Dancing with Degas
Updated: Feb 28, 2019
How I approached making a copy of The Blue Dancers by Edgar Degas, what I learned, and what I would do differently.
When I saw the call for entry for a Fakes and Forgeries show at Spring Bull Gallery in Newport, I thought, that might be fun. I quickly decided to do a painting by Degas, whose work I have loved since long before I took up painting. Looking online at all his work, I gravitated to the dancers--after all, I was a ballet dancer myself--and quickly was captivated by the composition and color in The Blue Dancers. My decision was, to my ultimate regret, purely visceral. I gave not one thought to what would be involved in actually painting it. Why couldn't I have picked L'Etoile? Only one figure! By the time that occurred to me, I was, to paraphrase Macbeth, in blood so far that to turn back would be to go twice over.
The original of The Blue Dancers is in the Pushkin Museum. I visited their excellent website and was able to zoom in on the painting on the wall to see the the colors as true as possible (posters and prints are notoriously unreliable). I downloaded a copy to use as my reference.
Before tackling it, I decided to try a freehand sketch, and a color study, shown below. I did them very quickly, but it didn't take long to discover two things: this was a very complicated composition, with body parts and tutus colliding to carve out shapes; and many of the values were very similar, almost running together. I realized then that it was going to be tough to capture the combination of unification and distinctiveness in the original, and that I might actually have to plan this out in some detail to get it right. Not my usual quick-and-dirty style. But I can be disciplined when needed, and this was the time.
Before starting in earnest, I had to make some choices: Size? Paper? The original is 25"x25," but not quite being able to face that, I settled on 16"x16." I chose Pastel Premier Italian Clay for my support; it was really more of an instinct than a decision. I then decided, also intuitively, not to do a wet underpainting, which I almost always do. I think part of me wondered whether Degas would have done a wet underpainting. Maybe I could have pinned that down through research, but I didn't.
I had printed out my photo of the painting for my reference, and had to scale it. I had never done this before, so I researched online how to scale a painting, and made a grid, first using my grid app on the photo and printing that out, then enlarging it onto the paper so I could do the drawing. Here is the grid on the printout and the drawing I did from an enlarged grid on my paper.
I had to do the drawing twice to get it right--my enlarged grid was a little off--and it took me several hours of the kind of detailed attention I don't love. When I finished it, I thought the hard part was over. Ha. I began laying in the color, putting in big patches of the dominant blues, greens, and orange to mark their places. Knowing that the value transitions were going to be difficult, I had printed out a black and white version, as well as a clean, ungridded version, to use as I worked. I thought I was prepared.
As I began to lay down the color, I noticed a few things. First, which is rather fun, I discovered that most of the colors I was picking up, and that seemed to be closest to my reference, were Senneliers and Giraults. Was that a coincidence, or was it because those are what might have been available to Degas, even those same colors? I preferred to think the latter, as they were perfect in ways that others were not. I did use some Terry Ludwigs for the intense blues and some Unison yellow-greens in the background along with Senneliers and Giraults.
Second, as I worked, I started to see that there were many layers of color, and that I had perhaps started with the wrong ones. I began dabbing color in on top of things that probably
should have been underneath. As you can see from the progress shots below, I had begun blocking in the background and the dresses, deciding to leave the skin of the figures, except for marking out some patches of color and highlights. As I progressed, I saw so many color nuances in the skin tone, which I had to try to retroactively incorporate; by the time I was between shots 5 and 6, I was working on the skin with abandon, very aggressively, just moving closer to what felt right rather than trying too hard anymore to figure it out. By then I think I had, but it had taken me too long, so now I just had to make it right. The last thing I did was finish the background, toning it down so that it was clearly the background without divorcing it from the figures, as this painting truly has an intensely holistic feel.
Because of this unity of composition, one of hardest things was to actually see where one figure left off and another began; until you actually start trying to paint this, it is hard to appreciate how elusive this was. I found I had to tackle the figures one by one to get them straight. For example, I completed the partial bottom figure first (progress shot 2), working to make sure that her shoulder ruffles and the tutu of the large dancer on the left could be distinguished. I then moved on to the dancer with both hands to her shoulders. It seems obvious now, and will be as I am pointing it out explicitly, but it took me a bit to figure out that it is the shadow of that dancer's elbow that is carving out the other dancer's waist. And so on. The ruffles and tutus shape the other figures, and determine what is in front and what behind. The dancers are crushed up against one another, bits of the tutu of one figure who is further behind coming in front of the back of another dancer who is further front. If that sounds a little confusing, it's because it is. It required very close looking to see how these figures, all jumbled together as if in a tiny area behind the scenery, were distinct.
Ultimately, I came up against a deadline, and had to stop and submit it. All things considered, I was happy with this first effort at doing a master copy. I found a really nice frame for it that brought out the orange in the painting and suited the period.
What I Learned, and What I Would Do Differently
After doing this painting, I can't help but believe that Degas was a terrifically free-spirited, supremely confident and fearless painter. There is nothing tight or cautious about this painting, and it feels very much as if it was joyously worked out as he went along from a general concept. Whether this is true or not, I found myself worrying less about every mark the further along I got, and feeling emboldened by his work to be more bold and carefree about the painting myself.
But since Degas was a genius and I am not, here are some things I wish I had done and that I would do differently were I to try this same painting (or perhaps other Degas work) again.
1. I would have spent much more time studying this painting than I did. I thought I had, but really, I think I might have saved myself a lot of correcting and back-and-forth had I just stared at that painting longer, and really noted how all the pieces fit together: the painting is a puzzle. There is nothing formulaic or even that familiar about the way it's put together. This may be the biggest lesson I take away from this.
2. I would have covered the entire paper with orange in a sort of dry underpainting, and drawn the angled line of the scenery, before doing anything else--that is, I would have done the sketch over the orange. Here is the palette (a few of these could have been let go). That little piece of orange is the most important color in the entire painting, and it lies beneath everything else. I had to work it in and on top of things in places where I had not first put it down.
3. I would have pulled blue through the painting from top to bottom, early on. I worked at pulling the blue lines through in little sections--on the scenery, on the skin, and so on--and mostly toward the end of my painting. I ultimately realized that most of these lines are continuous and integral, much like the orange. I also would have laid in a little yellow early on; dragging the blue through it creates some of the greens.
4. Consistent with 2 and 3 above, I would specifically experiment with the skin tones, as opposed to doing them trial and error as I went. One of the figures is a little overworked as a result. Probably I would do little color studies of arms, for example, experimenting with a mix of warm and cool colors.
I will be interested to see when I try another Degas whether this experience has put me high enough up the learning curve that it will take me less time than this one did. I suspect so. Degas was a sly magician, but I think I've figured out a few of his tricks. And I plan to use them.
P.S. Good news! Pastels are incredibly hardy! I left my barn door open by mistake for two and a half days during our little polar-vortex storm of ice and wind while painting this, maybe around progress shot 4. When the storm was over and I went out to paint again I had a moment of horror: the painting exposed to the elements, protective covers blown off, references scattered to the winds, water all over the floor, and just plain damp air. With some trepidation, I put pastel to paper and . . . it was fine. Phew.