a dream of a shadow

At Wimbledon in 2023, I had the chance to watch Jannik play a Frenchman, Quentin Halys, from the first row of Court 3 in the third round. Jannik was still ranked outside the Top 10 at that point in the tournament, but would make it inside the Top 10 following a semifinal result.

While watching him play that day, I had a feeling it would be the last time I would see him on such a small court. Indeed, from that point on, he would play only on Centre Court or Court No. 1, and I would never get to photograph him from that close--at least, not during a match! One of the photos from that day became the first thing that popped into my head when I thought to try pastels, one year on when I had started to do art. At that point I'd only done graphite pencil sketches, and didn't know if I could handle color.

I ended up going for other projects--a good thing, because while I might've done OK with this two years ago, I lacked the confidence in drawing and understanding of the pastel medium to achieve the effects that would have suited this image best. While you'll often hear people refer to "pastel paintings," or "painting with pastel," pastel is, like all dry media, properly a drawing medium. Unlike true painting with wet media and a brush, the act of applying color directly from a stick of dry pigment encourages a form of mark-making that is distinctive to the medium--if you will let it be.

In my first big pastel, I tried blending everything out for a nearly photorealistic look (mainly because I didn't know which details I could do away with, and the more I looked the more detail I saw). While that's all fine and well--academic realism is a legitimate style--it doesn't really let the medium of pastel sing. As I've learned since, pastel really shines when you lean into the mark-making, when instead of broad patches of flat or perfectly blended color you go for optical layering, squiggles, crosshatching--drawing, in other words.

That's what I've been trying to do since, and what I've attempted to do here: produce something that looks like a drawing as opposed to a painting (or a photograph).

Some process pics and comments:

(1) Face--I've changed up a bit how I tackle things: instead of beginning with it and worrying about the background later, I laid in some of the background around the face to help judge the values of the latter better (very important, because against a dark background even fair skin tones are actually darker than you think).

I consistently have this problem of making people's faces not dark enough--and even with the background put in first, I still ended up doing it this time! I had to edit his face later, after the whole thing was done. I have two theories as to what is causing this issue. The first is to do with the fact that I work from a reference image projected from a screen. These days devices often automatically adjust the screen's brightness to match ambient light levels; I generally complete my faces in one day, working late into the afternoon, and this means that the time I'm putting finishing touches on a face--highlights and so forth--is right around the time daylight is declining and the screen is brighter than it should be, brighter certainly than it was when I started working in the morning.

The second, not mutually exclusive possibility is to do with ... I guess this is a phenomenon related to atmospheric perspective? If you work with traditional media, you'll probably know that colors and values look different when you are really close to a piece, i.e. arm's length or working distance away, versus when you are at a typical viewing distance of two or three meters away. I was getting the colors and values right from close up; but from further away dark passages become less dark unless you really go bold with them. This is precisely why art instructors will bang on about the importance of stepping back from your work frequently. Something I'm still not consistently good about; but I digress ...

(2) Hands. Pretty straightforward here.

(3) Clothes. Like hands, they don't generally trouble me, but strongly backlit white fabric is always a special challenge. Before tackling the shirt I spent a good 20 minutes squinting at the reference photo wondering, What the !@#$ is that base color?? It looks grayish-blue, but the closest match was actually a violet. Here in the left-hand images you can see the initial (blocking) layer of the shirt and shorts being composed of a variety of these purples bordering on blue (ultramarine, to be specific).

After two years of pastel-collecting, I have a lot of candy-colored tools at my disposal. The flat pebble-shaped pastels, from Diane Townsend, are nice for the initial blocking layer: they contain pumice, which provides grit for subsequent layers to adhere to. The square ones from Terry Ludwig have become my workhorses, but despite coming in 789 colors, they did not have the deep, saturated greens and teals I needed for the background.

Enter my Unisons, and a random stick of Sennelier No. 957 Imperial Green that I'd bought online two years ago, using a digital color swatch, hoping it would be a match for the background of this exact piece. (It was.)

So. Two years late, there he is, memorialized as he was on a sunny English summer day, twenty-two years old and on the cusp of greatness.

I say, better late than never!

Still trying to get the hang of this oil painting stuff. Some say watercolor is harder, but having done both, idk, I find oil to be next-level. Different paints drying at different rates. Some colors drying glossy, others matte (and does one "oil out" or not if that happens?). Fat over lean. Mediums, solvents, what primer to use for your surface. Color mixing: I got away in watercolor with layering individual colors to get the hues and values I wanted, but in oil you can't do this; you have to mix everything bang on from the start.

The hardest part, imo, is that you can't hide your brushwork. OK, well, you can: you can blend everything out or paint with the tiniest brush known to man and achieve a hyperrealistic look, but I was going for a kind of brushy, John Singer Sargent-like look, which turns out to involve laying each brushstroke down over paint that is still wet (working "wet-in-wet" as painters call it) and not correcting the stroke afterwards.

Detail of facial area showing brushwork.

At most, you can soften some edges (going over the mark gently with a clean brush), and knowing which edges to do that for and which to leave alone is Its Own Thing. (Edge control is one of the biggest things separating really good from not-so-good realist work: the latter tends to leave everything sharp or to over-soften everything.) Trying to do more than that, however, will unavoidably make things worse, so if you get the stroke wrong, you have to scrape off what you did and try again. And when you're working into paint that's still wet, you have to know exactly how much paint to put on your brush, and what angle to hold your brush at when applying the next layer: otherwise it just won't take.

Now I understand why even a genius like Sargent spent four years training under the best portrait painter in Paris to master this style.

****

Process pics + yapping:

STAGE ONE. The outlines of the head were roughly sketched in charcoal, then an abstract background was laid in using terre verte (a green natural earth pigment, the light green color) and viridian mixed with a touch of transparent red oxide (the dark green color), applied with a large flat brush. The background was done first to aid in accurate judging of the tones and values of the face, which is difficult to near impossible on a white canvas.

Some artists will paint a light wash of color over the entire canvas to begin, also in order to make the judging of colors/values easier. But since I was aiming to do this work alla prima, painting fresh layers on top of still-wet paint, I didn't want to contend with layering warm flesh tones over green. (An old Italian technique called verdaccio exists where an artist paints a portrait entirely in shades of green, establishing the correct values or contrasts between light and dark, before painting over the image in its final colors. This is supposed to lend the flesh tones greater vibrancy, especially if bits of the underpainting are left to show through. However, these artists would always have allowed the green underpainting to dry thoroughly before proceeding.)

STAGE TWO. A base of flesh pink (venetian red tinted with lead white) was laid over part of the face. The eye and nose area were then defined, and the tonality of other areas of skin varied with pinks, reds, and yellows.

A lot of portraitists would have done a first pass over all the major features, instead of completely finishing one area before moving on to the next. I have always found it natural to do the latter and to start with the eye region specifically: it serves as an anchor for the rest of the face, and if done properly, there will instantly be life, a human presence, in the canvas. That's a shot of confidence and a great source of motivation to continue.

STAGE THREE. When painting a subject with facial hair, as a rule, I first layer other colors underneath the area bordering on or to be covered by hair. Where Novak's hair is showing under his cap, too, I painted flesh tones beyond where the hairline would have begun. If I were to try to guess exactly where the skin ended and the hair began and painted up to there, I would get a hard boundary or even a tiny gap of white between skin and hair, which looks bad. Even in thick facial hair you can also have slivers of skin (or some other color) showing through: so layering is a must.

Note the red on the jawline: in the source photo, there was a strong reflection of red/orange light from the shirt Novak was wearing. After the hair is layered in, the secondary lighting will look natural.

STAGE FOUR. The facial hair was laid in over the mouth, chin and jawline. The hair in front of the ear and at the back of the head was painted, and the form of the neck at the nape further modeled with shading.

STAGE FIVE. The cap and shirt were painted. At this point, with all areas of the canvas worked to more or less a finished state (only the ear remained to be done), some flaws in the face and neck became noticeable. Some were small details, but the biggest issue was that the face and neck were too pale compared to the source photo, especially after the very intense orange/red of the shirt was incorporated.

Fortunately, I determined that repainting the whole face/neck would not be necessary. I did repaint the neck, because there were a couple of spots of wrong-colored facial hair that needed to be eliminated, but on the face I darkened the nose, eye socket, and cheek area, and that was enough to get the whole thing looking half a notch darker.

This took 14 hours over two days; I was hoping to finish it in one day (to paint alla prima is literally to "do it in one go"), but some areas, the cap and the shirt particularly, really threw me for a loop. (They look drop dead simple, but they're not! Or rather, painting something so simple-looking, and having the result look simple-looking, is anything but simple ... at least for an oil noob like me.)

But it's OK, I'm improving! My first head took four days, my second one (the one before this one) three ... so I guess my next one ought to be one day's work??

Shirt (creases upon creases)

lobster claws hands

When sharing our work, we artists tend to show just the final product, maybe snapshots of a few steps along the way. We tend not to show how things went wrong in the middle, if they did. (Trust me: they go wrong pretty often, at least for me.)

On this one, a lot of things went wrong, and the final result, while I like it now, was far enough away from what I had envisioned that I couldn't look at this piece for months after finishing it. For one: the skin tones were way too red. As in, my man Novak was a lobster being sous-vided on court. This was caused by a combination of things--light reflecting from the clay surface onto him; and my iPhone employing a setting called True Tone which supposedly "automatically adapt[s the] iPhone display based on ambient lighting conditions to make colors appear consistent in different environments." There was nothing to be done about the first: I have to draw what I see in the photo. But the second wreaked havoc. The phone boosted the saturation of the photo, probably while I was working late in the afternoon in declining light. Without it, I would not have gotten the colors this wrong. As it was, I was too focused drawing to step back and realize that his legs were, well, lobster-red. (Needless to say, after realizing my mistake, I went and deactivated True Tone for good.)

Fortunately, I was able to correct the legs so that they didn't quite look as sous-vided. But the clay surface was another matter. While on the rest of the drawing I used my workhorse Unisons, for the clay I was trying out a few sticks of Henri Roché pastels purchased from Paris during my Roland-Garros trip. The House of Roché was founded in 1720; they've been in business since, and they're currently a two-woman team who make, pack, and ship upwards of 2000 different colors of pastels sticks all by themselves. Degas used their pastels back in the day: 'nuff said. They also happen to be, by some margin, the most expensive ones on the market. €20 a stick, folks. By contrast, other professional grade pastels are around $3-$8 a stick.

Naturally, being curious to know what the fuss was all about, I went and bought some sticks in RG colors. I thought to do this drawing specifically to see how Rochés differed from what I had: I would use the Rochés just for the court surface. Well, I learned the difference--to the detriment of the drawing. I thought my Unisons (a top quality pastel, used by many professional pastellists, handmade in England) were vibrant, but next to the Rochés they looked dull and chalky. The Rochés glowed with their own light.

Problem is, the clay surface was supposed to be a dull, pale pink/orange. That is how clay looks under overcast daylight, and, to own the truth, that is how it looked on my phone even with True Tone turned on. Rendered with Rochés, my court was now on fire. And this being pastel not oil painting, there was no way I could cover it up in another layer, scrape down and start over. Not over an area this large.

I won't bore you any further with my travails getting the court texture right. (It looked like a Martian beach volleyball court after my first pass, and I spent a day editing it to look like it was flat even though churned up by footfalls.) Also, look closely at the racquet, and you'll see that the string bed is all wonky: this is what happened when I was trying to get just one more thing done after having already worked on the drawing something like seven hours that day, and decided to eyeball the string pattern instead of using a straight edge and measuring.

I got a lot wrong on this one.

A more disciplined artist would perhaps have thrown the work out and started the whole thing over--or gone for version 2.0 after having finished the first attempt. Certainly, Degas was known to do umpteen iterations of a given image until he found exactly the colors, contours, poses etc he was happy with. I do not have the discipline or the patience for this. I justify my laziness by thinking of the finished, flawed work as a record of my shortcomings as an artist--and as a record of growth, when looked back upon. Once I decide I'm done with something, I never go back to change it, even if there are some pretty bad-looking mistakes in there--because I don't want to think that I was better back then than I actually was.

So, yeah. May I present Novak tapping the clay off his shoe while looking a little sous-vided on a sea of lava, folks. Happy birthday, old man.