This page was revised in August, 2020. There is, at this point, far more detailed information in Living Craft about everything touched on here. At this point I feel the best type of resin is fused damar, no solvent, no weakening of the paint film, very little yellowing in years of tests. You can find a video I put up about making this medium on You Tube.
Typical questions and answers about the putty medium in practice come first, a more in-depth text follows. A step-by-step tutorial about the putty medium is here.
|quick start guide|
What is the historical origin of the putty medium system?
It has long been known that Velasquez used a medium composed of sun oil and ground calcite. In reading various publications of the National Gallery, but especially their book "Rembrandt: Art in the Making", it became clear that Rembrandt used a combination of a somewhat thickened oil and chalk in some way. How these mediums were incorporated into handmade paint historically remains unknown. I began to experiment in early 2007 making a putty that could be mixed with commercial paint and alter its characteristics to be less slippery and more adhesive. This was a technical improvement for many reasons detailed below. Developments with the medium have simply gone on, all paintings on this website since 2007 were made with variations of the putty medium.
What ingredients are used to make putty?
Basic putty is made from a combination of oil and a variety of calcium carbonate. Varieties of silica can also be used, but do not aid drying, or strengthen the paint film, the way calcium carbonate does. Avoid ALL other types of stone dust in putty: magnesium carbonate, gypsum, kaolin, etc. In order to produce a stronger paint film, the oil used in the putty should be stronger than the raw oil of commercial paint. This can be painter-refined organic linseed oil, and/or a semi-heat bodied oil, and can include the addition of a small percentage of thicker oil such as sun oil. There are many variations possible within this deceptively simple recipe, these result in paint with different rheologies or types of behavior under the brush. It is also possible to incorporate a small amount of whole egg, egg yolk, or egg white. All "other" additions must be in small amounts in order for the inherent strength of the oil-chalk mixture to remain dominant.
What are the advantages of the putty system?
Using a putty allows the painter to capitalize on the convenience of commercial tube paint but also to overcome or at least ameliorate its drawbacks. A great many different consistencies of putty may be used, producing a significant range of techniques from smooth and detailed to roughly broken impasto. Putties may be engineered to any level of final saturation and gloss. Finally, the putty system contains no resin and allows the painter to work in a completely solvent-free environment. Brushes are kept in oil as has been documented in older painting practice and cleaned with a rag before use.
This is so simple. Why haven't I heard of this before?
The research linking this idea to Rembrandt's work is relatively new. There is also still a prevailing misunderstanding that resins are universally a part of older painting practice when resin use has in fact been proven at the molecular level to have had very limited use by painters such as Raphael, Rembrandt, Van Eyck, and Vermeer. Also, the putty medium is inexpensive, needs to be homemade to be fully effective, and extends your paint supply. What manufacturer would promote an item like that?
In what proportion is the putty mixed with the paint?
You determine this proportion based on the pigments in your palette, there is no functional limit. One part putty to two to four parts paint is a good rule of thumb, this depends on the strength of the pigments, whether you're underpainting (where more lean putty is possible) or overpainting (where less rich putty works better), and the way the style uses colour.
How does the putty affect the paint optically?
This depends to a certain extent on the type of calcium carbonate and oil used. Marble dust tends to be whiter but also more opaque than chalk. Increased saturation can be achieved through a small amount of thicker oil in the putty. There is often a greater sense of life or aeration in the pigment because the particles are more separate. It is also possible to make textured putty "glazes" in the manner of Rembrandt, which are optically between a transparent glaze and a scumble.
How does the putty affect the consistency of the paint?
This depends entirely on the consistency and ingredients of the putty, paint can be made with any degree of mobility or adhesion. A chalk putty is more thixotropic and mobile than a marble dust or whiting putty. Putty can be used quite densely with bristle brushes, or more finely with softer brushes.
Why should I bother to make the putty myself when there are commercial alternatives?
Manufacturers are in business to make money. They combine raw ingredients and resell them at often large profit margins in the case of art materials. Especially in the case of the quality of the oil, there has been degradation over the years as a result of this system. When you make the material yourself, you will have access to much greater quality for the same price by eliminating the middlemen. You are also able to work with a much greater variety of ingredients and subsequent effects.
Does the putty dry fast?
Calcium carbonate -- chalk, marble dust, whiting, calcite -- is often spoken of as a secondary drier. Used with painter-refined organic linseed oil refined, this putty can dry overnight. A putty made with walnut oil takes a bit longer. Putty can be enhanced with a small amount of stand oil, safflower, or walnut oil for longer open time, or sun oil for shorter.
How well does the putty medium hold up over time?
The single technical error Rembrandt made was in the use of smalt as a drier, which ultimately turned brown in many cases due to the acid in the linseed oil. Once the older mastic varnish is removed, Rembrandt paintings are in a technical condition which conservators often remark on as being excellent. The paint film produced by putty enhanced paint is remarkably resilient and stable.
How should I paint with this medium?
Painting with putty is very flexible in terms of system because of its proven ultimate stability. Work can be done from loose to tight on stretched canvas by starting with a loose putty and dipping the brush from time to time in chalk as the work progresses. On panels, painting can also be done from tight to loose as well, including various types of ploughing or carving. For various putty method slideshows, go here.
What sort of learning curve can I expect after putting this medium into practice?
The best way to begin with this idea is just to make small studies using a stone dust and an oil or oil mix on the palette. The putty enhanced paint is inherently adhesive, so doesn't need to be as thick as one might think at first. Simply adapt the consistency of the putty-paint to the way your hand wants to work and learn more from there, the range of possibilities is endless: work can be done from loose to tight, tight to loose, or both. For further information, continue below.
There are two distinct ways of painting with oil: one uses resins, one does not. The permutations of both methods are essentially endless. The origins of oil paint as a decorative medium made with boiled oil used outdoors make early recipes containing resin logical. The resin in this case was a thick cooked oil varnish such as amber, copal, or sandarac, which would contribute to the gloss and durability of the paint film. The use of uncooked resin varnishes -- spirit varnishes, resin dissolved in solvent -- in older painting was generally limited to a spirit varnish being used to help a specific glaze -- such as madder lake -- dry, and the work early painters such as Durer and Leonardo did with a final varnish of sandarac dissolved in spike lavender. In an era in which the formal perfection of a work was a sine qua non, technical simplicity was only logical and wise business practice.
By the early 19th century is was clear that something valuable had been lost from older painting technique: in England the many technical issues with Reynolds paintings (1723-1792) as he struggled to find the answer made this uncomfortably clear. Books began to appear which speculated about what these older techniques were. One such book is Eastlake's famous "Methods and Materials" (first volume, 1847) but this comes after a book in French by Merimee, "The Art of Painting in Oil and in Fresco," translated into English in 1839 . Both of these books conclude that a hard resin varnish such as amber or copal was the foundational component lost from older painting. This led to a tremendous revival of interest in amber in later 19th century painting in England, Parisian academic painters kept using copal. But, as is shown in the tremendously well-researched "The Artist's Assistant" by Leslie Carlyle, neither Eastlake, Merimee, nor the more conservative colorman George Field, author of "Field's Chromatography" in 1835, could stem the tide of complex medium formulas designed to make the new tube paint of commerce behave like the craftsman-made paint of an earlier time. While certain painters of the period, such as Constable, wrote in favor of simple ingredients, the most common recipes collected by Carlyle from the 19th century English sources are for various gel mediums based on mastic resin dissolved in turpentine combined with a leaded oil. This quick setting, easy to use formula, revived by French researcher Jaques Maroger in "Secret Formulas and Techniques of the Masters ,"1948, is perhaps the single most controversial formula ever developed. When fortified with copal varnish, as in the well-known Roberson's medium used sometimes by the Pre-Raphaelites, it has proven at least moderately sound over time. But in spite of the fact that varieties of "Maroger Medium" continue to be used by painters today, it is almost universally deprecated by responsible painters and often seriously lambasted by modern writers for the inevitable instability of the complex paint film it produces. Not in a year or even a decade, but over the longer time period expected in the life of a traditional painting.
Throughout in the 20th century, painting professors write books about painting, including Max Doerner and A.P. Laurie, both of whom are aware that something has been lost from older practice it might be a good idea to recover. Ultimately, the materials and techniques of older painting are either openly deprecated or ignored by later writers: no plausible modern guide to older painting technique exists. The modern system of using damar varnish and commercial stand oil, first written about by Max Doerner and promoted by Ralph Mayer in The Artist's Handbook, is not based on any known older practice. In over thirty years of looking at older paint films, technical art history and conservation research has never found soft or hard resin globally in the paint of a well-known older painter. Damar first makes it's appearance in early 19th century German painting, and a vacuum bodied oil like stand oil was beyond the technology of the 17th century. One might think that some sort of analogue of this system would exist historically, but, according to the findings technical art history, global use of resin in the paint film comes into play in the 19th century. The closest older painting comes is the egg, oil, and resin medium found in some Lotto paintings, the copal found in Gentileschi, and the mastic found in Bombelli. So, the whole secret resin concoction as the lost secret of older painting story is fiction.
Enter the modern conservator, armed with gas chromatography, mass spectrometry, and a host of allied high-tech paint film analysis techniques which enable the detailed analysis of minute samples. In Technical Bulletin #15, (1994) London's National Gallery published an extensive article titled "Rembrandt and his Circle: Seventeenth-Century Dutch Paint Media Re-examined." The conclusion was that, while some of Rembrandt's pupils used amber varnish in isolated cases, Rembrandt himself did not. With the exception of the occasional glaze using resin, Rembrandt's paintings were made with linseed oil. The most common additions were chalk and egg, they've also found starch and cherry gum. This same group of conservators and scientists, writing in the Rembrandt volume of their great Art in the Making series, makes the same case with more paintings in more detail. This conclusion proved so controversial with Rembrandt scholars that in the updated edition of the book in 2006, the authors go out of their way to point out that they can and do find resin, it's just not in Rembrandt's paintings. The same has been true of Van Eyck research: the medium was linseed oil, egg white may be involved in some way, but it may not. The recent Bosch research also says linseed oil, and is more definitive about egg. The same has been true of Vermeer research, linseed oil with protein, probably egg, possibly a hide size like rabbit skin glue.
Reading all this for the first time, long ago at this point, I also felt challenged. I had spent almost six years navigating the labyrinth of older sources and materials. Having figured out how to make amber varnish, copal varnish, sandarac varnish, and all kinds of egg emulsion mediums with them, having worked with Roberson's medium and done a few studies in the mastic gel mediums, I was attached to what I had learned. Was it all wrong? More importantly, was it all totally unnecessary? A few uncomfortable weeks after reading the Rembrandt book in the Art in the Making series, I decided I had to give the other side of the coin a fair trial. It was time to try green eggs and ham.
So, beginning in 2007, I worked on two things: how to make a successful all oil medium with the few traditional additives, and how to refine linseed oil so that it would dry quickly and not yellow. Then, in 2008, I started the process of putting those two lines of inquiry together.
I decided to take on linseed oil separately because, like many modern painters, I had become very suspicious of modern "alkali-refined" linseed oil, which is sometimes seen oozing from the corners of a paint tube in all it's dark orange glory. Not exactly an Old Master product. So all my working experience was with walnut oil: the oil of Leonardo, Raphael and Perugino. Not as fast to dry, not as strong a paint film, but reliably less yellowing.
The first time I worked with an oil and marble dust putty in a painting I realized that it was all possible. After six years in the labyrinth of older painting technique, a large, economy size Philip Guston light bulb went on over my head. Many mixtures, a year of trial and error later followed: once again, simplicity proved inherently complex. What follows is a description of this technology in action. It flies, quacks, swims, has feathers, webbed feet and a bill. Is it, then, a duck? Without a time machine, this is not for me to say. Hopefully, in the best older tradition, you will simply let your own experience with the materials be your guide. This is the essence of Rembrandt's advice to Van Hoogstraten: the authentic craft develops naturally from one's own experience. So, it seems reasonable to suggest that the search is not for the lost secrets, but for one's own practice. This is in fact easy, you start making things. At first they might not be perfect, but the information here will give you with a running start. And, if you are cut out for this, the learning curve will not be steep, because you will realize that you are finally headed in the right direction: towards the living craft that you make yourself.
|medium and system|
The ingredients of this medium are simple: an oil or combination of oils, chalk or marble dust, a small amount of egg yolk, egg white, or whole egg. When working with the medium at first, it's best to keep it simple so that you can develop a solid understanding of what is doing what. Putties can be complex, but do not need to be to work well. At first, it is best to make small amounts of a simple putty on the palette and then mix a specific amount of putty into each nurdle of paint. Three or four parts paint to one putty works well for most colours, strong can colours take more putty, brighter colour schemes can be finished with less putty that is richer, etc. The putty lightens the colors slightly, but not in the manner of white. It's more like adding air or light: Venetian Red becomes brighter and more orange. Texture can be made soft or or crisp, impasto low or high, all controlled on the palette by the introduction of more oil or more stone dust. It may be logical to make light colors thicker, dark colors thinner. How much mobility, how many layers, how much texture, where and when, all can be dialed in. This is easy to conceive, but can take time to implement in practice.
The use of the putty offers two technical advantages. The first is that it gives commercial tube paint an element of thixotropy or viscosity or "boing" that trumps the false "congealed grease" consistency of most tube paint, the fabled "butteriness" from aluminum stearate which is increasingly thought to cause darkening over time. The second is that it allows the painter to dispense with solvent and keep the brushes horizontally in a tray of oil, as is well documented in Old Master practice. I use raw walnut oil for this. Brushes are simply cleaned on a rag before use, clean well if you use safflower oil: it dries very slowly. Please do not use linseed oil for this! Rags soaked in linseed oil have a long history of auto-combustion due to the heat released during the polymerization of the oil.
Oil: Not getting into hand-refined oil here, there are four basic oils for putty. First in volume is Spectrum Naturals Walnut Oil, available in quality grocery stores or health food markets, about 8 dollars a pint. You can also get refined walnut oil in bulk from soap supply places, a good place on the east coast of America is Jedwards outside Boston. They have a great refined walnut oil from Italy. Specify the Italian oil, it's better! Second is Allback Boiled Linseed oil from Sweden, available online here. This is a traditionally refined organic cold-pressed oil, an unusual product that will give you an idea about the difference in hand-refined linseed oil, about 7-8 dollars a pint. Don't let the word 'boiled' turn you off, this is very thin. The third is sun oil, either linseed or walnut, made the traditional way by exposing a tray of refined, it must be refined, oil to the sun for a long time in the summer. It needs to be covered with glass but also breathe. Bugs and debris still get in, you heat the oil gently and sieve it all out in the end. You can buy Sun Oil, but it is expensive, and, surprise, may not be sun oil at all! You can substitute Stand Oil for this, the Kremer 53201 stand oil is made with high quality oil, but stand oil is overkill in terms of polymerization, is beginning to accumulate caveats from conservation. Long term, and in terms of giving you more period bounce, sun oil is better, and yellows less. The last oil is sort of a misnomer, called Burnt Plate Oil. Only it's not burnt like traditional burnt plate oil, it's a bodied linseed oil used in printmaking to cut the viscosity of the ink. The volatile parts are sparked off as they rise in the chamber, resulting in progressively thicker oils which dry slowly but are non-yellowing and add an unusual increase in pigment saturation. This oil is very cheap and available from Graphic Chemical here. I use mostly #5, which is about the consistency of stand oil, although sometimes a little #7 as well, which is thicker. It is incredibly concentrated, a little goes a long way, more is less.
The Spectrum Naturals Walnut Oil is further processed by heating. One of the things that came out of reading the National Gallery Technical Bulletins was the concept of the semi heat bodied oil:
"Heat pre-polymerization has several effects on the oil. Drying properties are improved and are further enhanced by the addition of metal salts (usually those of lead) during the process. The refractive index of the oil is increased, thus reducing light scattering at the pigment-medium interface and thereby increasing the saturation of the pigment colour; the paint film may also have a glossier appearance. The pigment is less liable to sink in the oil film, which itself decreases less in volume than a conventional oil film, reducing the amount of wrinkling that may occur. White paints appear less discoloured because, as the polyunsaturated fatty acids initially present in the paint film are destroyed by the formation of carbon-carbon single bonds, there is less scope for the formation of chromophoric and auxochemic groups, the presence of which give the yellow appearance to the film." From Rembrandt and his Circle: Seventeenth-Century Dutch Paint Media Re-Examined by Raymond White and Jo Kirby. National Gallery Technical Bulletin volume 15, 1994.
There are two basic ways to do this. One is to heat the oil to a high temperature for a short amount of time: thirty minutes, an hour, two hours at 180-190C will produce an oil which is noticeably thick. The oil is never actually boiled, and should only be heated just to below its smoking point using electric heat outside. This method is documented repeatedly in the oldest existing texts on oil painting. The other method is to mimic the effect of an oil aged for a long time or kept near a stove during the winter. This involves controlled low heat for a longer period of time. I've made this oil in the 48 to 96 hour range, temperature range of 204 to 245 degrees Fahrenheit. At this point, because of the other oils in the mix, I'm using the 48 hour oil and the lower end of the temperature range. The simplest way to accomplish this is with the small "dip heater" Crock Pot, which heats about a pint of oil to about 245 F. Twenty-four hours would be a good amount of time for this. You could also get a little more high tech and use a small deep fryer with a thermostat. If you are in funds or know someone in a lab, you can use a magnetic hot plate with a flask on a ring stand and a stir bar. There are often older ones by Corning from college labs available cheaply used. Always do this outside in a safe place!
So, in the tube paint, the oil is raw: weaker, more volatile, more likely to yellow. In the putty NONE of the oil is raw: stronger paint film on polymerization, more stable drying characteristics. The putty provides a stronger, higher quality vehicle than any commercial manufacturer has ever put in their paint.
Stone Dusts: A great variety of inert ingredients can be used, and some people just have to look under every rock. Having done so, it seems safe to say that the most important ingredients here are the calcium carbonates: the ground calcite of Velasquez, the chalk of Rembrandt, Chardin, Vermeer, Cezanne, and the readily available marble dust. The Venetians used ground silica, cristobalite is more transparent than chalk but doesn't aid drying and fine silica does present a health hazard. There's also fumed silica such as Cab-o-sil, or Aer-o-sil in Europe, this produces a gel in oil that makes everything more mobile in a hurry, but needs to be used in small amounts and definitely handled with a particle mask. Ground leaded glass was used, but this is never that fine, adds nothing in the way of "sparkle", and any putty medium dries well anyway. Blanc fixe, barium sulphate, was often used as an extender in 20th century paints, especially cadmiums, but blanc fixe yellows in linseed oil. A final dry ingredient of possible interest in very small amounts is bentonite, the colloidal clay from which Tix-o-Gel is made. Not that light, but a small amount produces a great increase in spring. But, a lot of other things produce thixotropy, and clay attracts water, which is not the best thing in a paint film, so do tests with this one first.
Fredrix marble dust is available everywhere and makes a pure white putty with a deceptive grippiness or drag: it can be made quite thin and still hold. There's also a nice calcium carbonate from Duda Energy on Amazon. A natural chalk putty will absorb more oil and be more springy under the brush or knife. Rembrandt's late manner is an example of chalk used in an amazingly accomplished way. Chalk putties tend to be beige or gray, whether this color has a bearing on the resulting pigment color depends on how much you use and how bright the palette is. The best natural chalk I've found is from Kremer Pigments, 58162 stone chalk, this is 4 microns and relatively light in oil. They're often out of this, don't let them sub another chalk of you want something light!.
Egg: Adding even a small amount of egg to a paint or a putty makes it begin to seize, creating the possibility for more control or layering, possibly at the expense of vivacity or oomph but not always. Egg yolk also accelerates the setting and the drying times, and can ultimately contribute brittleness in larger amounts. Working on panels, this is not an issue. Working on unsupported canvases, especially if large, egg should be used minimally. It is such a strong element in the mix that it can still contribute. Egg yolk results in a putty which is more glutinous, beaten whole egg results in a putty with a strange combination of some initial flow but a quick set as well. Many avenues of enquiry here, just be safe and make sure the oil predominates unless working on panels.
|sample putty formulas|
Although there are some examples of set recipes below, the system is amorphous. The best way to work with the putty at first is in small amounts made to order, using the oil and chalk separately on the palette as well to make adjustments. This will develop an understanding of how the system works much more quickly. Once you get comfortable with a given formula, you can make a larger amount: the recipes below are for 100ml tubes. You can also just put the putty in wax paper, make sure to get all the air out. Wrapping this in Duct tape and inserting a pin or small nail in the end produces a sturdy way to get very fine control over how much putty goes where.
1 cup chalk
2T 72 hour walnut oil
4T 48 hour walnut oil
2T Allback boiled linseed oil
1t sun oil
A nice combination of oils for general work, good amount of bounce, or boing, dries well. Chalk putties absorb more oil but also will break if stored. This is not an issue in practice, comes back together by simple mixing.
Marble Dust Putty with BPO
1 cup Marble Dust
1t BPO #5
1T sun oil
3T+1t 72 hour walnut oil
A little loose, a little gluey, dries with an increase in depth from the BPO.
Smidge of Egg Putty
1 cup Marble Dust
4T 72 hour walnut oil
1T sun oil
1t whole egg
More set from even this amount of egg, tight detail, clean line. More egg would cause more seizure, the need for more oil to make it move. This type of putty can get a little rococo in practice, fun outside or for loose work. Can be tubed without going bad.
Prehensile Putty Underpainting
Make a dense putty with some thicker oil such as boiled or sun oil, add a small amount of a chosen color like raw sienna. Put this thinly and evenly on a panel with a large knife, carve into it to draw. A very layerable surface results. Can also be thinned with a little solvent for more control.
As you become familiar with this system and group of ingredients, remember to let the energy of your hand be your guide. Asking "What wants to happen here?" always gets an answer that moves things forward. The more you pay attention to the materials, the more their behavior will tell you what to do. Keep a small container of chalk on the palette, it stays clean if you dip into it with a wet brush and will tighten up a passage. Too tight, try a little extra oil. Remember also to tune your brushes to the consistency of the paint: fine brushes, soft putty, firm brushes, firm putty. Fine brushes can do firm putty, but make sure they're synthetic. You can also set up a system of firmer underlayers to softer overlayers and carve back into the firmer underlayers. On it goes. Hopefully it will provide you with the same sense of finally being in a grounded yet versatile place with the paint that it's given me.
The conservation research that led to this way of working is relatively small, and listed below. A more extensive bibliography is at the end of the formulas section.
Rembrandt (Art in the Making Series). David Bomford, Jo Kirby, Ashok Roy, Axel Ruger, Raymond White. National Gallery Company, London. 2006
National Gallery Technical Bulletins: Volume 15, 1994, article on Rembrandt and his Circle. Volume 20, 1999, extensive cover article on 17th Century painting in Antwerp and London, technique of Rubens and Van Dyck.
|a note on magnesium carbonate|
Magnesium carbonate was used as a filler in student oil paints in the 20th century. Recent conservation research has shown that magnesium carbonate in the paint film generates the mineral epsomite over time, whose presence makes the paint water-soluble. So magnesium carbonate is something to avoid completely.
A text handout from a workshop using the putty medium in the manner explained by Hoogstraten.
Older painters were often faced with complex issues regarding pigments that we are not. Pigments could be toxic, fugitive, incompatible with specific other pigments, in short supply and/or tremendously expensive. When painters were working for the church, cost was typically not as issue, but as an apprentice or student it would be logical to learn using a palette of inexpensive and readily available pigments. The natural earth pigments -- yellow, brown, orange, red, and green -- plus black and white provided this with the added bonus of being permanent. A system evolved which made the most of the lower chroma of these pigments by using them transparently and whenever possible without white. Value was often emphasized -- strong nearly white lights, strong nearly black darks -- and within this stark monochromatic envelope the earth colors are perceived by the eye as being more lively. In the hands of skilled practitioners -- painters as diverse as Giulio Romano, Velasquez, Rembrandt, Hammershoi, and Morandi have specialized in earth pigments -- it is possible to see the many variations of color-style possible.
An important contribution of this palette was the development of a way of using color abstractly. Older painters did not copy colors, because often their palette did not provide that ability. Instead, they came to an understanding of the way a triad of red-yellow-blue can be used to create the illusion of dimension on a flat surface. Even if the yellow is ochre, the red is burnt sienna, and the blue is black, a convincing 3D reality can be made. So, older painters were often interpreting color, but were copying color relationships themselves more exactly. This worked because all color in a painting exists in a relational way to the eye.
In any painting system white operates as the lightest shade of blue. When we add white to a color, we make it lighter, but also cooler. The earlier systems took advantage of the blueness of white by using a warm dark transparent brown for the shadows. The relatively orange shadows and the relatively blue highlights easily made for a convincing depiction of space in spite of the fact that light is rarely white and shadows are rarely brown.
Another logical outgrowth of this system is the idea of not mixing the color on the palette, but on the painting itself. Whether the painting was begun on a midtone ground or white, the darks were laid in transparently first, with the lighter colors coming next. Color using white was last. The sequence of colors was always from dark and transparent, to translucent midtones,to cool and opaque lights. Spatially this is logical because the light is literally "on top of" the dark.
Although the earth colors lend themselves well to this way of working, it can be done with any set of two triads of color: one dark and transparent, the other lighter and more opaque. It can of course be done with one triad as well, and this might be a good way to start if you have a favorite set of three primaries, but the dual triad method gives more possibilities.
This way of working can be inferred from the very small palettes which older painters used, and is written about by Samuel von Hoogstraten, a 17th Century pupil of Rembrandt. As a method it takes getting used to because you are beginning by making something relatively strong and chromatically jumpy -- on purpose. The brighter the colors, the jumpier things can get. This leads to brighter color accents in the end via the overbold beginning. So in working this way it's good to keep in mind that a painting only has to look good once: when it's done. A typical dark triad would be raw sienna dark, a modern crimson lake, and ultramarine blue, put on in that order. The consistency of the paint is determined by how much you want to accomplish in one sitting. Looser paint allows for more blending but less layering, would be good for developing lots of mood in a first layer on something to be more detailed over time. Tighter paint -- more chalk -- can be used on panels to create areas of color which are more discreet, to the point of being able to be carved. This paint can then be loosened bit by bit -- more oil -- in subsequent passes until the correct paint viscosity for finishing the painting is achieved.
By becoming familiar with various consistencies of putty, by using small additions of a denser oil such as sun oil or stand oil, by adding a very small amount of egg white to the putty, an almost unlimited number of variations in texture and paint rheology -- viscosity, handling characteristics -- can be achieved. The putty also allows pigments which are tremendously strong -- such as the phthalos, or titanium white -- to be cut in such a way that they are easier to work with. The putty can be used with a very dense pigment such as Mars Red to help aerate it visually. The white of older painting -- lead carbonate -- is 10 percent as opaque as titanium white. And this was often cut further with chalk, especially in the beginning of the painting. Putty with a very small amount of pigment can be used to make a translucent veil over dry paint, different in feeling than a scumble or a glaze.
Colors on the palette are all cut with putty. The white should be cut in several increments, the work with white is from the more extended (transparent) to the less extended paint. The colors are organized around two triads, one of which is composed of dark and transparent colors, the other of colors which are lighter, brighter, and more opaque. Colors can be chosen for the maximum warm-translucent/cool-opaque difference or for their ability to create a certain situation in daylight. It is up to you whether to work from a beginning which is fluid -- loose putty, moves easily -- or a beginning using more chalk and thicker paint. It's a good idea to become conversant with both methods since, while they give different results, they both work and can with practice be mixed within a painting. Paint which is too loose may run or preclude layering, paint which is too tight may give results which seem frozen or stiff. The key here is to learn to move slowly away from your initial viscosity: if thin towards thick, if thick towards thin.
Example Dark Triad: Raw Sienna Dark, Crimson Lake, Ultramarine Blue Dark.
Example Light Triad: Yellow Ochre, Golden Ochre or Venetian Red, Cobalt or Manganese Blue.
A typical beginning color on a white ground would be a raw sienna or raw sienna dark.
The same brush can be cleaned with a rag or not and then be used to make a pass with the Crimson Lake where things are darker -- red as part of a shadow -- or red as the local color . Then a pass with the ultramarine can be made, working first from the darker shadows towards the lighter areas which are most truly blue: the brush will become less purple and more blue as it is used. The idea is not to blend the colors at this stage, but to leave them in a more pure and patchy state: the blending comes later. The general feeling of the painting is worked out at this stage with the dark triad and no white.
For a larger or more detailed painting the work can be stopped once this stage is complete and reviewed the next day while the paint is still somewhat wet but capable of more subtle manipulation.
In the next pass with the lighter triad, some blending will occur unless you use softer paint with softer brushes or have made the earlier paint quite tight. I tend to work from the darker colors to the lighter: for example putting manganese blue where I'll want bright green, then coming back on top of that with yellow ochre later to make the "actual" color.
After the pass with the lighter triad, the painting should seem a little more normal looking although still on the dark side in terms of value, on the vivid side in terms of color.
With this method the blending takes place naturally as more paint goes on. This is where the mobile set of chalk putty helps greatly, allowing paint to be laid on top or mixed with pressure. Underpaint can be excavated, the tip of the brush can be used to restate contours. At first while getting used to this system you may overshoot the mark and end up with something which goes too far into blending or finish, and simply looks realistic. But as you gain experience you'll begin to see and feel where and how to leave fragments of color behind as traces of where you've been in the layers. The longer you can stand for it to look "unfinished", the better chance it has of showing you something new.
Another set of triads that were often used has earth colors for the first triad, and vivid primaries for the second. Painters such as Fra Angelico, Vermeer, and Constable in his studies used this type of palette.
Dark Triad: Raw sienna or Trans Yellow Oxide, Mars Red, Burnt Siena, or Trans Red Oxide, and Black for the blue. The bluest black I've used is the Blockx Ivory black.
Lighter Triad: Primary yellow or similar, Cadmium Red or similar, Cobalt blue or Ultramarine.