'Siberian Chiffchaffs' : colour issues & photographs
Light conditions and camera issues
When observing any bird, it is apparent that its appearance varies, being affected by variations in light conditions, the angle of the bird and even the surrounding vegetation, which can not only influence the way that the eye interprets the colours of the bird but can also impart a colour sheen to areas of the bird's plumage. Photographs are even more susceptible, as the camera white-balance and the 'demosaicing algorithm' will influence the result. Also, colour 'balance' (or density) tends to differ consistently between different makes of camera. All other things being equal, Nikon DSLRs appear to produce paler and greyer images while Canon DSLRs produce darker and browner images. There are no such things as 'correct' or 'absolute' colours in a bird's plumage, of course, but there are certainly distortions and artefacts which need to be avoided if the eye and camera are to capture 'representative' and 'full' colour values. In RGB capture and display of colours (as in digital cameras and on computer screens), quite small variations in the proportions of red, green and blue can produce quite significant shifts in the displayed and perceived hues (see RGB analyses below).
When a species’ identification depends upon subtle colour hues, ‘random’ or 'isolated' digital images are of little use. Photographs taken in overly bright light are likely to suffer from over-exposure and colour dilution or loss ('burn out'). The underparts of paler species can easily appear overly white. The portrayal of upperparts may well be glazed, with a surface 'bloom' or 'sheen' of light which masks completely the true underlying hues of the feathers. Photographs taken in heavy shade will lose colour subtleties and gradations. The upperparts can appear as a uniform 'slab' of monochrome hue, with colour highlights and variations between feather tracts lost. Images of a given individual can vary widely in such circumstances and, in extreme cases, a photograph of a Chiffchaff can even suggest the 'wrong' subspecies. In diagnosing 'Siberian Chiffchaff' from photographs, this is a serious source of problems. Photographs taken at long range are particularly suspect, as they are more prone to colour distortions, while certain hues are more visible at a distance than others (to the eye as well as the camera). Further comments on this issue, and examples of varying images, were included in Noeske and Dean (2006). Written several years ago, some of the captions in that feature now need revision, with increased knowledge about tristis, but the plates still provide useful illustrations of how inconsistent photographs can be. Even at close range and with in-the-hand photos, colour depiction can be volatile. Those who trap and photograph birds regularly try to reduce such effects by taking images against a 'neutral grey' card and ensuring that the bird and card are in acceptable light conditions (soft lighting or slight shade are generally best). Including a Kodak grey-scale in the image also enables an objective evaluation to be made.
In the field, it is not possible to choose the light conditions, of course, nor the bird's surroundings. However, even if not expert in optics and colour rendition, we all appreciate that fleeting impressions are unreliable and - if light conditions are adverse - it is generally evident at once that the appearance of the bird is being distorted. If a Chiffchaff remains positioned in unhelpful light-conditions, it will simply be impossible to evaluate plumage hues adequately. In the vast majority of cases, however, a Chiffchaff will be on the move and will be observed in more favourable (or neutral) light conditions in the course of observation. I have very rarely encountered a Chiffchaff in which radical variation in appearance was incessant and for which a settled appearance did not emerge eventually. Thankfully, the human brain accumulates data, can then evaluate variation and its causes, eliminate extremes, and reach a balanced assessment over time. If a Chiffchaff is observed for a considerable period, it will be very unlikely indeed that such a settled appearance does not become evident. Conversely, a single photograph captures just an instant in time (literally a fraction of a second). To publish two photographs of a single individual which portray radically different appearances, and infer that each has equal 'value' or is equally representative is misleading. This happens not infrequently and can exaggerate the problem described as 'plumage morphing' (for discussions on 'plumage morphing' see the main text here, and also this thread and this thread on the web-forum Birding Frontiers). Under the requisite prolonged, close-range and careful observation, the characteristic appearance will nearly always become apparent – not ‘inflexible’ but broadly accurate and reliable. Having made this evaluation, then a long-series of photos is required, to provide a good chance of obtaining some which are free of the problems discussed above and which match the appearance of the bird as carefully evaluated in the field. It is clearly tempting to publish each and every photograph of an 'interesting' species. However, publication of photographs with an appearance not validated by careful field observations should be avoided, as such images portray artefacts which impair rather than facilitate discussions.
Fresh plumage versus moult and wear
When using photographs to assess the plumage features which distinguish tristis from other subspecies, it is also important to use images of individuals in reasonably fresh plumage. As birds wear, critical hues can fade and disappear while - as some colours wear and fade more than others - some hues may even become more apparent, owing to enhanced contrast with faded areas in the plumage. During mid-summer, for example, some individuals may become so faded and deficient in colour that it is impossible to assign them to any particular subspecies on the basis of plumage colour.
In a UK context, 'Siberian Chiffchaffs' are unlikely to be encountered in mid-summer. At least from their arrival in autumn (October/November) until mid- to late-winter, their characteristic appearance should be evident. However, the pre-breeding moult of tristis is relatively late compared with abietinus and collybita and, in consequence, tristis encountered during March and April can appear confusingly 'washed out' and dishevelled as a result of colour-loss and body-moult. (In fact, the state of moult can be a useful ancillary factor in identification.) Temporary colour loss in this state can produce a much 'greyer' appearance than before and after the moult is completed. The two images below illustrate just how radically the appearance of an individual can change at this time. Thus, evaluation of the natural hues and 'plumage limits' of tristis must be made only on the basis of individuals in reasonably fresh plumage.
|November 2007||April 2008|
Plate 1. Siberian Chiffchaff, Portland, Dorset, winter 2007/08
The two images above show the same individual, which wintered at Portland Bill, Dorset in 2007/08. In the left-hand image, taken in November 2007, it is in fresh, definitive plumage and the 'natural' and characteristic hues of tristis are well displayed. Its identity as tristis is evident. By April of the following year, when the right-hand image was taken, the bird was in moult, including body feathers and with missing/growing tertials and central tail-feathers. In this condition, there has been considerable loss of the natural hues of fresh plumage and its identity is much less evident. These two images illustrate how important it is to evaluate hues on fresh-plumaged individuals. When birds are worn or are in heavy moult, their appearance can be deceptive and evaluation of their 'natural' colours becomes difficult or impossible. Thus, generally, autumn and early winter individuals are easiest while spring individuals can be more problematic.
See also Case Study 4 and Case Study 5.
All these issues have been taken into account when choosing photos to include in the Gallery of images of Siberian Chiffchaffs.
HSB values in three photographs of Chiffchaffs
There is perennial discussion of the appropriate 'colour nomenclature' in describing Chiffchaffs and, in particular, what constitutes a 'grey' Chiffchaff. The photos below depict three Chiffchaffs carefully studied in the field or in the hand. There is certainly an element of semantics in such definitions but, applying my own short-hand colour nomenclature, they portray good examples of the 'olive and yellow' livery of a typical nominate collybita, the relatively 'grey and white' livery of enigmatic 'Bonelli's-like' Chiffchaffs and the 'brown and buff' livery of a typical tristis : i.e. individuals with appearances which figure prominently on this website. It is worth at this point repeating the explanation of this 'short hand nomenclature', which was included at the outset in the introductory note on colour nomenclature in the main text:
'In my own texts, I employ shorthand terms for certain colour liveries, such as 'olive and yellow', 'brown and buff' and 'grey and white'. Clearly, there is nothing very 'nuanced' about these terms and they are no more than labels of convenience. They are relative, not absolute terms, designed to indicate the key colours and shades which distinguish the Chiffchaffs to which they are applied. In a wider context, 'grey' can span any shade between white and black, as in the 'Kodak Grey Scale' below. In the term 'grey and white' livery, the label 'grey' is intended to convey a prominence of a 'stony grey' shade (nomenclature again!). It denotes a shade resembling 'pale neutral grey' in Smithe's 'Naturalist's Color Guide' or shades in the range 2 to 3 on the Kodak Grey Scale i.e. similar to or a little darker than the outer borders to the scale in the image below). 'Buff' is here used to designate the hue which infuses the breast and flanks of a truly 'classic' Siberian Chiffchaff. In this usage it lacks yellow and has a slightly 'rusty' tinge (in more detailed discussions in the text, the term 'rusty-buff' is often employed). Thus, it signifies a somewhat warmer shade than is sometimes inferred by the term 'buff'.
● It is essential that all these shorthand terms are interpreted with reference to the definitive and carefully chosen photographs from which they derive their meaning (see Plate below).'
Of course, any colour other than a fully saturated hue includes an element of grey. In the HSB (Hue, Saturation, Brightness) colour model, for example, hues are reduced from full saturation by the addition of grey. S represents the extent to which a hue approaches full saturation while its inverse indicates the level of grey. Thus, an S value of 80% means that a hue is 80% saturated, with a 'grey level' of 20%. How the eye perceives the resulting colour will also depend upon the shade of 'grey' (paler, darker), how visually dominant is the 'hue' (red is a dominant colour, for example, while green is a softer and less dominant hue) and how intimately component colours are admixed.
Plate 2. Three Chiffchaff liveries.
Even after careful assessment in the field, a photograph will never provide a thoroughly accurate 'colour map' of the individual, as it is a simulation using photographic algorithms (and see discussion above). However, the appearance of each individual in these photos has been judged to be a good representation of its field appearance. It is instructive, then, to examine the 'pixel maps' and HSB values involved in these photographic depictions.
Olive, yellow-olive and greyish-olive hues dominate the pixel map. The HSB
values indicate a 'grey level' of 46%.
The nearest match in Smithe's 'A Naturalists' Color Guide' is 'Olive-Yellow'
Bluish-grey, greyish-olive and olive hues dominate the pixel map. The HSB values
indicate a 'grey level' of 79%.
The nearest match in Smithe's 'A Naturalists' Color Guide' is 'Olive Grey'
Greyish-brown and buffish-grey hues dominate the pixel map. The HSB values
indicate a 'grey level' of 74%.
The nearest match in Smithe's 'A Naturalists' Color Guide' is 'Drab'
Thus, the computer colour simulations include a significant 'grey' component in the upperparts' hues of both the 'grey and white' Chiffchaff and the Siberian Chiffchaff (79% and 74%, respectively). In nominate collybita the level of grey is much less, at around 45%. These percentages can be used to reconstruct the mantle hues of 'olive and yellow', 'grey and white' and 'brown and buff' Chiffchaffs from their H and S values, with the B (brightness value) retained (Plate 6).
Plate 6. HSB reconstructions of mean mantle colour, sampled from photographs of a nominate collybita, a 'grey and white' Chiffchaff and a Siberian Chiffchaff (tristis).
Thus, in the HSB colour model, the nominate collybita has a yellowish-olive hue of 54%, diluted by a grey-level of 46%. The nearest match in Smithe's 'A Naturalists' Color Guide' is 'Olive-Yellow'. The 'grey and white' Chiffchaff has a greenish-olive hue of 21%, diluted by a grey-level of 79%. Appropriately again, the nearest match in Smithe's 'A Naturalists' Color Guide' is 'Olive Grey'. The Siberian Chiffchaff has a russet hue of 26%, diluted by a grey-level of 74%. The nearest match in Smithe's 'A Naturalists' Color Guide' is 'Drab', which in colour terminology denotes a 'grey-brown' colour but has other less appropriate colloquial connotations.
The mean mantle colours of the collybita and the 'grey and white' Chiffchaff combine (different) shades of olive with grey but with a significantly higher 'grey level' in the latter (79% c.f. 46%). In contrast, the mantle of the tristis combines russet and grey in place of olive and grey - but with a significant 'grey level' approaching that of the 'grey and white' Chiffchaff. In the tristis, the russet hue at 26% saturation produces a distinctive 'tan-brown' colour, which the other two Chiffchaffs lack entirely. This confirms that there is more to the appearance of tristis than an absence of 'misplaced' yellow and olive.
Note also that, in the collybita and the 'grey and white' Chiffchaff, converging olive hues are reflected in the similar RGB values whereas, in the tristis, there is a significantly lower G (green) component. Although both have a high 'grey level' in their composition, visually the mantle colour of the 'grey and white' Chiffchaff blends with 'neutral grey' while the mantle colour of the tristis is evidently more contrasted (Plate 7).
Plate 7. Comparison of mean mantle colour and neutral grey component in a 'grey and white' Chiffchaff and a 'Siberian Chiffchaff'.
Colour-corrected images from 'Siberian Chiffchaff revisited'
Unfortunately, the photographs of specimens which appeared in the paper 'Siberian Chiffchaff revisited' (Dean & Svensson, 2005, British Birds 98:396-410) were badly underexposed and suffered from correspondingly inadequate colour reproduction. Colour-corrected versions are reproduced below.
Siberian Chiffchaff, Irkutsk, Siberia, May.
Photo copyright Natural History Museum. London.
Plate 237 from 'Siberian Chiffchaff revisited', Dean & Svensson, 2005, British Birds 98: 396-410.
Siberian Chiffchaffs, Krasnoyarsk, Siberia. Upper, May. Lower,
Photo copyright Natural History Museum. London.
Plate 238 from 'Siberian Chiffchaff revisited'. Dean & Svensson, 2005, British Birds 98: 396-410.
Siberian Chiffchaff, India, January.
Photo copyright Natural History Museum. London.
Plate 239 from 'Siberian Chiffchaff revisited'. Dean & Svensson, 2005, British Birds 98: 396-410.