly visible. We can agree that adjustments with a certain effect are acceptable for a given purpose e.g. photojournalism, but can we can go on to state that any greater effect takes the image into camera club pictorialism which we may decide is not allowable in photojournalism.
For the purpose of evaluating images, institutions may calibrate their metrics by reference to test images; that would be the next step of development for this scheme. It is not intended to set any standards beyond dispute, but to provide a much firmer basis for discussion than different judges arguing in vague abstractions about the differences between ‘moderate’ and ‘strong’, ‘dark-room’ or ‘beyond dark-room’.
• The modifier ‘allowable’ refers to what is permissible, tolerable, admissible, compliant with certain rules or acceptable within a given framework of norms. It does not refer to any aesthetic measure or judgement.
• It would be lovely to be able to aggregate the score for each image and compare it to the allowable sum. This works well for low sums i.e. those intolerant of any manipulation, but could lead to unwanted results where an image use is more tolerant of manipulation. But it is easy to set maximum allowable scores in each parameter.
To what extent is image manipulation allowable?
Score the following on a 10-point scale of 0 (zero) to 9 (nine) as follows: This is the scale that we will apply to each of the separate parameters or characteristics that we wish to adjudicate.
0 = no change at all between the captured image and the image under review
1 = minor global changes reflecting differences between RAW conversion engines
2 = almost invisible change for the purpose of correcting to normal values or machine calibration
3 = just visible change for the purpose of correcting to normal values
4 = visible or light change for the purpose of visual enhancement
5 = clearly visible or moderate change for the purpose of visual improvement just beyond normal values, and without visible artefacts
6 = clearly visible change for the purpose of making image more appealing with values beyond normal but not unnatural or hyper-real, with small visible artefacts allowed
7 = obvious or strong change emphasising a feature or quality for purpose of producing a striking image, with values clearly beyond normal i.e. unnatural or hyper-real
8 = radical change to exaggerate an effect in a feature or quality, with strong differences from parent image
9 = very radical, exaggerated change for the purpose of graphic effect resulting in image markedly different from parent image, even unrecognisable
How this might work
Let’s take three scenarios or work situations in which the extent or degree of image manipulation makes some difference to whether the image is regarded as acceptable i.e. fit for a particular purpose. Let’s keep it simple by aiming for a straight binary ‘Yes’ or ‘No’ decision. And for the present, I assume we don’t need an explanation of what is meant by various photographic qualities such as ‘Curve’ or ‘Highlight/Shadow’. I’ll explain these in more detail later.
Forensic, scientific or evidence-based recording
Where purity or integrity of the actual capture is crucial – as when gathering forensic scene-of-crime records, or scientific experiments subject to third-party scrutiny – we won’t allow any manipulation at all. We cannot accept anything that would disrupt or threaten the integrity of the record (which also forbids changes to file name, and camera metadata). We want to ensure, as it were, as straight and uninterrupted a causal line between the subject and the image as possible.
So we’d want the image to score zero in all qualities. For example, we look at Exposure/Brightness to see if that has been altered. If not at all, that scores zero, and we check the next one. If Exposure/Brightness has not been altered at all, Tone Curve will probably score zero too, as will Highlight/Shadow. But White Balance is independent of these, so could be varied. But we don’t allow any changes here either, so this should also score zero. In the case of raw files, the conversion should, of course, be at the setting used at point of capture.
In some circumstances, we may wish to allow a small adjustment in White Balance, in which case, we may stipulate that the score for this parameter is allowed to score 1.
And so on, through all eleven parameters. You may stipulate, for most strictness, that the total score cannot exceed 1. Or you may relax conditions and allow a total score up to 10 with no category allowed to score more than 1.
Fine art, experimental, graphic design
Let’s go to the other end of the scale. We may characterise allowable changes being those that almost sever the link between the original capture and the manipulated image. In short, anything is allowed. In fact we may not even need to know – or have any interest in – how the final image is created or in how it compares with the original capture or captures. These images may score as much as 99, but in practice can score somewhat less as some parameters are likely to be only lightly adjusted.
So far, this metric – the Angometer – is producing the results we expect and it reflects current norms and practices. The tougher test now is to see how it performs in the middle range.
Photojournalism, news reporting
The stress test for any image manipulation scoring is how it deals with photojournalism and news reporting. Here, we are prepared to accept global changes in tone, exposure, mid-tone contrast, white balance and colour balance, monochrome conversion which broadly enhance the visual qualities of the image without a substantial deviation from the scene as originally captured.
This context therefore covers the middle ground in which some manipulation is allowed – usually constrained to those which shelter under the banner of ‘dark-room effects’ – but no more: that is, not too much. This begs the question of how mush is ‘too much’. Any metric should be able to offer a crisp answer. It should be able to unpick the bundle of ‘dark-room’ effects so that, for instance, we can allow some removal of minor capture defects without stepping onto the slippery slope of allowing cloning in the image.
For example, if in photojournalism and news reporting we wish to keep scores in all parameters to around 5, we are broadly allowing corrections to enhance visual impact without shifting, distorting or abridging content, then a clearly visible application of dodge and burn would clearly score more than 4 or 5. So the controversial Hansen image of World Press Photo might have scored 7 or even 8 to reflect the appearance that the tone changes on the faces of the anguished men suggested auxiliary light sources or reflecting surfaces. Therefore, if World Press Photo forbids scores greater than 4 – 5, a score of 7 or 8 would then press the red buzzer for a closer appraisal.
Still in news and photojournalism, if we look at cumulative scores over the eleven parameters, we could allow an image to score a maximum of, say, 50 so long as no single parameter scores more than, say, 7 or 8. This would describe an image which has received adjustments in exposure, tonality, colour and removal of minor defects, and so forth. But it has not transgressed by making a large change in any feature. For our use we would judge this image OK in terms of post processing.
The Angometer is admittedly rather cumbersome, plus details will vary with the user and the context (e.g. whether it’s applied to a competition for amateurs, or to a journalism award on the world stage).
It does offer an advantage in being able capture some nuances that are easily conflated. For example, while burn and dodge effects are widely acceptable to some degree, the Angometer helps you define just how much dodge and burn is acceptable. It helps to define the difference between a ‘minor’ dodge effect from a ‘major’ dodge effect – one suggesting new light sources or reflective surfaces.
This details how the scale of scoring may be applied to each parameter or image characteristic that we’re interested in. Clearly this list and descriptions is the core of the scheme, and can be rewritten, added to or simplified by the competition or award administrators. It enables administrators to brief judges clearly and explicitly.
(Overall lightness plus mid-tone value: 0 would not allow corrections even if parent image is badly exposed; 5 allows capture errors to be corrected to normalise image i.e. spread Levels histogram to fill the range; 6-8 allows increasingly exaggerated effects with 7-8 being pseudo-high key or pseudo-low key; 9 allows heavy adjustments to e.g. Levels to make image extremely dark or very light).
0 1 2 3 4 5 6 7 8 9
2 Tone curve
(Shape and position of Curves and white and black point globally applied, with no local adaptation: 1 allows application of default conversion to RAW image; 5 allows normalised curve to increase mid-tone contrast; 6 allows boost in contrast to enhance contrast compared to original capture; 7-8 allows boost in contrast in shadow, mid-tone and highlights to obvious change from original capture; 9 allows extreme flattening, posterised effects or bi-level curve, even reversal)
0 1 2 3 4 5 6 7 8 9
3 Highlight and shadow
(Adaptive highlight and shadow control or tone-mapping applied over whole image (not burn/dodge or tone curve: 0 forbids any compensation or recovery even if done in-camera; 1 allows just-visible recovery of highlight and shadow as applied by default RAW conversion or low levels of e.g. DRO (Dynamic Range Optimisation) or Highlight Tone Priority; 3 allows visible recovery but without appearing to alter lighting effect; 5 allows visible recovery that maps shadows and bright areas to mid-tone; 6-7 allows increasingly strong recovery of shadows and highlights mapping dark and bright zones to mid-tone; 9 allows exaggerated recovery including edge artefacts and increase saturation in recovered areas.)
0 1 2 3 4 5 6 7 8 9
4 White and colour balance
(Colour temperature and magenta-green correction: 3 allows incorrect WB setting to be corrected in main lighting, even if it distorts another light source. 5 allows correction of secondary sources to normal i.e. two or more points of correction. 9 allows dramatic change to image WB e.g. to turn daylight balance to tungsten. )
0 1 2 3 4 5 6 7 8 9
5 Colour saturation
(Either increase or decrease in saturation in all or specific wave bands. 2 allows slight increase or decrease to render image closer to scene as captured. 5 allows visible increase or decrease to improve photographic effect. 6-7 allows obvious increase or decrease for deliberate emphasis or exaggerated effect. 9 allows extreme change – to posterised colours or black-and-white. )
0 1 2 3 4 5 6 7 8 9
(Noise reduction applied post-capture to JPEG or TIFF images (some cameras reduce noise as part of RAW capture): 3 allows noise reduction just visible at 100%; 5 allows moderate noise reduction visible at 100% with just-visible detail smoothing; 6-7 allows strong, easily visible noise reduction or introduction of grain e.g. pseudo-film effect with detail smoothing; 9 allows aggressive noise removal or addition of noise or film-like grain which leads to easily seen smoothing in case of the former, or the easily seen obscuring of detail in case of the latter.)
0 1 2 3 4 5 6 7 8 9
(Matrix convolutions e.g. USM or similar which operate over entire image: 1 allows very light overall, but non-adaptive, sharpening, 4 allows light adaptive sharpening visible only under zoom; 5 allows just-visible sharpening e.g. for online use; 6-7 allows visible, adaptive sharpening with just-visible artefacts; 9 allows exaggerated sharpening with easily visible halo artefacts.)
0 1 2 3 4 5 6 7 8 9
8 Local tone
(Burn, dodge, local curve and similar localised or brush-limited tonal controls; 1-2 allows just-visible changes in tone distribution; 5 allows visible changes in tone distribution to enhance the image without suggesting changes in lighting compared to original e.g. enhancing catchlight in eye; 6 allows clearly visible, 7 allows very visible adjustments which also imply or suggest changes in light source or reflecting surfaces e.g. burn in shadows to remove detail; 8: allows obvious visible adjustments including easily visible vignetting or reverse vignettes; 9 allows extensive tonal changes implying changes in light sources, new light sources or reflecting surfaces or obvious vignette/reverse vignette .)
0 1 2 3 4 5 6 7 8 9
9 Monochrome and toning
(Conversion of the RGB capture to black-and-white or monochrome, with or without toning and film-like effects. 0 allows conversion to black-and-white by desaturation, 1 allows conversion to black-and-white using trichromatic response of eye, 2-4 allow increased emphasis on limited tone bands, 5 allows pictorially acceptable results similar to those obtained with black-and-white film photography, 6 allows introduction of light toning tints similar to those of light dark-room toning e.g. selenium, 7 – 8 allow increasing stronger colours and tonal variation from original e.g. to simulate sulphide, sepia toning, 9 allows metal substitution or dye toning e.g. gold, ferricyanide, Colovir resulting in srong colour and tonal changes.
0 1 2 3 4 5 6 7 8 9
10 Small defects
(Clone or repair effects applied to specific defects: 0 forbids any change, including removal of e.g. dust specks or hair; 3 allows removal of dust and hair specks but no more; 5 allows removal of just-visible defects e.g. stray hair, but not e.g. freckles or other unwanted element; 6-7 allows cosmetic removal e.g. freckles, wrinkles without strongly altering character of face and skin; 8 allows removal of non-essential but distracting element e.g. telephone wires in sky, telegraph pole behind head; 9 allows extensive removal causing an obvious alteration in character of face and skin or unwanted element e.g. figure in background.)
0 1 2 3 4 5 6 7 8 9
(Copying of part of image, whether from the same image or another, to replace parts of the original image: 3 allows just-noticeable cloning to incidental or peripheral detail which does not substantially alter content of image e.g. tree or cloud at edge of image, tear in background paper; 5 allows noticeable cloning to repair error e.g. JPEG or sensor read-out error; 6-7 allows substantial change e.g. adjustment of model’s pose, tidying up drape of clothes; 9 allows any level of cloning from multiple images.)
0 1 2 3 4 5 6 7 8 9
Please! Comments and suggestions on this, so we can work together to develop and improve our discussions. Thanks for getting this far (assuming you didn’t cheat! ;) ). It’s been a long post.