The Subtleties of Color
Small Differences That Make All the Difference
Every painter knows the dance, taking a few steps back from the painting, their head tilted slightly askew, the eyes pulled tight into a squint, or the hand held in front to block off an area from view. The to and fro of action and adjustment, of sense and sensibility. And if color happens to be foremost in that dance, the search is often for still unseen subtleties to be coaxed from the colors at hand. Even if the effect is one of clash, there remains the desire for it to be 'just so'; a clash tuned to the highest pitch or given a particular piquant flavor. In the end, artists thrive on subtleties, on the small differences that make all the difference, and their search for colors that can respond to those needs is endless.
All these immeasurable things, no matter how rarefied they might seem, ultimately have a material basis rooted in the nature of pigments and paint, the common tools of the trade. In the pages ahead we will start by reviewing some of these underlying factors and then examine why some of the most subtlest of variations can make all the difference in choosing the right color matching your intent.
Beyond All Measure: the Limitation of Colorimetry
When trying to describe what makes any particular color unique it is tempting to point to a color's location within a well defined system such as CIE L*a*b* or Munsell. Doing so allows us to feel that we can place the color's uniqueness within a mapped and measured space, and even calculate the degree of difference it has from all those other colors that jostle for a treasured spot of their own. But we would quickly learn that it is precisely those subtleties that are lost in the process. For all the accuracy of our spectrophotometer in reading the exact makeup of the light being reflected back from a sample, nothing in that information really tells the artist what they need to know about actually using that paint: how it mixes with other colors, its degree of opacity, tinting strength, or any number of physical attributes. Paint is ultimately color on the move, dynamic and energetic, and no single snapshot can capture that more vibrant life lived on the palette of the studio.
While clearly no system of measurement is perfect, spectral reflectance curves perhaps come closest to capturing the nuance of a color, especially if the spectrum is available for both masstone and a tint of a known percentage. With these two points as reference one can roughly gauge how a particular paint might perform in mixtures. However, care must be taken as even here there are difficulties. Cadmium and Hansa Yellow Medium, for example, might share nearly identical spectra at full strength but no one would mistake them for the same in practice. Additionally, one must factor in the responsiveness of the eye to various wavelengths, as this can greatly shape how the eye perceives the color -- which can be very different from the data itself.
Underlying Causes of Subtle Differences
Nearly all of the subtleties of a particular color can be traced back to the physical attributes of the pigments and the way light interacts with those particles within a paint film through absorption, reflectance, scattering, and transmission. Pigments, in turn, are largely characterized by their underlying chemical composition, along with such factors as particle size, refractive index, and scattering coefficient, while paint films impact color through their pigment load, thickness, and sheen. Ahead we will touch on all of these briefly, as a way to introduce some of the complexity behind many of the subtle differences we see.
Physical Properties of Pigments
Crystalline Structure of Pigments
All pigments, with rare exception, have crystalline structures that dictate their color, and even small changes at this level can alter which wavelengths are absorbed or reflected. Phthalocyanine Blue, for example, has two types of crystal formations (α and β ) that are responsible for their slight leanings toward the Red or Green Shade, while changes to the crystal lattice of Quinacridone is responsible for its broad range that runs the gamut of bright Quinacridone Reds to the ever deeper Magentas and Violet. A third example includes the entire array of Cadmium colors, where cadmium sulfide, which is yellow in its pure state, is made progressively redder and deeper by replacing the sulfur in the crystal lattice with increasing amounts of selenium. This substitution broadens the amount of the spectrum that can be absorbed, and if enough selenium is added, Cadmium can actually appear black.
Transparency/Opacity and Tint Strength
A particle's opacity is greatly dependent on its ability to scatter light, which relies primarily on a particle's refractive index and size. The larger the difference in the refractive index between a particle and its surrounding medium, the more light is scattered and the underlying layer obscured; a phenomenon similar to the way fog scatters a car's headlights. Conversely, the closer these numbers are, the more transparent a particle will appear. The high refractive index of Cadmium Yellow and Titanium White, for instance, is almost solely responsible for their tremendous hiding power and sense of opacity, while Zinc White and Hansa Yellow appear more transparent because their refractive indexes are considerably closer to that of an acrylic polymer. Because dark pigments with low refractive indexes, such as the Phthalocyanines, do not scatter much light, their hiding power resides almost completely in their ability to absorb light, the pigment loading, and the thickness of the film.
The other aspect of particle size has equally dramatic consequences on both scattering and tinting strength. As a particle becomes smaller it scatters light more effectively until a certain optimal size is reached, after which this aspect begins to drop off sharply. As one continues further below this threshold, the pigment particle grows increasingly transparent while simultaneously reaching a maximum of tinting strength. Here is where the magic of the Transparent Iron Oxides reside, as the normally opaque iron oxide pigments are manufactured to such small particle sizes that they become wonderfully translucent and far more effective in glazing and the production of cleaner, higher chroma tints. Titanium White, on the other hand, is carefully manufactured to optimize its particle size for maximum light scattering, and hence opacity. In fact, a one centimeter wide crystal of Titanium Dioxide is completely transparent, and it is only as the crystals get smaller that scattering becomes dominant and we sense the pigment as inherently 'white'; an effect similar to the whiteness of finely ground glass. Should the Titanium Dioxide be ground even further down to a nano-particle scale, it would actually become completely transparent, a feat that seems almost magical given how strongly we associate opacity with Titanium White.
Purity and Uniformity
Differences in the chemical purity of a particular pigment, as well as the uniformity of its shape and size distributions, are responsible for still other quirks of coloration. For example, natural earths owe their particular flavors and nuances to varying amounts of trace elements, such as manganese oxide, silica, alumina and clays, as well as their wide assortment of particle sizes. While this accounts for many of their prized undertones, and explains why particular regions in the world become coveted for their mined ochres, siennas, and umbers, it is also the reason why these colors are generally weak tinters and lower in chroma than the parallel range of synthetic oxides. Also, because they are mined, these pigments have a wide lot to lot color variation depending on the level of impurities in the next shovel full. Ultramarine Blue presents another example; one of the earliest synthetic pigments, it is richer and more saturated than the genuine Lapis Lazuli it replaced, which as a mined rock always came with impurities of calcite, sodalite, and pyrite, that muted its tone.
The Physical Properties of Paint
As most artists know, colors do not necessarily stay the same through thick and thin. In thick films of densely packed pigment, the masstone is dominant and the color will appear more saturated and deeper. As the film becomes thinner, the undertone becomes more pronounced and the overall color can appear more transparent, lighter in value, and sometimes higher in chroma as well, assuming the underlying substrate is very light in tone. These effects are ultimately caused by having an increased amount of light reflected from both the pigments and the underlying substrate in the form of backscattering.
Beyond film thickness, simply altering the pigment load or density in a paint film can markedly change the perception of a color. For example, in a film of densely packed translucent pigments, much of the interior scattering and transmittance of light can be lost through subsequent and repeated absorption, and one primarily sees just the reflected light coming from the surface. This reduction in light reads as a deepening in hue and a reinforcement of the dominant absorption band. As pigment load is decreased, and light begins to penetrate through the film, the interplay of scattering and absorption has a larger impact on the overall color. One can imagine a similar effect if placing identical sheets of stained glass on top of each other, one after another. As the pile grows thicker, the color will get increasingly deeper and more saturated.
At its most extreme, the spectral reflectance curve can change considerably as more and more light is able to penetrate deeply into the material. This phenomenon can be seen in such transparent colors as Green Gold and Nickel Azo Yellow, where a dramatic difference emerges between the mass and undertone, as well as a subtler shift in spectra for Phthalocyanine Blue G/S.
Sheen and Surface
Whether a surface is glossy or matte, smooth or textured, will ultimately impact a color's expression as well. As a paint film becomes glossier and smoother, there is less scattering of light at the surface and more penetration and absorption of the light by the pigments themselves. This causes darker colors to typically appear deeper and more saturated when they have a gloss sheen, and conversely, appear to lighten if matte; not unlike the phenomenon of removing a darker colored stone from the bottom of a riverbed and watching the seemingly rich color dissolve before ones eyes with the evaporation of the water.
The spectral data used in the following case studies was obtained using a Minolta® Spectrophotometer. Samples were as 10 mil drawdowns on lacquered cards, with each color represented both at full strength and mixed with varying percentages of GOLDEN Regular Gel (Gloss) or Titanium White. Many of the graphs used in this article are spectral curves, which might be unfamiliar to many artists as they are not that common outside of laboratory settings. The easiest way to understand them is simply as showing the amount of light that is reflected from the surface for each wavelength in the visible spectrum. The more that is reflected, the higher the curve will be at that point. To make the readability a little easier, along the top of each graph we include markings showing the approximate range for each band of color, running from Violet through Blue, Green, Yellow, Orange, and Red. The x-axis, running along the bottom, is marked with the actual wavelengths themselves.
|Quinacridone Violet and Quinacridone Magenta Masstones|
Quinacridone Violet and Quinacridone Magenta Masstones
Quinacridone Magenta and Violet are one of those common cases of colors that seem so close together surely it couldn't matter all that much which an artist reaches for. Of course the answer depends somewhat on your needs. Quinacridone Violet is more opaque and bluer in the undertone than its more transparent, redder cousin. While these features go easily unnoticed when used full strength, the subtleties become much more pronounced when tinted or used in transparent glazes, as can be seen in the spectral graphs. Notice how the spectral curves of the masstones of Quinacridone Violet and Magenta (Graph 1) have virtually identical shapes, with almost negligible levels of reflectance from Violet all the way through Yellow (400-600nm) until finally rising sharply within the cooler, outlying regions of Red. If one looked at these two colors, it would be difficult to tell them apart. However, mixing these colors 1:10 with Titanium White or 1:50 with Regular Gel not only dramatically changes the shapes of their respective spectra but clearly highlight their differences as well. In both the transparent let downs (Graph 2) and the tints (Graph 3) of these colors one can 'see' the warmer aspect inherent in the Quinacridone Magenta, where its spectra now rises much earlier, indicating a new more orange component, while the Quinacridone Violet continues to exhibit strong absorption even past 600nm.
|Quinacridone Magenta Mixed 1:50 With Regular Gel Gloss and 1:10 Titanium White|
Phthalo Blue (GS) (PB 15:3) / Phthalo Blue (RS) (PB 15:1)
|Phthalo Blue (RS) and Pthalo Blue (GS) Mixed 10:1, 3:1, 1:1, 1:3, and 1:10 with Titanium White|
Carbon Black (PBk 7) / Mars Black (PBk 11)
|Carbon Black and Mars Black Masstone and Mixed 1:50 With Regular Gel Gloss|
When Colors Coincide
|Phthalo Blue (GS) Mixed 1:3 with Hansa Yellow Medium and Cadmium Yellow Medium|
Phthalo Blue (GS) Mixed 1:3 with Hansa Yellow Medium and Cadmium Yellow Medium
The Hansa Yellows sit across from the Cadmiums like a row of twins arriving late and uninvited to a family dinner. Contrary to many notions these are not the poor substitutes for the 'real' thing, but truly flushed with their own sense of flash and purpose within the painter's toolbox. The Hansas might not have the opacity of the Cadmiums, but their transparency allows them to be an essential ingredient for transparent glazes, deep greens, and composite blacks. For all the brash and brawn of the Cadmiums, the Hansas speak in their own bright voice.
A good way to experience these subtleties is to watch how the two colors impact various mixtures. With a Phthalocyanine Blue or Quinacridone Magenta, for example, Cadmium Yellow Medium creates dense lighter-valued tints with a sense that white has somehow strayed into the mix. In the accompanying graph (Graph 7), notice how it produces a sharp spike in value after 500nm, and an elevated reflectance throughout the oranges and reds. With Hansa Yellow Medium, on the other hand, the saturation of Phthalo Blue (GS) is largely preserved and the hue is simply shifted towards green with a minimum increase in value. What is not as well shown here is the fact that the translucency is held onto as well, the mixture remaining ideal for glazing and developing other rich, dark greens.Link to additional supporting visuals
Pyrrole Red (PR 254) / Cadmium Red Medium (PR 108)
Like the Hansas, the drama of the Pyrrole Reds and their Cadmium doppelgangers is often played out through mixtures more than masstones. At full strength both Cadmium Red Medium and Pyrrole Red are opaque with spectra that closely echo each other with just the slightest amounts of difference, causing Pyrrole Red to appear slightly warmer and higher in chroma. Little there, however, prepares us for the sharper divergence that occurs in the tints with Titanium White. The Pyrrole and Cadmium are now equivalent for the majority of red wavelengths, though Pyrrole Red still maintains a greater reflectance in the oranges and, most significantly, shows much stronger absorbance throughout the greens and on into the blue regions before a slight bump up in the violet. As a result, the Pyrrole Red tint appears richer, deeper and less chalky then the more neutralized Cadmium Red Medium. As one can imagine, these differences in tint strength and chroma get repeated in nearly every mixture where these colors play a major role.Link to additional supporting visuals
Seeing Through Opaque Pigments
This section begins with pairs of synthetic and natural iron oxides whose differences revolve almost entirely around particle size, with the synthetic oxides being exceptionally small when compared to the usually chunky, larger-scaled pigments of natural earth colors. As mentioned earlier, when particles grow smaller not only does the total surface area increase rapidly, but their ability to scatter light diminishes as well. With scattering held to a minimum, the pigments' interaction with light is solely through absorption and reflection, which both maximizes their tinting strength and increases their translucency. As a result, the synthetic oxides will often be the preferred choice when needing brighter mixtures and cleaner glazes, while the more standard earths can provide a wonderful opacity and density when relying on their masstone.
Burnt Sienna (PBr7) / Transparent Red Iron Oxide (PR 101)
|Transparent Red Iron Oxide and Burnt Sienna Mixed 10:1, 3:1, 1:1, 1:3, 1:10 with Titanium White|
In this grouping, well-known Burnt Sienna is contrasted with the similarly hued Transparent Red Iron Oxide. Both start off as mid-toned earths, the Burnt Sienna a touch brighter and with a slightly higher reflectance in the warmer orange to yellow range, while the Transparent Red Iron Oxide reads as a ruddy and rich mahogany brown, with its peak reflectance deep within the cooler range of reds. None of that, however, quite prepares one for the transformations that happen when the samples are tinted with white or mixed with gel to form a glaze. As the graph shows (Graph 10), for example, Transparent Red Iron Oxide jumps dramatically in Chroma, or Saturation, even when mixed as high as 1:1 with Titanium White. By contrast, Burnt Sienna remains very low in Chroma, never raising much beyond its starting point, as it forms the tell-tale cold and pasty pastels of nearly every brown when mixed with white alone. Similarly, when making glazes, the Transparent Red Iron Oxide blooms into rich browns with bright, fiery undertones of orange while the Burnt Sienna will always carry a slight sense of murkiness.Link to additional supporting visuals
Yellow Oxide (PY42) / Transparent Yellow Iron Oxide (PY42)
Yellow Oxide and Transparent Yellow Iron Oxide have differences that are a little obscured, perhaps, by their identical Color Index designation as PY42. In fact, many artists assume far too often, that pigments with matching Color Index names are unvaryingly the same. But nothing could be further from the truth, especially if reaching for a yellow earth for glazing or to use in tints or mixtures. And some differences can be seen fresh out of the tube, where the Yellow Oxide starts out brighter and very opaque, while the Transparent Yellow Iron Oxide has a much deeper, almost Raw Sienna masstone and is one of our most transparent colors. From there the differences grow, all tied to the singular issue of particle size more than anything in their chemistry.
Yellow Oxide (PY42) / Transparent Yellow Iron Oxide (PY42)
In the first accompanying graphs we show both colors in their masstone and mixed 1:10 with Regular Gel Gloss, a ratio similar to one often used for glazing. The Yellow Oxide, shown in the two lighter lines, is true to its strong opacity and changes very little considering it has been mixed with ten times its weight of transparent gel. Beyond being somewhat brighter than before, with a more pronounced warmth, the overall shape of its spectral curve is preserved. On the other hand, the Transparent Yellow Oxide changes dramatically, rising steeply through the bands of Yellow, Orange, and Red wavelengths to take on a very warm tone.
The accompanying graph (Graph 12) traces changes in Chroma when varying amounts of gel are added. As one can see, Yellow Oxide remains relatively flat throughout, increasing only slightly as more and more gel is added. No matter how transparent you make it, Yellow Oxide remains a muted color with moderate saturation. On the other hand, while Transparent Yellow Iron Oxide starts appreciably lower in overall Chroma, it actually increases dramatically in saturation as gel is added, eventually surpassing Yellow Oxide at the 1:1 mark and continuing to rise even further. Paradoxically, perhaps, the color grows in brilliance as it is extended with gel, providing proof -- should one ever be needed -- that it is the better choice for creating luminous glazes.Link to additional supporting visuals
Nickel Azo Yellow (PY 150) / Green Gold (PY 150, PG 36, PY 3)
These two colors, one a single pigment and the other a mix, present cases where there are amazing changes in color when they are used to make a tint or glaze. Green Gold, for example, starts its life as a darker, lower chroma lime green in the masstone, but scrap it against a white surface or extend it with Gel to reveal the undertone, and one quickly is confronted with the much higher chroma of a bright yellow-green that is not easily created by any other means. In the spectral curves where this is charted, one can see the initial low-lying dark line that represents the color at full strength. As increasing amounts of Regular Gel Gloss is added, there is a dramatic upward swing in reflectance throughout the Green and Yellow ranges (500- 600nm) and a simultaneous almost complete absorption of Blues and Violets that persists even at the extreme dilution of 1 part Green Gold to 50 parts Regular Gel. It is this total suppression of the cooler bands, and the peaking around 560nm, which is the region where the eye is most sensitive to luminosity, that helps provide the striking vividness of this color.
Lastly, Nickel Azo Yellow (one of the component colors in Green Gold) does a very similar dance, moving from a brownish, Raw Sienna-like hue at full strength, then taking on increasingly translucent orange and yellowish earth tones as more Gel is added, until finally reaching high-pitched yellow notes reminiscent of a bright, transparent Hansa Yellow Medium. In the accompanying graph, we track this movement in chroma when mixed with Gel as well as with Titanium White. It is interesting to note that Titanium White initially increases the chroma of Nickel Azo Yellow all the way through a 1:1 addition until finally succumbing to the neutralizing affect that Titanium White will eventually have when added to any color.Link to additional supporting visuals (Nickle Azo Yellow)
Link to additional supporting visuals (Green Gold)
© Golden Artist Colors, Inc.