The painting shown above is van Dyck’s Portrait of Cornelis van der Geest, now in London’s National Gallery. He was around 20 years old when he painted it. After you’ve caught your breath, read on.

The painting is just under life-size and I recall reading somewhere that van Dyck carted it around on his travels in order to show his talents as a painter to potential patrons. Apparently he had a custom box made for the purpose. As you can read in the gallery’s write up, the painting was enlarged years later. This post is all about the technical aspects which help it and many of his paintings appear so striking.

Support and Ground
Any consideration of painting technique needs to begin with the support and ground being used for the painting. These two aspects directly affect the final result more than the general public would guess.

Preparing one’s own supports and grounds is now uncommon, both because of the ease of acquiring ready-mades and the time and skill needed to make them by hand in the studio. Although it is true that in van Dyck’s time there were materials suppliers, back then art materials manufacture was regulated by a professional guild, based upon standards set by artists.¹ Today, acrylic-primed cotton canvas is almost ubiquitous and because of that few artists are even aware of what they are missing. Canvas, sized with hide glue and primed with lead, or panel primed with chalk (and then possibly lead) are dramatically different to paint on than what you’ll pick up in the local or ‘big-box’ internet art store. The absorbency of the ground is a feature of 17th century oil painting methods which is very rarely put to use nowadays and artists who attempt to copy Baroque-era paintings (Rubens, van Dyck, Velazquez, etc.) often fail to realize that half the battle is in that feature alone.

The van der Geest is on an oak panel. So why did van Dyck choose a panel support as opposed to canvas? Due to its smoothness, a properly prepared panel allows the artist to achieve a tighter finish than what is normally possible on canvas. I also believe that artists of the past were more aware of paying attention to where their viewers would be seeing the painting from than we are today. And that’s not just for large frescos. While clearly conjecture on my part I would say that given the small size of the van der Geest a panel support was chosen because the painting would likely be viewed at close range. Seeing canvas weave would destroy the illusion. In addition, had van Dyck intended the painting to be a show piece all along, panel was a safer choice given his travels.

Below is a detail from a self portrait his painted a year or so after the van der Geest.

This painting is currently in the Alte Pinakothek and I am purposefully showing only a detail here so that it can be viewed similarly to the van der Geest. One of the initial impressions is of canvas. In fact, the painting is not solely of a head, it’s a half figure and is most easily viewed as a whole from farther back. Bear in mind that any version of a painting viewed on the internet or in a book cannot possibly elicit the same effect on the viewer as it does in person but you can view it as a whole on the museum’s website here.

Colored Ground or Imprimatura?
Another important aspect of the ground is its color and this is especially true for Venetian and Flemish paintings. While they may seem similar, a colored ground and an imprimatura are not. A colored ground, also called a campitura, is opaque and is simply composed of colored pigments mixed into the priming before it is put on. An imprimatura is colored pigment as well but it is added on top of the dried ground in a thinned, translucent way. Think ‘stain’ and you’ve got it. Imprimature² are not always visibly even, which we can see in Rubens’ Lion Hunt, again in the National Gallery, London (detail shown here). His streaky imprimature often show through in the shadow areas of his paintings.

It seems that van Dyck used both a colored ground and an imprimatura, together, in many of his early paintings and that they were previously sized with hide glue. The resulting color was often a light, translucent gray but a bone or biscuit color was also common.³

Palette
Van Dyck’s color choices were naturally more limited than what we have available to us today. However, he had access to most of the same natural earth and mineral pigments that we do. His white was always lead and for this we generally have a far lesser quality of selection than he did (and we now also have to deal with anti-lead concerns, whether they are valid or not). His blacks were charcoal and bone black.

Van Dyck also used lakes.⁴ Lakes are paint made from various plant and other organic matter, like a color called Dutch Pink (which is really a yellow).⁵ They are, for the most part, fugitive. This means non-permanent at best and damaging to other pigments at worst. While one can still purchase the base materials of many lakes, finding ready-made lake colors is difficult.

White
I am calling out a section on van Dyck’s white as his impastos are such an obviously important feature of his work. The image below is a detail from van der Geest’s right cheekbone, showing the thickness of the white on the ruff.

The reasons for the thickness of his whites as well as their leveling properties is undergoing some debate and this is not restricted to van Dyck but also to other Dutch and Flemish artists of the period. The known factors are lead, particle size and the oil being used. The lead white of van Dyck’s time was more varied in particle size than anything we have mass-produced today.⁶ Researchers are convinced that the oil in van Dyck’s whites was usually linseed and, as I comment on below in the section on mediums, bodied or heat-bodied. Distinguishing his vehicle (what the paint was mulled in) from the medium is likely impossible without a known control sample so any conclusion may only be a guess. Knowing how easily Stack Lead White handles while being mulled⁷ it is quite possible that van Dyck mulled his whites in his medium. If true, this would allow him more control over their thickness and settling properties because the base white would not have been pre-wetted via an initial mulling in linseed or walnut oil.

Van Dyck may have added calcite to his white (and some of his other colors) to adjust their handling properties.⁸ Calcite adds volume without color and can also help the artist fuse edges. Due to its prevalence in Velazquez’s paintings calcite may have also been part of a medium mixture but here again this is difficult to prove.⁹

Another theory is that he added a protein to his white. This means egg (yolk, white or both). This theory is not without merit as tempera preceded oil painting and in the sixteenth century it was somewhat common to begin a painting in tempera and finish it in oil. Given the age it is likely that most artists knew how to do both. Adding egg to oil paint is a different story however and although promising this possibility needs more study.¹⁰

Finally there is burnt plate oil which is something etchers know a lot about. Remember that both van Dyck and Rembrandt were etchers as well as painters. Although I know of no studies about burnt plate oil and I have no experience with it, the images on this page are suggestive. Here again this is something in need of more scientific examination.

A close up of the forehead in van der Geest. This is impasto composed principally of lead white.

Medium
The subject of mediums is heated and the artist’s choice a potentially risky one.¹¹ It seems that van Dyck generally used nothing more complicated than linseed or walnut oil. For certain colors, notably his whites, he apparently used a bodied or heat-bodied linseed oil.¹²  This may mean thickening in the sun (usually termed ‘bodied’) or stand oil (‘heat-bodied’). The difference is that when sun-thickening the oil is exposed to the air (possibly in lead trays or with the addition of powdered forms of lead) and as such begins the oxidation process.¹³ Stand oil is produced over heat in a vacuum sealed container, therefore it is artificially thickened but not oxidized. The sun-thickened oil process was known long before stand oil. As I mentioned earlier, relative to his vehicle (what the pigments are mulled in), it is also possible that calcite was added to his medium.

Procedure
The following is attributed to van Dyck and is found in a manuscript titled, Observations of Ant Dykii. It is quoted directly and in full in the 1990 van Dyck show catalog¹⁴ and that is where I am quoting it from. Whether van Dyck followed these procedures himself or not is up for debate. We know that he had a workshop, as did many painters of that time, and at a minimum these procedures would be solid guidelines for his assistants. The quotes below are italicized and my comments are not. I have also rearranged some of the quoted text for clarity.

It would be desirable that the artists at the start of their work would pay particular attention to all of the following issues (namely, sketching, dead-colouring, painting, shadowing, heightening, final touching).

Sketch
It is important to sketch in the forms, so perfectly, that afterwards there will be no reason to make a change.

-The word sketch here implies a painted sketch, not one drawn in charcoal or other non-paint medium.

Dead-colouring
Dead-colouring is called “la maniera lavata” because it is a washing process; because the space within the contours seems to have been touched up with a wash only.

He must, following the rules, choose lean colours (tempered colours), so that the first layer, when it will be dry enough, will have a light tone; he has to see to it, of course, that those colours used in the dead-colouring are basically the same as those with which he is going to do the final painting, that is, that the dead-colouring of nudes will always be done with carnation colours, and clothed figures will have a dead-colouring in a darker paint, without trying to accomplish in this rough work a degree of finish which will closely anticipate the final colour; it must be noted that this dead-colouring should be lighter in some places and darker (browner) in others as is seen fit by the artist in view of his final conception.

-’Dead-colouring’ means the color of a corpse in old Dutch. Looking at a number of van Dyck studies one can see that his version of ‘dead-colouring’ was not simply a grisaille (in grays). Rather, it was a colored underpainting composed of lighter colors similar to those intended to be in the final painting. I would call this stage a lay-in. Below is his study of The Princesses Elizabeth and Anne, currently in the National Galleries of Scotland.

Notice that the flesh color is slightly dull and that it is not in grayscale. The thinness of the paint is best seen on the girls’ foreheads where the ground shows through. You can also see areas of ‘the sketch’ as described above. Based upon the priming showing in the upper-righthand corner, this canvas was toned with an imprimatura, not a colored ground.

Modelling
The second manner is called “la maniera sbozzata,” or the modelling or forming process; because it is supposed to give the entire work its final form.

The painter should realize…, when the dead-colouring is dry, that when he gives final form to his figures he will use darker colours; which means that he will pay special attention to a correct rendering of those parts of his final modelling in accordances with the pre-modelling he did in the dead-colouring. But here the painter should see to it that he will not attempt to accentuate the round forms with highlights. Which means that the round forms should not be brought forward by highlights but only through the use of darker (browner) colours.

-Taken together, la maniera lavata and la maniera sbozzata clearly imply a series of layers complete unto themselves. This description is not window shading, which means finishing one part at a time.

Finish
The third one is called “la maniera finita,” which is the finishing or completing process; because this one gives the work it’s final touch.

After the round forms have dried the painter should put in the darkest shadows, whereby the colours of the clothing should be glazed in some places with lakes, ultramarine, and other colours in accordance with the work. This is the way the work was carried out by Titian, Giorgione, Pordenone, Palma Vecchio, and many others who left us such extremely fine works.

-Here again we see in the system a series of layers, each interacting with and accenting those below. This is not alla prima painting.

Variation In Paint Thickness and Its Edges
Once again, let’s compare van Dyck’s van der Geest with his self portrait. Below are details of the eyes on both paintings.

Clearly there is thick paint. But what may not be clear, especially in the van der Geest, is that the thick paint is mostly reserved for the highlights. This is where the bodied oil, calcite and possibly egg may come into play. In a shallow way van Dyck almost sculpts with the paint, like a low-bas relief, and if we could see a cross-section of these two images they would reveal that. The shadows are mere warm-toned stains (note that the color in the eye sockets of the van der Geest is warmer in the original painting than in the image I show here) and the paint gets progressively thicker as it nears the highlight on the noses. Most certainly this aspect is not mechanically done, but the concept of using thin shadows and thick highlights was followed.

The halftones and many of the lights also have their stain-like qualities. Go back up in the post to where I show the full head of his self portrait and notice how thin the shadows, halftones and most of the non-highlight lights are. These areas make the most use of the previous layers of paint and were possibly put on as velature or more likely sfregazzi, methods Titian was also known to have used. Given van Dyck’s great interest in Titian,¹⁵ helped along by Rubens’ obsession, this is an educated guess. The thinness of these areas, especially in the self portrait, may be also evidence of scraping and then over-painting.¹⁶

Now, take a look at this close-up of van der Geest’s left eye.

The painting is sometimes called, The Man With the Watery Eyes and it should be clear why. Those thick highlights are lead and most probably a lead with large particles. Thick oil paint like what is seen here normally has smoothed edges because of the way the oil in the paint dries. Some researchers studying Rembrandt’s paintings have noted that the addition of a varnish (probably mastic) keeps those kinds of edges more crisp, as is seen above.¹⁷ But relative to the van Dyck I have my doubts and lean towards particle size along with the other characteristics of his lead. Sadly, we may never know for certain as getting a sample to test from these areas would be damaging to the painting.

Silver and Gold
Some have compared flemish painting to silver and gold. This theory, relative to flesh, posits that shadows are to be painted warm (gold), the halftones cool (silver), the lights warm (gold) and the highlights cool (silver). The warm glow (gold) in the shadow of the eye socket (seen far more clearly in person) was achieved by keeping white out of the shadow mixture entirely. Any lightness within is simply the ground showing through the thinner areas of the shadow paint.

Eastlake claimed that the principle difference between Flemish and Venetian art was that the Flemings did not add blue to their flesh. Rather, they relied on the fact that a translucent lighter note over a darker warm results in an optical cool (a scumble). In essence, these are the halftones. Something like this can be seen in the temporal line (right side in the image above) and also on the nose-side of his brow. Despite Eastlake’s theory, blue pigments have been found in paint samples taken from flesh areas of van Dyck’s paintings.

New Findings?
Researching Flemish painting methods is of great interest to many contemporary artists and the resulting theories often spark heated debates. In addition to the resources in the footnotes, below is a short list of other sources for further study. I make no claims, either way, as to their validity but each offers something to mull over.

• An english translation of the De Mayrne Manuscript. Also available here. The original book, written by a doctor who knew van Dyck, is one of the few accounts we have from the time. The translation is good, if expensive.
• Tad Spurgeon’s Living Craft. Tad is a representational artist and has gone to great lengths to test oils and additives to lead white, all with a mind to reproducing Flemish painting qualities.
• James Groves’ site has a lot of information on his views of Flemish technique, some based upon research and others educated guess. His ideas on how pigments were dispersed (pulverized, ground, mulled, etc.) are interesting.

References

¹ Although about the times of the Impressionists, Anthea Callen’s book, The Art of Impressionism: Painting Technique and the Making of Modernity, has a great chapter on artist’s suppliers.

² Imprimature is the plural form of imprimatura.

³ Wheelock, Arthur K, and Susan J. Barnes. Anthony Van Dyck. New York: Harry N. Abrams, Inc., 1990, pages 47-50. This entire book can be viewed and downloaded for free here.

⁴ Matthew Inness has recently done a solid post on Rose Madder, a lake pigment. See here.

⁵ See the article When Pink Was a Color by George O’Hanlon here.

⁶ Recently, George O’Hanlon of Natural Pigments has begun producing a lead white that may well be as close to what painters in pre-industrial ages had as we’ll ever get. If you are into mulling your own paint, this white is a must try. I use it and highly recommend it. See an article he wrote about the process, Stack Process Lead White here.

⁷ It does not really have to be mulled, per se, but can be made usable by some serious mixing with only a palette knife. Also, see note 6 on George O’Hanlon’s process above.

⁸ Calcite use in oil painting.

⁹ Brown discusses this, relative to Velazquez, in his book, Velazquez: The Technique of Genius.

¹⁰ Wetering, Ernst van de. Rembrandt: The Painter At Work, page 239. This page is also viewable on Google Books here.

¹¹ Norwegian painter Odd Nerdrum sentenced to jail and his response.

¹² London’s National Gallery Technical Bulletin Volume 20 has a number of chapters related to his materials and procedures. This volume and many of their other volumes are now available free online at their website.

¹³ See Marc Dalessio’s blog post on sun-thickening oil, here.

¹⁴ Wheelock, op cit. pages 45-46.

¹⁵ Based upon an inventory taken at the time of his death, Van Dyck owned a number of Titians. Van Dyck’s ‘Cabinet de Titien’: The Contents and Dispersal of His Collection, Jeremy Wood, The Burlington Magazine Vol. 132, No. 1051 (Oct., 1990), pp. 680-695.

¹⁶As I discuss in my post about Velazquez’s painting of Innocent X.

¹⁷ Whether varnish has been proven to have been used in a vehicle or medium by Rembrandt, Rubens, van Dyck, etc. is currently under debate. The latest consensus is that none or very little was used.