EVOLUTION OF A PAINTING
What follows is an in depth study of the making of an oil painting. In considering the techniques described and the materials enumerated, the reader should keep in mind that this is the story of the development of a particular painting as executed by a particular artist. Another artist may work in quite an unlike manner or make very different choices of materials. It is also possible - actually, it is a fact, that with another painting the same artist might pursue some other approach than that which was followed here.
All this having been said, however, many students will hopefullyfind this narrative illuminating and instructive.
- PART I: Preparation and Foundation
- PART II: The Grisaille
- PART III: Introducing Color
- PART IV: Completion
preparation and foundation
Because of time limitations and other constraints, demonstrations I do in classes are abbreviated and simplified. They are not representative of my usual studio practices, yet many students want to know how it is that I do what I do. While it is impractical to demonstrate my working methods fully in a classroom setting (it would require tens and even hundreds of hours spread over many weeks, and possibly months), I can use this forum to give some insight into my process and provide a better understanding of the thinking and methodology behind the work.
“I’m Still Here” was inspired by a sea shell in my prop collection. Its poetical shape of repeating curves is evocative of the human heart. That resemblance coincided with my desire to do a painting that spoke of the congestive heart failure I suffered in the spring of 2012 and my subsequent heart surgery. Thus, it became the foundation upon which I erected the painting.
While I had a vague image in my head of what I wanted, I needed something more concrete before I put paint to canvas. Therefore, I did a number of small, rough sketches.
A few things had already been settled in my mind before I did the sketches. I wanted a diptych-like layout, but on a single canvas with the two sections divided by a narrow band of color; the band of color is an idea adapted from the "zips" of Barnett Newman's paintings. One section of the diptych would contain the shell, and the other would be filled with abstract, blocky forms hanging like heavy clouds in a sky. My first doodle (Figure-1.01) captures this initial idea.
Figure 1.01: The first concept was unacceptable due to the clash of the curvlinear shell versus the geometric blocks.
I wasn’t particularly happy with the sketch…on several counts.
For one thing, the blocky shapes seemed out of place. Their severely angularity contradicted the organic and curvilinear character of the shell. Part of my intent was to convey the constant tiredness and difficulty in breathing associated with congestive heart failure that comes from fluids collecting around the heart and lungs, squeezing the energy out of them. The blocks were intended to suggest the weight of the fluids, but instead they ended up looking forced and purposeless, as though they were an easy answer to filling up the space of the canvas.
I then decided that thick clouds or a mist would be more appropriate to the narrative, so I tried integrating the block-like forms into violent storm clouds in the second sketch. This is portrayed in Figure-1.02 with its bolts of lightning at the lower right intended to suggest the sudden pain of a heart attack.
Figure 1.02: This second sketch is more cohesive than my first effort.
The picture now began to absorb the block forms, a definite positive, although they were still more noticeable than I thought they should be. The clouds had a pleasing organic quality, so clouds and shell together now formed a more harmonious relationship than blocks and shell had. Inadvertently, the quick arc of the clouds also suggested a compositional scheme: repeating ovals.
This was all an improvement, but still not quite right. For instance, the lightning wasn’t really true to my personal experience. My heart attacks had been silent; there was no pain at all. In fact, I hadn’t even been aware I had two heart attacks until my doctor ran tests that confirmed they had occurred. The lightning was also a bit “over the top,” much too melodramatic for my tastes.
A third sketch (Figure 1.03) appeared to be the answer. In it, a massive cloud strongly suggested the effects of congestive heart failure, and the blocks were now well embedded in the cloud to the point of being nearly unseen.
(At this point you may be wondering about my insistence on including slab-like forms in the picture. There are several reasons, which are explained in the sidebar to the right.)
Figure 1.03: This is basically a refinement of the second sketch.
I now began to refine the idea with a rough tonal study. In the process I began to work out the value structure of the composition, which emphasizes arcs and ovals.
Figure 1.04: To help better visualize what the cloud section of the painting might look like, a tonal study was made.
In the sketches done thus far, I had inadvertently made the shell panel about twice as tall as wide. Out of curiosity I measured the shell itself and discovered that at its widest point it was about half as wide as it was long. This seemed to be too convenient to ignore. Using those proportions, a ratio of one to two, the left (shell) panel ended up with its height being twice its width, and I made the right (sky) section twice as wide as its height. Further, the sky panel is divided into four sections that repeat the dimensions of the shell section. (See Figures 1.05 and 1.06.) Each of those divisions would be treated as a critical point in the final painting.
Figure 1.05: In this sketch the proportions of the painting are established.
All these ideas were combined in a final composition sketch.
Figure 1.06: Compositional study confirms the idea of repeating arcs and ovals, as well as the breaking up of the canvas into a series of 1:2 rectangles.
It is in this sketch (Figure 1.06) that a crack in the shell first appeared. Notice that the hole is in the shape of an inverted shell.
getting ready to paint
Now that I had an acceptable sketch, a fully resolved study was needed as a roadmap during execution of the painting. But first, in order to carry out the study, a little research was in order. I needed to know what clouds of the sort I visualized truly looked like so I made use of that marvelous technology, the Internet, as well as my own file of clippings from magazines and books. Here are the images I selected (Figure 1.07.).
Figure 1.07: Collection of photographs as candidates for the clouds in my painting. Those bordered in red ended up being the two selected as models.
Ultimately I ended up just the two outlined images as references, along with my own vision.
With the shell on a small table in front of me, and my cloud photographs near at hand, the composition was fleshed out taking into account value structure as well as the shapes, sizes, and positions of the major elements. This is shown in Figure 1.08.
Figure 1.08: A fully realized pencil rendering of how the finished painting is expected to appear.
At this point, small color studies would normally be done to determine the color scheme (see sidebar at right). I had already visualized the painting in color, though, and felt that color studies weren’t needed. Being able to skip that step, I went on to prepare the canvas (see sidebar).
Figure 1.09: Based upon my pencil rendering, the composition is laid out full size with pencil on prepared canvas.
Once the canvas was ready, a tape measure was used to divide it up into 10" wide segments with the leftmost section being reserved for the shell. A 2" wide strip separated the shell area from the cloud portion.
With my pencil study propped up on a nearby easel as reference, I then used a 2B pencil to draw the large elements on the canvas and sprayed the drawing with two layers of workable fixative to protect it (Figure 1.09). Because it is easy to wipe off the canvas, vine charcoal is normally used for this step, but pencil was selected instead for greater precision.
The canvas was now ready to receive paint.
a very old technique
Some painters work directly on the white gessoed canvas, others tint the canvas first with paint. They may let the tint dry before developing the painting, or they may work directly into the wet color. The technique I most often use (as shown in Figure 1.10) was adopted from the great Baroque master, Peter Paul Rubens.
Before embarking on a painting, Rubens prepared his canvases by toning them in an uneven, streaky pattern. This pattern has a direct effect on the character of the finished painting.
Even though the paint may appear opaque in a completed picture, light still penetrates it. Light makes its way through the paint to the white canvas and then is reflected back through the layers of paint and into our eyes. Where the variegated layer is thicker and darker, slightly less light is able to reflect back to us than in those areas where the layer is thinner. The outcome is that wonderful light flickering across Rubens’ paintings that we so admire.
Figure 1.10: Borrowing an idea from Peter Paul Rubens, streaks of color were applied as an underlayer.
Taking that lesson from Rubens, I used a large bristle round and a large bristle flat brush to sweep color onto my canvas. The color chosen was brown ochre light (see sidebar) heavily diluted in odorless paint thinner. Normally I wouldn’t recommend thinning the paint so much, but in the first layer and for the purpose that I used it, such extreme dilution is perfectly safe from a technical standpoint.
At this stage, the canvas was placed against a wall and the paint permitted to dry for a week.
In the last installment we left off with toning the canvas in an irregular pattern. It is important that the paint be well cured so that it remains undisturbed when worked on top of; the toning was therefore allowed to dry for about a week before proceeding.
Figure 2.01: Here we see the canvas toned in an irregular pattern. The color has dried for a week and is now ready to receive the next layer of paint.
When the surface was ready, a thin veneer of perylene black was spread more-or-less evenly across the entire canvas except for the dividing strip, which was left white. The paint was diluted very slightly with the tiniest bit of odorless paint thinner, and then daubed onto the canvas with a large brush. The daubs of paint were subsequently spread out into a thin layer by working them firmly and vigorously with same large brush.
While the paint was still wet, the light-dark structure of the painting was broadly established by “wiping out” the lights. Wiping out is a centuries old practice and is a simple idea: remove or reduce the layer of black paint in some areas with rags and brushes soaked in paint thinner (or more traditionally, in turpentine). Areas where a little black paint is lifted become paler, and where a lot of black paint is erased the surfaced is restored nearly to its original whiteness. The result can be seen on the rightmost portion of the painting below.
Once the black paint has dried, execution of the grisaille may begin.
Figure 2.02: Black paint has been thinly laid onto the canvas, and then some areas have been “wiped out” with rags and brushes soaked in odorless paint thinner to establish the general light-dark structure of the painting.
general information about the grisaille
The term “grisaille” comes from the French word “gris” for gray. A grisaille is a tonal painting, often done with grays, but it may also be carried out with shades of brown or blue or almost any color.
Ideally, the grisaille is done with colors that are low in oil content. The reason for this has to do with the oft heard admonition, “paint fat over lean.” This means to avoid applying pigments of low oil content on top of colors with high oil content. The reason is quite simple: to produce a structurally sound picture. Another good practice to follow is to begin a picture with faster drying colors, later on using slower drying colors if desired. (For more information about this, visit the Archives page on this site and select “Oil Painting” from the sub-menu to see “Lo-Fat Diet.”) I chose perylene black with cremnitz white for my grisaille. My reasoning is explained in a sidebar.
A purpose of the grisaille is to help simplify the process of making a painting by dividing the process into two manageable steps.
The first step is the grisaille itself, which allows the artist to concentrate on drawing issues and value structure without worrying about color. When done, a grisaille is like a finished painting, complete in all respects but lacking color. The second stage is to apply color over the grisaille. Fully transparent pigments may be used so that nothing more needs to be done, or a combination of transparent, translucent, and even opaque colors may be employed.
Because it is difficult to judge slight differences of value, I prepared a series of grays in advance on my palette. With a palette knife, I mixed up a big batch of the darkest tone needed, and then used the knife to separate out a portion of the batch and added a small amount of white paint to the separated mound. The process was continued – separating and then adding white, separating and then adding white, and so on – until I eventually made the palest tone needed along with many steps between. When I was done, I had about 20 different color samples, which actually turned out not to be enough. In the process of executing the grisaille, I found that I had to combine pairs of pre-mixed gray-tones to create one or more intermediary tones in order to better meet my needs.
Figure 2.03: Shown is the series of pre-mixed grays used for the grisaille; additional shades were made while the painting was in progress by combining some of the mixtures.
The tools and materials required for the grisaille were basic: palette knife, a few brushes in several sizes, plus a very little painting medium (recipe given in sidebar).
Paint was put on the canvas with mid-sized stiff or semi-stiff brushes (#6 and #8 filberts and brights). For blending and other manipulations, smaller soft haired brushes were chosen (#1, #2, and #4 filberts and flats). Many were a bit unruly and beat up since older brushes are so good for blending. Newer brushes in better shape were selected mainly when precision was an issue.
Figure 2.04: Among the brushes for executing the grisaille are several that are a bit beat up; I find these very useful in blending. At right is the palette knife I typically use to mix colors.
In keeping with the fat-over-lean guideline, I used painting medium sparingly; only enough to moisten the brush on rare occasion. For the most part, though, I used the paint exactly as it came from the tube with nothing added to it.
the shell section
To keep things manageable, I began with the zone containing the shell. I took this route partly because the amount of time available was limited; it was already past noon when I was set to go. The shell section was the smaller of the two parts of the painting, plus if things took longer than anticipated I had a convenient stopping point: I could paint the background to the edges of the shell and then save the shell for another day. The cloud unit had no good stopping point; once begun I would have to continue until the section was finished or risk leaving behind unsightly seams.
As it happened, things worked out and I was able to complete both the shell and its background in the same day.
The procedure for the background was really rather simple: lay in patches of color adjacent to one another and then blend their edges together to create transitions. (Refer to “Methodology of Blending” in the sidebar.) You can see in the following illustration that the dark upper-left portion becomes gradually paler toward the lower right. The mottled appearance gives testimony to the fact that a many patches of color were laid down in this section and then blended into one another.
Figure 2.05: The background in this upper quadrant was carefully painted to create a gradual transition from dark in the upper left to paler on the lower right.
In most instances I work from dark to pale, as was true with the above example. It is also true of how the cast shadows were created. The entire section in which the shadows appear was rendered darkly. Then paler grays – and sometimes white paint –were picked up on the brush and worked into the wet paint to heighten the areas surrounding the shadows and to define their shapes.
Figure 2.06: Cast shadows are formed by first painting the general area with a dark-ish color and then painting around the shadows with paler paint.
This produces a more convincing shadow than does painting a shadow on top of the background.
The shell was painted in a way similar to the background, but with smaller and softer brushes.
More care needed to be taken here than with the background since detailed shapes and surface textures had to be taken into account. Many areas were worked and then reworked, painted and adjusted. Reworking and adjustments constituted most of the labor, and was performed by adding tiny dabs of paler or darker color to the base coat of paint where needed. In some cases a dab or stroke of paint was allowed to stand, in other instances it was blended partially or completely into the surrounding color. This procedure often had to be repeated several times in order to achieve the desired results.
Figure 2.07: Here is the grisaille of the shell section in its final state. The shell itself is a product of firm strokes of color combined with delicate blending.
You would think that something as detailed and specific as the shell would be more difficult and require more work than a cloud with its vague masses. However, it was just the other way around. Painting the skyscape turned out to be extremely demanding and required a long work session.
As with most procedures in picture making, one begins with the large generalities and works toward the smaller and more specific details. That is exactly the approach taken with the cloud section.
Work commenced by laying in large patches of paint for the darker regions, and then middle values and paler patches were added where needed. After the major areas of dark and light were well established, a great deal of refining and adjusting took place to more exactly describe the mass of the cloud, the sky that contains it, sea beneath it, and the boxy forms buried within it. The results of this process are illustrated in detail by Figure 2.08 through Figure 2.
Figure 2.08: The grisaille of the entire cloud section is seen here. It is clear and well organized into large zones of light and shadow that provide structure. A range of middle values make the picture more believable and serve as transitions between the light and dark areas.
Figure 2.09: This close-up demonstrates how large areas are broken up into smaller patches of light, dark, and middle tones. The following images are increasingly closer views of a small portion of the cloud section. They provide an intimate understanding of how the paint was thoughtfully applied and manipulated with small brushes.
Figure 2.10 Figure 2.11
Figure 2.12: An extreme close-up of a small area of the painting.
The grisaille was now complete. There remained only two steps to go before beginning the glaze and velatura: trim the edges of the canvas, and lay a foundation for the pale dividing strip.
Perched on an easel, the grisaille awaits a few final touches. In the background on the right of the photograph, the shell used as a model for the painting is mounted on a tilted sheet of foam core board where it is easily visible from my painting station (Figure 2.13).
Figure 2.13: The complete grisaille on a studio easel with the shell nearby.
Paint remaining on the palette from executing the grisaille was scrapped into a big pile, mixed together, and then applied to the edges of the canvas to make an attractive presentation for exhibition.
Figure 2.14: The edges of the canvas have been trimmed and it now sits drying. On the right in the photo the preparatory drawing for the project is mounted on a secondary easel.
After several days, the pale dividing band was given its first coat of paint (Figure 2.15). In keeping with the theme of the band representing a brilliant light shining out from between the shell and cloud panels, cold, bright and vibrant colors were used in contrast to the warmer and duller colors of the end sections. They were: titanium-zinc white, ultramarine violet, cobalt blue, viridian, and cadmium yellow lemon. For more about these colors, see the sidebar.
Figure 2.15: The vertical dividing band is given its first coat of color.
A series of deep transparent colors was mixed up in advance for glazing the background surrounding the shell in "I'm Still Here." (See "Colors for the Shell Background" sidebar for list of colors.)
glazing the shell section
With a stiff #4 synthetic filbert brush, a transparent hue was daubed in near the top of the section (Figure 3.01a). A small quantity of glaze medium (see "Glaze Medium Recipe" sidebar) had been mixed into the paint and using the same brush, the paint was spread around into a thin veneer over the upper portion of the image (Figure 3.01b).
Figure 3.01a (left): Spots of dark, transparent color are dabbed around the area to be glazed.
Figure 3.01b (right): By manipulating the paint firmly with a stiff brush, it is spread out into a homogenous film.
This technique continued as I worked my way down toward the bottom of the image. A color was dabbed over a small area and spread into a thin layer, blending one region of color into the next where two colors abut in order to create gradual transitions.
Since the strongest light was to be focused around the stem of the shell (near the bottom of the picture), the color was generally brightened as I moved downwards. The very bottom of the image was allowed to fall into shadow, however, as can be seen in Figures 3.01c and 3.01d.
Glazing of the shell’s background is nearly complete in Figure 3.01c (left), with only a small portion of the original grisaille still visible at the very bottom of the picture. Glazing of the section is entirely finished in Figure 3.01d on the right. Also shown is the brush used for the majority of the glazing operation.
scumbles and velaturas
The next step, which involved the use of opaque colors, proceeded immediately while the glaze layer was still wet. This allowed the opaque paints to be easily blended, or feathered, seamlessly into the glazed colors.
Two traditional painting techniques were employed: the scumble and the velatura. (The colors used, and a description of the techniques are explained in the sidebars.)
Near the top of the section, cool and pale color was brought in from the right as though a fog was wafting into the painting. Farther down, approaching the area around the stem, brighter and warmer color was used to imply that a light was focused in this area. These brighter patches are critical to the dynamics of the shell section of the painting. Alternating as they do with darker bands of color, they create a pulsating rhythm that animates the shell as though it is a beating heart. They also serve to break up what would otherwise be a dull expanse of essentially a single color.
After bringing the top half of the background to the state desired, a splash of cool, pale opaque color was worked in near the upper right (Figure 3.02a, left). As work continued downward (Figure 3.02b, center), brighter and warmer color was introduced to suggest that the light was focused in the area around the stem of the shell. Finally, in Figure 3.02c on the right, darker tones were reinforced with mostly transparent colors at the bottom.
It is important to understand that the thinness of the glaze and velatura paint layers was achieved through vigorous manipulation of the paint, not by adding an excessive amount of painting medium to the paint. In areas that needed to be especially pale, the opaque color was built up more heavily, such as the bluish-pink patch near the upper right.
Also note that the shadow cast by the shell was established by the dark tones of the grisaille and not by painting it with dark colors. The shadow was tinted by working on top of those dark tones with a thin layer of background color. Thus, the value structure (i.e., pattern of lights and darks) remained dependent on the grisaille while color was added by the glaze and velatura.
The main concern in building color on the shell was to give it its own coloristic appearance distinct from the ruddy background, and at the same time retain a sense of the red light bathing the entire scene.
As with the background, work began with the placing of daubs of color (left, below) that were then worked out into a thinnish layer of variegated color that modeled the surface of the shell. More care was required here than with the background for there was a lot of detail and surface texture to consider, as well as the effects of a pronounced chiaroscuro. To keep the form solid looking, large patches of color were first filled in, such as a dark, blue-green mass to represent an area of shadow, and pinkish or peachy tones for illuminated zones, and then the smaller details were then worked into these large regions. This is apparent in Figure 3.03b (center), where the light and shadow masses are quite clearly defined. The illustration on the right, Figure 3.03c, shows both background and shell completed.
Left to right: Figures 3.03a through 3.03c
Unlike the background colors, which were prepared in advance, colors for the shell were compiled during the process of painting it. Figure 3.04 below displays the remnants of the color mixtures remaining on my palette at the end of the session. Also laid out on the palette are the tools used in painting the shell, including several beat up and unruly brushes I find useful in putting down soft-edged marks and for blending.
Figure 3.04: Color mixtures used in painting the shell.
the cloud: initial tints
Initially, nothing dramatic was undertaken to elaborate upon the large cloud in “I’m Still Here.” Rather, the first steps were to establish the general warm-to-cool tone of the panel going from left to right.
Figure 3.05a: A tepid start to the cloud section of the painting yields uninspired color.
In addition to Perylene black and unbleached titanium white, as seen below in Figure 3.05b, a very limited palette of cobalt blue, mars violet, and olive green was used along with touches of yellow ochre half-burnt. The left portion of the painting was warmed up with reddish mars violet, with the color becoming increasingly cooler and bluer toward the right in Figure 3.05a above. Green and ochre provided warm accents of stronger light where needed, especially the glow peeking out from beneath the left underside of the cloud.
Colors for the initial work on the cloud section.
Upon reviewing the painting at this stage, the cloud appeared to be timid in treatment. It lacks the power it should have possessed, so I set about intensifying the darks to increase contrast and drama. The result appears in the illustration below at left, Figure 3.06.
More assertive darks in Figure 3.06a (left) now provide a stronger foundation for the improvements that evolved (Figure 3.06b, right).
With the darks intensified, it became possible to use more saturate color to give the lights greater richness and luminosity. The blocky forms within the cloud were now aggressively re-established, and as is evident in Figure 3.06b, and the masses of the cloud itself acquired more presence and weight.
"I'm Still Here" is essentially a diptych, a picture composed of two panels. One of the primary concerns with that format is that the two parts of the picture form a unified relationship and constitue a well structured composition as a whole.
Figure 3.07a demonstrates the basic compositional theme of the painting.
The intended pictorial theme of ovals and curvlinear shapes is clearly evident in Figure 3.07a, above, which also diagrams the continuity of the painting overall. Upon careful review, you are certain to discover many repititions of elipses, many with proportions identical to those in the illustration. Below, the painting is reproduced without the overlay.
Figure 3.07b: Development of the entire painting prior to repainting of the dividing band.
One step remained before I could stop painting and clean up: prepare the dividing strip to receive a velatura.
In contrast to the dull and warm-ish colors of the two panels of the painting, it was planned that the dividing strip was to be cold and bright to suggest brilliant light breaking through the murky darkness. To that end, strong, sparkling color was laid in for the strip. Ultimately, it will be covered over with a translucent layer of less colorful and more neutral shades of paint…but the vivid colors that would be laid in now would later have a direct effect upon the paint that will be layered over it in the end.
Color at the top was cold, weighted toward violet and blue, and became warmer as the strip descended. The palest and warmest area fell about two-thirds of the way down, and then began to shift back towards darker and cooler color, this time running more toward green.
The procedure and colors applied to the upper portion of the strip appear in Figures 3.08a through 3.08e. To make them paler, titanium white was added to all the colors used, especially in the bright region lower down. From top to bottom the colors were as follows: manganese violet and ultramarine violet (Figure 3.08a and 3.08b), and then icy ultramarine blue followed by the warmer cobalt blue (Figure 3.08c-d). The section ends with cerulean blue (Figure 3.08e), which leans ever so slightly toward green.
Figures 3.08a through 3.08e (left to right) chart progress from the reddish tone of manganese violet at the top of the dividing strip, through the cold of ultramarine blue to a pale mixture of titanium white with warm cerulean blue near the bottom.
Notice that each color was applied in a relatively small area, leaving a slight space between it and adjacent colors. Neighboring colors were then blended together with a soft-haired brush to produce smooth transitions. To achieve this, only the tiniest amount of painting medium was added to the paint, just enough to make the paint buttery, and the paint was put down in a meaty layer with some body, but not in overly thick deposits.
The next group of illustrations chart progress in the lower third of the strip. Cobalt turquoise combined with an abundant amount of white paint was the first color introduced (Figure 3.08f) in this section. Then, at this point, progress was reversed, working from the bottom of the strip upward. Figure 3.08g, for instance, shows cobalt blue at the very bottom, with some bluish viridian green immediately above it (Figure 3.08h). To complete the strip, strong passages of light were brought in (Figure 3.08i); a cool flare near the top and a warmer glow made with touches of cadmium yellow lemon toward the bottom.
Figures 3.08f through 3.08j continue the process of modeling the dividing strip with bright, cool notes of color.
The complete range of colors used for the dividing band were mixed ahead of time and are spread out on the palette shown in Figure 3.09 below, along with the brushes used to paint the strip.
Figure 3.09: Palette of colors for the dividing strip.
This is what the painting looked like at the end of the day. It was now a well unified image with a firm design structure and sound color/value relationships. The picture was well advanced at this point, enough so that it was easy to see the direction in which it must logically continue.
After having set the painting aside for a few days so I could return to it with a fresh eye, I saw immediately several issues that needed to be addressed. The first thing I attacked was the shell’s background, which was much too fiery next to the cold, greenish tones of the rectangle of the cloud section.
Figure 4.01: Returning to work on the painting after a break of a few days, several deficiencies were noted.
I began by neutralizing the fierce orange-red background with glazes, and then working back into the glaze with opaque colors to pull out areas of illumination and atmosphere. In addition to neutralizing the angry background, the glaze color helped unify the area by establishing an overall tone. I carried out this exercise piece-by-piece, working generally from to top to bottom. The entire operation is illustrated and explained below.
Figure 4.02 (left), and Figure 4.03 (right).
A mixture of ultramarine violet and Perylene black was glazed onto the uppermost portion of the painting (see Figure 4.02). With soft paper towels (Viva brand), the glaze was gently wiped to thin it down. Next a blend of opaque color (unbleached titanium white pale, mars violet, and German black) was daubed onto the glaze (visible at the upper left in the photograph). Using soft haired brushes, the patch of opaque color was modeled and feathered into the surrounding glaze (Figure 4.03). This process continued by extending the glaze downward (apparent in Figure 4.03) and then working back into the glaze with opaque color.
As work progressed, I began adding Antica green earth to the glaze color. By the time the lower third of the painting had been reached the glaze was entirely green earth.
Figure 4.04 (left) shows the finished shell section, and Figure 4.05 (right) is what it looked like before revision.
Just as the glaze color evolved, so too did the opaque color. It became warmer and more insistent to convey the effect bottom half of the image being washed in light. Pigments like Ercolano red and mars orange were used.
After spreading opaque paint into an area, it was often necessary to brighten part of that area. To do so, a second batch of opaque color would be made (using the first mixture as a “starter,” much the way sourdough bread is made). This second batch would be slightly paler and brighter, and sometimes a bit warmer in temperature. A little of this new batch would then be daubed into a spot in the opaquely painted area, and then be manipulated to blend it in. In this manner it was possible to produce a range of values representing stronger and weaker areas of light. This can be seen in Figure 4.04, where the color grows particular intense to the right of the shell’s stem.
After having redone the reddish background, I turned my attention to the shell, adjusting it so that it would be better integrated with the background. I also made it a bit more solid looking by deepening the darks and warming the lights.
Figure 4.04 and Figure 4.05 offer a comparison between the final result and what the shell looked like at the start. Below can be seen the completed painting.
Figure 4.06: The completed painting.
OF A PAINTING
Blocks and Slabs
“I’m Still Here is intended for an upcoming exhibit entitled “Particles of Time,” and the blocky forms represent tiny chunks of time.
The first appearance of my trademark slabs was in "A Curious Place," a painting completed in 2001 (see Relativity Gallery page on this site). Up until that time I had recently been making highly realistic depictions based on real-world observations.
I found this approach was limiting my development since I was working so hard to make my pictures look exactly like what I saw. Simultaneously, I had been reading numerous books on physics and mathematics and I wanted to make images in response to what I had been reading. My current approach would have been totally ineffectual.
I therefore decided to paint what I could not see, using blocky forms as metaphors for a variety of objects and ideas, such as particles of light, isomorphism, and so on. This tactic also meant I could more effectively address such abstract issues as composition and color theory without concern for having to mimic reality.
I liked the freedom this new method gave me as well as the architectural character of the forms (as a teen I had briefly considered becoming an architect), so brick-like shapes became a permanent fixture in my paintings.
Utrecht cotton duck canvas (#10) was mounted on heavy-duty stretcher bars measuring 20" X 52", which is proportional to the preliminary study.
Although I much prefer linen over cotton, linen is many times more costly. Cotton was chosen purely of economic necessity.
Cotton expands and contracts considerably in response to variations in humidity. I would be using acrylic gesso to prime the canvas, so I knew that the moisture in the gesso would cause the canvas to loosen up once the gesso had dried. To help counter this effect, I pre-stretch the canvas and then wet it down with a broad brush and distilled water. After it has dried (overnight), I may repeat the process (which I did with “I’m Still Here”), or go on to do a final stretching of the canvas. The final stage is done with the help of canvas pliers in order to get the canvas as taut as possible.
Two coats of Golden brand gesso were applied to the face and edges of the canvas with a 3" wide house-painter’s brush.
The first coat was applied with side-to-side (horizontal) strokes, allowed to dry overnight, and then lightly sanded before the second coat was put on in an up-down motion. When done, the canvas was set aside for a week to allow the gesso to cure properly.
A note about why the final coat of gesso is applied in an up-down direction:
Light in galleries and most homes comes from above (a ceiling fixture). If the last coat is applied horizontally, light raking across the slight ridges left by the brush causes glare. When gesso is applied vertically, this doesn't happen and glare is minimized.
I am fortunate to have a studio with plenty of natural light, and the windows face northwest, which is nearly perfect. I say nearly, because in late afternoon the sun shines almost directly into my eyes, so I lower the white cloth drape and sometimes place large sheets of foam board in the windows to moderate the harsh light.
For “I’m Still Here,” I had placed my reference sea shell on a tray-table near my easel (visible near the center of the following photograph) and directed a strong light at it. So that natural light from the windows would not interfere with my set-up, I propped several stretched canvases against the windows to block most of the sunlight and create a controlled environment.
Since I am right handed, the easel is positioned so that window light comes from my left. My main oil painting station is arranged so that everything I need is conveniently at hand, with easel before me, palette to my right, and brush cart on the left. The easel is capable of handling fairly large paintings and can be tilted back or forward. A sheet of thick, tempered class serves as a palette; the back of the glass is lined with gray paper so that colors can be mixed against a neutral background. My brush cart has a flap at each end that can double the cart’s length; one flap is raised in the photograph. Clamped to the flap is a plastic bag filled with slightly used paper towels; to avoid wastefulness, they will be put to work a second time cleaning palette and brushes. A row of brushes sits on the surface of the flap. On the right end of the cart are jars of painting medium, turpentine, and paint thinner. Between the jars and brushes are arranged tubes of paints currently in use.
On the far right in the picture, but behind my easel in “real life,” is a heavy metal cart. On the top shelf are tubes of paint, one tube of each color I own, which is nearly 100. Many jars crammed with brushes fill up the front of second shelf, and in back are gallon cans of paint thinner, turpentine, and other odds ‘n’ ends. On the bottom are tubs of acrylic gesso and painting medium, along with bottles and jars of varnishes, oils, balsams, and similar materials, along with canvas stretching and picture hanging tools and supplies.
In the bottom photograph, a second easel has been placed very close to where I am working. It holds my drawn and photographic reference materials well within eyesight as I work. Above, in the ceiling, is one of several fluorescent light fixtures that illuminate my studio at night. The fixtures have been outfitted with color corrected bulbs.
for the Grisaille
The medium used for the grisaille was a simple one: equal parts odorless paint thinner and refined linseed oil. This is a light bodied, slow drying medium and one that I regularly recommend to my beginning painting students.
Colors for the Grisaille
Building a structurally sound painting requires that the first layers not dry too quickly or contain much oil. The rules, stated in simple terms, are as follows:
Fat over lean
(oilier over less oily)
Slower drying over more rapid drying
Thus, the paints used in the first layers of a painting are ideally low in oil and dry swiftly, with cremnitz white and perylene black fulfilling both these requirements.
In addition, perylene black has a distinctly greenish cast to it, an excellent foil to the reddish tones intended for the finished painting. While all white pigments are cool in temperature, lead white (cremnitz is a type of lead white) is the warmest of the whites and thus fit in well with my plans.
Glaze Medium Recipe
For glazing, a painting medium that dries relatively quickly and increases the transparency of the paint is desirable. Therefore, the slow-drying refined linseed oil used in painting the grisaille was replaced with faster drying stand linseed oil. Because stand oil is so thick (it is about the consistency of molasses), the ratio of paint thinner to oil was about 5 to 1 so as to achieve a more fluid consistency. To this mixture was added 1-part dammar varnish, which makes the paint more translucent and has the additional benefit of accelerating the drying of the paint.
Three Traditional Techniques
Glazing is a remarkably simple idea that for some reason many students seem to find very mysterious. It is accomplished by spreading transparent color over a dry, underlying painting. When the underlying picture is a purely tonal study (usually shades of grays), it is generally called a grisaille, which comes from the French word “gris” for gray. The following provide excellent illustrations of the technique.
First is a grisaille (shown above) in which all the compositional, drawing, and tonal issues are resolved. It is, in effect, a finished painting lacking only color.
The artist then gave color to the jar, wall, and table with transparent paints, leaving the colander mostly as shades of grays. In some areas, the glaze was applied as just a whisper of color to give only a slight tint. In others, multiple layers of color were built one atop the other to provide stronger notes, allowing each layer to dry before the next was applied.
Glazes and velaturas are similar to one another. Both rely on the idea that a dry underpainting remains at least somewhat visible beneath overlayers of fresh color, except that a glaze is transparent and a velatura opaque. An effect of this difference is that a glazed picture typically possess beautiful clarity and rich color, and the glaze almost always darkens whatever it is applied to. On the other hand, a velatura is normally carried out with pale opaque colors. So, even though it is applied as an extremely thin film, the velatura lightens what it is laid over while obscuring it to some degree to produce a somewhat hazy or smoky look.
In “I’m Still Here,” velaturas were used primarily to heighten areas in light. Outstanding examples of exquisite use of velaturas are some of the paintings made by Thomas Chimes, such as the one above.
A scumble is often applied on top of an area that had been glazed, usually after the glaze has dried but sometimes while it is still active. Scumbling is carried out with opaque colors put down in a patchy manner so that that the scumble does not completely hide what is beneath it. In other words, some of the underlying color peeks out here and there. Much impressionist art is constructed by scumbling, especially the work of Monet and Seurat.
The image above is a painting by Wolf Kahn. It is made lively due largely to scumbling. Instead of laying in color of homogeneous passages or by blending colors gently into one another, Kahn worked layer of color atop layer of dry color to create an active and energetic surface.
OF A PAINTING
This is the sea shell that spurred the creation of “I’m Still Here.”
As you can see, it is very suggestive of a human heart.
Trying to work out problems while a major painting is in progress inevitably yields unsatisfactory results. For this reason, studies are advisable in most cases. Depending upon the complexity and character of the painting, preliminary studies may be as crude as a few doodles on a napkin or they could be quite elaborate. For highly complex and challenging paintings, I do a series of rough idea sketches followed by a compositional rendering. There may also be pencil studies to improve understanding of important elements (such as a hand or an onion plant).
Once a satisfactory compositional study is completed, I might then execute one or several oil sketches to resolve color issues. If color studies are done, there would certainly be a sketch of the painting in its entirety. In addition, there may be color studies of major elements (for instance, individual figures or the effect of candle light).
Because I already had in my head a good image of the finished picture, including color structure, I didn’t feel the need for color studies in this instance. Although, admittedly, if I run into questions during the execution of the painting there is always the possibility of setting it aside while a color sketch or two is done to find answers to those questions.
Brown Ochre Light?
To create a structurally sound painting, the early layers of paint should contain less oil and dry more rapidly than later layers (go to the menu above and choose OIl Painting, and then "Lo-Fat Diet" for more detailed information and technical data). Ochres fulfill these requirements so they make a good initial layer.
Brown ochre light was specifically chosen for several other reasons, though.
I had already decided that blacks and reds would predominate in the finished painting. For blacks I would mostly be using perylene black, which is decidedly greenish and therefore an excellent foil to the reds. Thus, the grisaille would be executed with perylene black.
Brown ochre is reddish and would contrast well with the greenish black that would lie directly on top of it. It had the added benefits of being inexpensive (anybody who has read this page knows that Fred is a true cheapskate) and a very rapid drier.
Beginning students often think that by thinning the paint to an almost watery consistency it is easier to blend. Just the opposite is true. Little or no painting medium (or paint thinner) should be added to the paint, and the paint layer should have some body without being thick or globular.
Transitions are achieved by laying patches of color next to one another and then blending them together where they meet. Fast transitions are the result of abutting two strongly contrasting colors. Slower transitions are produced by placing arranging a series of colors in a row and then working each into its neighbors.
A method I frequently use is to gently work the edge of a filbert brush or a beat-up soft-haired flat brush in small circles along the seam where two colors meet. This causes the colors to mingle with one another. For a short, quick transition, I confine my activity to just a narrow area along the seam. When a slower transition is desired, I continue the action as I gradually expand away from the seam. In the latter approach, the action is in one direction only, such as from left to right, and it is never reversed.
Paint Brands and Colors
For interested readers, here is a complete list of all the colors by brand of the paints used in executing “I’m Still Here.” You may notice that several brands of terre verte appear on the list; that is due to the fact that each has a slightly different character from the others. Comparisons and additional information about the different paint brands can be found on the Material Advice pages of this web site; oil painters should consult the Material Advice page “For Oil Painters.”
Blockx: lapis lazuli (genuine ultramarine blue)
Holbein: cobalt violet
Maimeri Puro: terre verte
Michael Harding: alizarin crimson, manganese violet, raw sienna, ultramarine blue (French), ultramarine violet
Old Holland: manganese blue, yellow ochre half-burnt
Old Holland (left) and Williamsburg are two of the finest brands of oil paints made in this writer’s opinion.
Rembrandt: rose madder light (a natural pigment that is warmer than and has been largely replaced by alizarin crimson), terre verte
Rublev Natural Pigments: antica green earth (natural terra verte pigment found in northern Italy near Verona), Ercolano red (named for the town at the foot of Mt. Vesuvius in Italy near which deposits of the pigment are obtained)
There are no synthetic pigments in the Rublev line (left). The company seeks out only the finest naturally occurring pigments from all around the world, and offers many rare and historical colors such as genuine vermilion and true lapis lazuli. On the right is a sample of oil paint from Michael Harding, a British color maker. Their paints are excellent, and I have found their transparent hues to be particularly lush.
Williamsburg: cobalt turquoise, Italian burnt sienna, mars violet, titanium white, titanium/zinc white, unbleached titanium white, unbleached pale titanium white, viridian
Winsor & Newton Artist’s: bright red (pyrrol red), cadmium yellow lemon, cerulean blue, cobalt blue, cremnitz white, Indian red, mars orange, olive green, Perylene black, transparent yellow, yellow ochre pale
Colors for the
Work on the background began by glazing with the series of mixtures shown in the first photograph below. They were made with the following transparent colors: ultramarine blue, alizarin crimson, bright red (pyrol red), Italian burnt sienna, and raw sienna.
As work on the shell background proceeded, opaque colors were eventually introduced to heighten areas in light and give more body to the paint. (The mixtures are along the left in the picture below). The opaque colors used were unbleached pale titanium white, Ercolano red (a type of English Red Light pigment obtained from quarries near the town of Ercolano in Italy), mars orange, Indian red, and yellow ochre light.