Chalky Flesh Tones Under Mixed Warm and Cool Light
A four-hour acrylic examination requires a seated portrait, 50 × 70 cm, from a model lit by a 60 W tungsten lamp placed about 1 m to the model's right at shoulder height, with a large north-facing window behind and to the left giving cool daylight. The candidate may use only five colours: titanium white, cadmium yellow medium, cadmium red, ultramarine blue and burnt umber. Two hours in, the lit side of the face is a pale pink-white with no visible form, the midtones are a chalky beige, the shadows are a grey-violet that reads as dirty, and the shadow side of the head merges into the dark background. The candidate's response has been to keep adding white to 'brighten the face', which has made the problem worse, and they now propose to add black from a neighbouring student's set to 'put the darks in'.
1. Explain why white is causing chalkiness here, and name the variables the candidate should be controlling instead.
Read the worked answer
White in acrylic is a high-opacity, high-value pigment with almost no hue of its own. Adding it raises value while simultaneously lowering saturation, so every area it touches drifts toward the same pale neutral. Because flesh in warm lamplight is not a high-value, low-saturation colour but a mid-value, high-saturation one, whitening produces the double fault the examiner is seeing: the lights get lighter but not warmer, and the colour differences between light, midtone and shadow collapse into one family of pinks and beiges. The form disappears for a simple reason - if every plane is the same value, there is nothing for the eye to read as a turn.
The variables that actually need controlling are three:
- Value, but achieved by dilution and layering, not by whitening: thin the mixture with water or medium to let the ground show, or glaze a warm transparent passage over a mid-value base, so a light can be light without being chalky.
- Hue temperature, which is doing most of the descriptive work here. Under tungsten the lights are warm (yellow-orange), and the shadows of a warm-lit form are comparatively cool. The portrait's structure is carried by that warm-cool opposition, not by lightness alone.
- Saturation placement. Keep the highest chroma in the midtones and in the transition zones - the cheek, the ear, the nostril - and let the lights desaturate slightly toward the highlight, exactly as real skin does.
The proposal to borrow black should be refused. Black is a dead neutral that greys every mixture and kills the violet-grey that shadow needs; the palette already contains the correct darks (ultramarine plus burnt umber) and adding black would make the muddiness permanent rather than correctable.
2. Give a concrete mixing plan, using only the five permitted colours, for the light, midtone and shadow of the face, and justify each choice.
Read the worked answer
The logic is to build three families, not three single colours, and to keep the mixtures related so the face reads as one substance.
Light family (warm, high value, moderate chroma). Start with white plus a small amount of cadmium yellow medium and a touch of cadmium red. To stop it going chalky, keep the yellow dominant over the red and add a pinhead of burnt umber to knock back the purity - umber supplies the slight greenish-grey that keeps flesh from looking like icing. The very highest highlight should be the least saturated mixture on the face, not the whitest paint applied at full strength.
Midtone family (the chroma carrier). Cadmium red plus cadmium yellow medium in roughly equal parts, cooled with a very small amount of ultramarine, then lightened slightly. This is where the model's actual skin colour lives, and it should be the most saturated passage on the canvas. Keeping the midtone saturated is what makes the light side look luminous by comparison.
Shadow family (cool, lower value, still chromatic). Ultramarine plus burnt umber gives a warm-cool grey that is far more useful than black; add a little cadmium red so the shadow stays fleshy rather than stony. Under tungsten the core shadow should lean cool-violet, and the reflected light inside the shadow - bounced from the lit shirt or table back into the jaw - should be warmer and slightly lighter than the core.
Justification: because the light is warm, the shadow must be comparatively cool, and because the paint is opaque acrylic, the value range must be built from the darkest mixture upward rather than rescued by white. Mixing the darkest note first and testing it against the background prevents the whole painting from being keyed too light, which is the underlying cause of the chalky read.
3. The shadow side of the head is merging into the dark background. What should be changed, and what should not?
Read the worked answer
Merging is a value-and-edge problem before it is a colour problem, and the candidate should diagnose which of the two is operating, because the fixes are opposite.
If the shadow side and the background are the same value, no amount of hue change will separate them at three metres; the separation has to come from a value step. The correct move is usually to lighten the background behind the shadow side - not to darken the shadow, which would flatten the head into a silhouette and destroy the reflected light that makes the form turn. A background that is dark behind the lit side and lighter behind the shadow side produces the standard 'light against dark, dark against light' reading and makes the head appear to come forward.
If the values already differ but the edges are identical - a hard shadow contour against a hard background contour - then the fix is edge control. In front of the model, the shadow side's outer contour is typically the softest edge on the head, because the form is turning away from both light sources; painting it as a soft, slightly broken edge separates it from a background that can remain comparatively crisp.
What should not be changed is the temperature logic. Candidates often 'solve' merging by warming the shadow, which destroys the warm-light/cool-shadow relationship the whole portrait depends on. The shadow stays cool; it is the background that moves. A useful test is to look at the two areas through half-closed eyes: if the boundary vanishes, it is a value error and the background must be adjusted.