Oil Painting
My Current Methods and Recommendations
Oil painting is a very complex craft that takes many years to master. With every passing year, I discover new aspects of the medium that I didn't know or fully understand in the past, and new solutions to persistent problems. I've only very recently arrived at a sense of a really confident command of the medium - although I suspect that I'll be saying the same thing yet again five or ten years from now.
There are many misconceptions about oil paint. For instance, it's widely believed that oil colors remain wet on the canvas for many days or even weeks after application, and can be reworked during that time; in reality, I find that most applications are dry to the touch within 24 hours (sometimes much less). But I typically work in very thin, transparent layers, and I normally can't expect those to be workable after much more than a few hours.
Another misconception is that oils do not shift in value (darken) as they dry - this is generally true, but some pigments are exceptions to this rule, and they can be a real nuisance. The most troublesome of these is Titanium White - I use it only sparingly. More about this later as I discuss my palette.
Surfaces
I rarely work on linen canvas anymore, and never on stretched canvas. I haven't stretched a canvas in years. Instead, I've made the switch to aluminum composite panel. Since these panels are very resistant to changes in atmospheric heat and humidity, they offer the most stable substrate that an artist can ask for. That's why art conservators recommend this product strongly.
These rigid, lightweight panels consist of two thin sheets of aluminum, with a polyethylene core sandwiched between them. They are most often only 3 millimeters thick, but can be found as thick as 6 millimeters. For smaller paintings, the 3 mm solid-core variety is perfectly suitable. For larger work, I use a thicker panel with a corrugated or honeycombed plastic core, which is much lighter in weight than solid cores.
Typically, I prime the surface of the metal directly with a lead alkyd primer manufactured by Natural Pigments, and then I paint directly on that. However, I've found that the primer needs to dry for a few months before it behaves properly, so it requires a lot of advanced planning.
For more spontaneous work, preprimed linen can be glued to aluminum composite panel, but you'll probably find that liquid glue options (such as acylic soft gel medium) don't work so well - the canvas tends to lift up along the edges despite every effort to prevent this. Unprimed linen can be glued to metal with little difficulty, but primed linens are trouble. When it comes to gluing primed linen to metal, the most reliable method that I know of involves the use of BEVA film, an adhesive used by art conservators. This is a heat-activated adhesive, and the use of it requires a heat press, which is the one big disadvantage of this method. I've done this successfully with small canvases, but larger works require a very large heat-press, and those are super expensive.
If you're interested in learning more about the use of linen on aluminum composite, I need not repeat what Canadian artist David Gluck has already described so well before me, so please allow me to direct you to his excellent treatment of the subject here:
http://paintingstufftolooklikestuff.blogspot.com/2013/04/panels-that-will-blow-your-face-off.html
Paint and Palette
Currently, my paint-box contains paint from Williamsburg, Old Holland, Windsor Newton, and Natural Pigments. Each brand offers a pigment, or some variation of a pigment, that I prefer over the others. For instance, only Old Holland produces Manganese Blue, and Natural Pigments produces the best Lead White. These are all serious brands that produce professional quality paint, but they're certainly not the only good brands to choose from.
Some brands, however, produce a line of cheaper, student-grade paint in addition to their professional grade; in general, you want to avoid these student-grade pigments. You'll save money, but they just don't perform anything like real paint, and they can give you the false impression that you're a bad painter when, in fact, you're simply facing the limitations of inferior materials. Good paint is expensive for a good reason. Use high-quality paint, even if you're a beginner.
I've been trending toward a more limited palette of colors, and now rarely use more than six colors (including two whites and one black).
Yeah, there's more than one white pigment. My choice of a white pigment differs from one circumstance to another, and is one of the most important palette selections to be made. One would think that any variety of white pigment would be pretty much the same as any other, but that's not true. My preferred white for most uses is Lead White, also known as Flake White. Lead White is usually quite transparent, which is sometimes useful, and sometimes difficult. But it's absolutely reliable for mixing exact colors matches for overpainting - in stark contrast to its rival, Titanium White.
Although Titanium is the most popular white pigment in use today, because of its opacity and strong covering power, I've found that this pigment darkens in value at it dries - colors mixed with titanium will shift in value, either noticeably or not, depending on whatever percentage of titanium is found in the mix. The change is not so noticeable in very bright colors - but in any color tinted to a mid-range value with titanium, I've found the problem to be totally unacceptable. Considering the popularity of this pigment and its wide usage, I'm amazed that so few painters have noticed its susceptibility to value drift. Because of this problem, I tend to avoid this pigment whenever an exact color match is required for overpainting. I typically use pure titanium only for the mixing of colors of the very brightest value, such as highlights, when a very opaque application is needed.
But I do find that it's possible to mix Lead White and Titanium White, either in a 50/50 combination, or in an even safer percentage in which the Titanium amounts to only about a third. In this case, the strong covering power of Titanium can be enjoyed, while its tendency toward value-drift seems to be reduced to an unnoticeable level.
Brushes
I employ ordinary bristle brushes (flats, typically), a variety of softer brushes (filberts, most often), as well as some very tiny brushes with fine points ("script" and "liner" brushes). A surprising amount of my work is done with the finest brushes, since I like to have very precise control over the surface qualities of my paintings. But the nature of the brushes to be chosen really depends greatly on my mood with any given piece, or on the demands of the specific subject; I've done some paintings almost entirely with large flats, which produce a very different sort of appearance. I'm not all that devoted to any specific brand of brushmaker - if a brush holds up well enough after repeated use, it's probably satisfactory for me.
Cleanup
I'm a very tidy painter, and I thoroughly clean my palette every day. Twice a day, actually, if I put in many hours of work. I like to maintain to a clean, organized palette and workspace.
Palette and Process
Before beginning a painting, I typically tone my canvas completely with a wash of neutral grey, typically a combination of manganese blue with burnt sienna, which produces a slightly warm, greenish grey.
After the undertone color has dried thoroughly, I begin the process of completing a fully-rendered, black-and-white under-painting prior to working with color at all. Color is introduced in thin, translucent layers only after the black-and-white rendering is completed. This painstaking method is historically known as "Grisaille." It's very time-consuming, but no other process that I've used has been able to so effectively deliver the results that I'm looking for. I appreciate having all of the drawing issues and value relationships resolved in the early stages of the work. This leaves me free to concern myself exclusively with color later on.
Virtually every square inch of my work receives at least two applications of paint, and more often three or four, since I'm a perfectionist and in search of a very specific result. The first pass in any area rarely hits the mark quite accurately, and requires further refinement after the first layer has dried.
In the earliest stages, I thin the paint with mineral spirits - I prefer the "Gamsol" variety made by Gamblin, since it lacks the sickening fumes that are common with most other solvents. I never use turpentine as a solvent, due to the harmful effects of its fumes - I always get terrible headaches from turpentine (in contrast, Gamblin's odorless mineral spirits seem to produce no ill effects for me). There is a balance to be achieved in the thinning of paint, since free-flowing paint is desirable, but too much thinning can be destructive to the integrity and binding power of the paint.
The addition of too much oil to the painting's surface can also lead to a yellowing of the work over time, so I strive to avoid the excessive use of oil as a thinning medium. My preferred medium at this time is a simple combination of Gamsol mineral spirits with stand oil, in which there's about twice as much solvent as oil.
There is a long-standing rule in painting, however, generally expressed as "fat over lean," which simply means that the initial, underlying layers in a painting should contain less oil than the final layers applied on top. The problem results from a disparity in the rate of expansion and contraction between the layers. "Fat" or oily layers are more likely to expand and contract during the drying process and in response to changes in atmospheric conditions - the more oil, the more lateral movement of the paint film. If the lower layers are more oily and mobile than those on top, they can pull the more brittle top layers to pieces, resulting in cracking. As a result, it's necessary to ensure that the greatest percentage of oil is to be found in the top surfaces of the picture, which can expand and contract freely without causing damage.
Once I begin to work with full color, however, I freely use more oily pigments without concern, since they'll be used only in very thin layers in the uppermost stages of the work.
Treating the Work Surface
I don't like to work on dry paint surfaces, so I always apply a thin veneer of refined linseed oil to any area where I intend to add an additional coat of paint. I always wipe off as much of this oil as possible - only the thinnest film is required to make the surface sufficiently slick, and to restore the glossy, saturated appearance of the underlying paint (which often takes on a dull, matte surface upon drying).
However, in applying oil to dry paint surfaces, I've found that the underlying layers often develop a glassy surface resistant to new oil - the oil can be seen to bead up, rather than spreading flat to coat the surface evenly. This apparently occurs because the oil in the underlying layer has risen up to accumulate at the surface during the process of drying. The more oil in any given layer, the more likely it is to develop this glassy surface - and, as in the case of the "fat over lean" rule, this is yet another argument for avoiding excessive use of oil in the early stages of the work. Since oily, glassy surfaces seem to resist new applications of oil, this would logically seem to undermine the chances of a reliable bonding between old and new paint layers - and it definitely interferes with the ability to manipulate fresh paint on such a surface - so I find it necessary to treat the troublesome surfaces before painting over them.
I do this by mixing some linseed oil with "Bon Ami" powder, a common, gentle abrasive used for household cleaning, made from harmless calcium carbonate. A small amount of powder is all that's required to perform the task. The powder is mixed thoroughly into a puddle of oil until the abrasive is reduced to tiny, invisible particles. I then use a brush, loaded with the powder/oil mixture, to carefully scour the surface of the painting in the area to be retouched.
Although the oil mixture initially beads up, it will soon begin to cling to the surface quite evenly - a sign that the powder is abrading the paint surface. If this is done gently, and without too much powder in the solution, the surface can be altered without any visible harm to the paint layer. I've found that trying to perform this same function with a solvent like mineral spirits will invariably cause damage to the underlying paint layer (for instance, removing too much paint or otherwise changing its appearance), which can be difficult to hide even with a great deal of overpainting, especially if one intends to apply transparent glazes. The Bon Ami powder is much more gentle than solvent, and has the added benefit of lacking noxious fumes. After the surface has been adequately prepared, I use a paper towel to rub away as much of the excess oil as possible. No residual trace of the powder can be detected, and the next application of paint will adhere to the preceding one perfectly.
Speaking of paper towels, I would be remiss if I failed to mention how important paper towels are in my studio. In particular, I make use of the blue Scott "shop towels" - they're readily available in stores, and they produce much less lint than ordinary paper towels. The last thing that I want to find in a wet painting is a bunch of lint - so these towels are a superior choice for cleaning brushes and palette knives while you work, and for applying oil or otherwise wiping the surface of the painting in some way. I never tear paper towerls into smaller pieces - I only cut them with scissors. Tearing releases fibers along the torn edge, but a clean cut with scissors keeps the fibers intact.