Back to the composition studies

It's been a while since I took the time to do some composition studies. Here's a Frazetta. Usual disclaimers about my having no idea if any of this is right or not, this is just what I'm taking away about a tiny bit of what's at work in the design. I find it helpful to do these because I could easily spend all day drawing people in action poses without making a picture that communicates any movement or force. There's obviously a lot more going on in here than I can put a finger on.

Mullins Study

I've wanted to do a study of this Craig Mullins piece for a while, since... y'know, it's Mullins. 

I figured out a lot of what I tend to do wrong with my pieces right from the beginning, just by studying the big shapes. Something you see consistently in all good painting, and which I still need to make myself pay attention to in my own work, is very strong silhouettes. Everything here is divided into light and shadow, but their overall shape is maintained. The light being cast on the floor is used to carve out the silhouettes of the people in the crowd, so that most of them read even with very little detail. Attention is first paid to the large abstract design. Notice how the big shape of the crowd blends seamlessly into the large vertical structure.

The shadows in this are also airy, with very dark accents only used in a few select spots. This allows them to carry much more color and texture information. Trying to figure out how this is done is one of the reasons I wanted to do the study, because in my own work I tend to either crush the shadows till they're nearly black, or go in the other direction and wash them out. That might be because I take the value too close to the light- in the Mullins piece, the shadows get very light in areas, but the degree of value separation from the lights is always consistent. 

Flow of light is important in the whole piece. The dark shapes in this get significantly lighter as they go back into atmosphere, but counterchange is used to keep them clear. Few of the shapes- the pillar on the left, for instance- have the same value from top to bottom though. If the pillar were the same value the whole way, it would look pasted on to the picture. It's darker than the arch behind it at the bottom, but as the arch lightens towards the top, the pillar does as well, without ever getting light enough to break the shape. You can see flow of light, and color shift, on the floor as well. It's more orange towards the right, near the windows. It cools and lightens as it goes back in space.

This is as far as I took the study. I didn't get into any detail, and didn't pay as close attention to edges, but still did pick up on how important they are to the piece. None of the big shapes are changed much from the earlier stage, aside from a little refining, but edges and texture are used so masterfully within those shapes that you can always tell what the material is. Definitely makes me want to hit the material studies, something I tend to neglect. Knowing those properties allows you to suggest how something looks without having to go in and render the crap out of it.

Another thing that the study reminded me of was the need to paint with a harder brush. It's not a problem I run into with oils, but in Photoshop I still find it so challenging to make the exact marks I want that I veer towards softer-edged brushes and pen pressure turned out. It makes it easier to inch towards what I want, but the flipside is that it becomes much harder to make a bold, clear shape with a sharp edge easily. I need to return to the hard round every now and then.

Edit: Notice how, even with the piece I'm copying right there, and full awareness of having this problem, I've still managed to make a lot of this just a little too dark. Note to self...

'Alien' Studies and some recent work

I am doing a really bad job of keeping this thing updated... I'll blame freelance. 

I've started doing comp studies from film stills- something I should have done years ago instead of only studying old paintings, especially since I want to work in that industry. These were done from the original 'Alien.' Such a gorgeous movie. It continues to blow my mind that these shots that seem so busy and full of mechanical details and complex lighting are actually so (deceptively) simple. Big shapes. Very clear silhouettes. You can tell what's going on from a mile away. 

This being an old movie, I also really appreciate how subtly a lot of the color is done, compared to the over-the-top complementary color usage these days. If anyone knows a good reason for the bright yellow-lit helmets in Prometheus, besides "It stands out against the blue backgrounds," I'm all ears. (Don't get me wrong though, that movie was still bloody gorgeous.)

Some recent work. These 2 are paintings for a client's sci-fi film pitch. Some of my more fun assignments. 

Cover illustration for Jared Sandman's new novel, 'Blackstone.' 

Finally, a portrait I did of the late Christopher Hitchens for the 'Notorious' group show at the Modern Eden Gallery in San Francisco. I haven't had a chance to work in oils much lately, so this was a real joy, and I finally got to do some experimenting again. I might post some of the sketches and process shots in a later post, once I get them scanned.

Frazetta Study

This was a painting I did back in January for a Crimson Daggers challenge.

Big step forward for when I made it, since there's a little bit of a sense of movement, but this was the extent of my planning it. I played with leading lines and all, and I liked the poses, but I'm still having trouble composing images with masses and shapes.

Frazetta's Conan the Destroyer. (Really, do I need to identify it :P?)

One of the things that's always gotten me about this piece is that it's so... chaotic. But, it's not confusing at all, the action, movement and shapes all read wonderfully. Most battle scenes, including the Pyles I've been trying to pick apart, resolve into big, simple shapes when I'm doing studies of them, but aside from the overall silhouette of Conan and the mass of bodies, I've had trouble figuring out how the masses are arranged.

Pushing it into value and raising the contrast helped a tiny bit. I can see how the big light shapes and dark masses spray out from the central figure of Conan, although I feel like a lot of the patterns I drew on it below are a little forced. I wouldn't count on them revealing anything useful. I also wasn't able to figure out anything about the pattern in the sky at first.

Opposing curved shapes?

This one was a little more revealing, I thought. There are 3 prominent axes in the painting, starting at the bottom right, moving up to Conan's raised weapon, and ending with the one buried in the guy's back. Highlighting those made me notice the similar shape echoed in the huge light break in the sky. Could be just seeing random patterns that aren't really there, although it suggests the idea that the lone figure's bringing the heavens crashing down on them. It also suggests a reason for that bright light shape on the left, which the axe and speared skull are silhouetted against.

This is all guesswork on my part, of course. This piece mainly works because Frazetta had an unrivaled imagination when it came to violent action, so everything about this piece- the arrangement, the poses, the thundering clouds- all ring with truth. I highly doubt that any of my cute lines and patterns were purposely installed.

More Pyle Studies

"The Battle of Bunker Hill."

Doing these comp studies at a small size forces you to ignore all the details - the minor separations of tone, the individual handing of each soldier, narrative details like the drummers, the bodies strewn about- and focus on the organization into large shapes. Most of the soldiers are organized into a big block. We can see the sheer number of people as it goes back in space, across the lit area in the midground. The large, simplified light and dark shapes are broken up by contrasting accents to add interest, something I'm starting to see more after reading a quote by Pyle's student, Harvey Dunn-

"Lick touches of light into the shadow field to give zip to the points of interest in the quiet tones and to tie the light field and the shadow filed together, similarly put a few flicks of dark in to the light field."

One of the most useful composition principles that we learn is to put the area of highest contrast your focal point. Note that it applies to much more than value- the highest value contrast we see here is in the background, with the dark smoke silhouetted against the white cloud. That's not where we first look, though. The focal point is on the soldiers in the foreground, where the tiny light, dark and red shapes form a pattern, contrasted against the economical shapes in the rest of the piece. Even more importantly, this is the area of the highest human interest. Unless you're doing something very wrong, a person will look at those areas first, not necessarily the point of highest value contrast.

As far as shape language, what I see most is the ordered, geometric shape of the soldiers marching forward, to meet the dark and looming, chaotic shape of the smoke from the battle. The contrast between order and chaos is the heart of the idea here. Note how perfectly in formation all of those soldiers are. I'd reckon that the vast majority of painters would try to ham-fist some "visual interest" into the scene by breaking up the poses in that formation. The rifles wouldn't all be straight; the light shapes formed by their legs and cuffs wouldn't all line up so perfectly. In Pyle's piece, he's at once staying true to the reality of how a military formation would have looked, and is emphasizing the idea of complete order and discipline, in contrast to the chaos of everything around them. We can almost hear the drum beat and the noise of battle, from the dark smoke cloud, and from the other bodies lying about- the ones nearby contorted, and the ones farther up the hill so strewn-about that their red uniforms make it look like a blood stain on the battlefield. That makes the main soldiers' bravery and discipline marching forward all the more incredible.

This one's another Pyle painting on Jekyll and Hyde. I couldn't pick out much from this one aside from, again, seeing how well and simply the light, dark and middltone shapes are organized. I did find one fun bit though.

The shape of the foreground guy's hand silhouetted against the light jumps out immediately, so it's obviously important. Look at how that shape is mirrored in the painting on the wall... and check out the sharp-angled shape of a beast's mouth about to bite it. Right there, that gives you an idea that Jekyll's about to turn into something quite dangerous. You don't need to know the story behind the painting to see it.

EDIT: I'll need to go back to studying that Jekyll/Hyde painting sometime. One of the risks of finding cool things like that painting on the wall is that it becomes easy to overemphasize the role they play. I thought about it a little and removed it in Photoshop to see what effect it would have. I don't think the piece suffered at all. Clearly, it's not as crucial to how we perceive Pyle's painting as I first thought. Lesson learned: always test your conclusions.

Comp Study of a Pyle

"When the World Seemed Young" by Howard Pyle

For this week's study I looked through a bunch of Pyle's paintings to find one where I could try to analyze how the shape language is supporting the central idea, without resorting to battle scenes, which tend to be more obvious (sharp angles, thrusting shapes, etc.). I can't begin to pick apart what makes most of Pyle's pictures work, but I tried to take a stab at this one. The main narrative of the man wooing the girl is already very clear, but like any good picture, that idea is being put across without needing to analyze what the characters are actually doing.

Before I started the study, I was already reading a few things. One is the common motif of linking the woman to nature, something you see in countless paintings. Her shape is linked to the flowering trees behind her, so the garden becomes a metaphor for her. The wall then becomes important- it's the barrier between us and the garden, and her by extension.

The block-in of the large shapes seems to support that idea. Rather than the silhouettes of the characters, the main thing we see is the wall, and the steps leading up to the arch-shaped entryway. The way is open, but we're not through it, just as the man is still outside the garden. His one foot on the step shows that he wants to go through and perhaps nearly is, which is the metaphor for his attempting to the woo the girl (since she is the garden).

After adding color and the smaller shapes, some more things start to jump out.

This I wasn't sure about, but I looked into it anyway, because the silhouetting of the man's dark hat of the light path behind it made me think it might be important. Possibly it's meant to show some affinity between the two characters, since the girl's own hat is also prominent with the dark shape within the lighter one. This affinity is expressed not only by the objects themselves, but by the similar shapes (circles in this case).

This was the very first thing that made me read into the idea of the wall and garden. The girl herself is the garden, but note that the pattern on her dress mimics the red, rectangular shape of the brick wall. It suggests that the man hasn't won her over yet- it is the wall keeping him out of the garden.

Another one I wasn't sure of, but I traced the path of the shapes formed by the pink, flowering trees to the read shape in the dress. Maybe it emphasizes her link to the background further?

That's all I have for this one. Of course, it's impossible for me to say whether any of it is accurate at this point. I can spend all day looking for metaphors and patterns, but unless I can use the knowledge to make strong pictures myself, I can't claim that it's the thing that makes the painting work in the first place. There's a lesson there- be very skeptical when someone shows you a diagram they drew over a painting and says "this is the secret to great painting!" If they can't use it to make a great painting, take it with a kilo of salt. Be scientific about the information you receive.

So, with that in mind, I'm learning here. If you have something to add, or think I messed up somewhere, please comment! We'll see if I can eventually use what I'm picking up from these studies to make my own paintings better.

Cornwell Studies

I'm still working on the sketch for the samurai vs. dragon painting from last week. The picture has a lot going on in it, but is basically broken into a dark, cool foreground and a sunlit middleground. I've seen tons of Dean Cornwell paintings masterfully pull of this arrangement, so I did some comp studies from him to try to figure out how to pull it off.

This painting has a lot going on in it. Several figures, bright primary colors, and (at first glance) a seemingly chaotic value pattern. The value study quickly shows how simply all the masses are arranged, though.

Howard Pyle and Harvey Dunn taught their students to compose using, at most, 3 values. Dark, Middletone (for the lights in the shadow, and the shadows in the light), and Light. You can see that at work here. The foreground figures are all in shadow, and clustered into one dark shape. The second cluster is in middletone, and the background is mostly light, with the other values used to create pattern.

The study in color shows how carefully Cornwell controlled the use of his 3 values. The light areas of the foreground figures are all done in middletone, never the light tone. To use that would break the shape and value pattern of the whole image. The lights on the middleground figures are done in the light tone. There is no dark tone used on that shape.

One thing that surprised me while doing these is how well the picture works in grayscale. I'd assumed that the color was much more important to helping it read.

Just drew on top of the image to see where things line up. The major shapes pointing to the main figure, who is dark and silhouetted against the light yellow drapery, are obvious. The diagonals fanning out in the second image was less expected, although I suspect that I'm just finding patterns there. I couldn't say whether it's critical to the image or not.

It's also important, while looking for patterns like this, to keep in mind that they're design aids. "Composition" includes many things, and design- the arrangement of shapes- is only one of them. That they are often mistakenly assumed to be the same thing, I feel, leads a lot of people to think that the design of a piece is all that's needed to make it work, thus you have so many magic bullet composition schemes being peddled. I'll rant about that another time.

This second painting is much simpler, and uses the same device of dropping the foreground in shadow. The foreground uses the dark and middletones. The two main characters are in full value range, however. They're the main actors in this, so it makes sense- the main figure in the previous painting is the focal point, but the story doesn't revolve around him. In this picture, we have full light against dark contrast on the main figures, and they're framed and silhouetted by the dark background and white pillars, focusing all attention on them.

I went back to my own sketch after doing these, and focused only on value this time. This is one of those things that I've been taught for years, but for some reason keep trying to skip- compose in value. If it works in value, it will work in color. I tried to arrange the values more carefully, and used a full range on the main rider. Hopefully it reads better now.

Composition Studies - Battle Scenes

Until now I've mostly been using this blog to dump new work as I make it, so it's tended to get neglected. I'm going to try to make a point of doing regular updates from now on, and will focus on my studies of composition and picture-making. I paint a lot of pictures with multiple figures and action, and want to try to pick apart how others have approached that- using shape language, pose, the overall graphic design of the picture, and generally the spirit of the thing.

The painting that I've been doing studies for is a rather complicated battle scene, of a bunch of mounted samurai being attacked by a dragon. The 2 sketches I ended up choosing between are-

Both of the 2 have merits. In particular, I like the more unified shapes in the second, but I'll be going with the first one. There's a little more immediacy to what's happening, as opposed to the second, where the danger is already upon them and passing. The first also has more of the look of screeching to a halt, like they've just been set upon, which is what I originally wanted. The focal point is clearer and the camera angle's more interesting.

Still, there are problems. The shape language needs to be better. That main horse in particular looks too calm, and I'm not sure how to lend more urgency to the movement. I decided to do some comp studies of older battle scenes and found even more that I don't like in my picture- the large, split, symmetrical shape of the two horses in the foreground, for one. There's little unity, it's too haphazard.

This is one of my favorite battle scenes by Howard Pyle. He was a master at capturing the spirit of a scene, and his sense of design was incredible. Even though there's a lot going on, there are really 2 large masses in the picture- that of the horses and the un-mounted soldiers, separated by the barrier of that rifle.

The large mass of the attacking horsemen drives a wedge into the foreground. Simple but very effective shape language.

Note the implied lines in the attacking shapes. The arms and swords silhouetted against the sky in particular give a real sense of those guys bearing down on that thin, opposing diagonal of the rifle, which is all that stands between them and the infantry. This is the kind of thing I want to be able to do in my own pictures- the shapes, the lines, the brushstrokes, they all communicate the idea before you even have a chance to analyze everything.

This was actually pointed out to me months ago by another artist. Not so much about the shape language, just more showing how Pyle controlled the focal area. Tons of implied lines point to the central horseman, who is also silhouetted against the light sky.

Same thing, in case you like spirals.

Here's another, even more complicated Pyle battle scene. This one is a mural that he did. Countless figures in all manner of poses, but...

...they're all grouped into large shapes, all of which are designed to convey the forward thrust of these charging soldiers.

Implied vectors within the shapes further carry the thrust. I don't even need to draw the lines over the light rifles that do the same, those are obvious. Another thing to take away from this picture- if some art student ever looks at your picture as says, without really looking, "I don't like how this line leads my eye out of the picture," please hit him. You can do the same if someone starts talking about the rule of the thirds.

Here's a Frederick Remington piece. This one's harder for me to figure out, but I wanted to get what I could about the shape language of the horses. The silhouettes of almost all of them are top-heavy triangles. This gives the impression of being off-balance, which conveys motion. Even when you walk slowly, from one step to the next, there's that brief moment when you're off balance.

Same thing, only applied to the larger masses and cast shadows.

Implied vectors, etc.

So, there's a bit that I picked up from these that I need to learn to apply to my own work, especially the beautiful simplicity, even in a hugely complicated scene. The figures and horses in these paintings are arranged into big shapes that convey the idea all on their own- the details seem equally well designed, placed as accents to increase the thrust, or emphasize something. I'll need to figure some stuff out in the sketch I'm working on, and will try to analyze what is and isn't working.