Sunday, January 11, 2015

A Little More on the Color Wheel

It seems there's a bit of confusion on the color wheel thing in the previous post. I had a couple of questions asking me about the relationship between chroma and value. I said that in the color wheel I presented, in each slice of the pie, the chroma changes (becomes grayer) but the value remains the same as we travel toward the center.

Sometimes our eyes are tricked into seeing lower chroma as lower (or higher) value, when they're actually the same. I really can't tell you what the scientific reason behind that is. I have noticed a tendency to confuse lighter (value) with brighter (chroma). But confusing the terminology is one thing, to actually perceive a color as lighter or darker than it is something we all experience to one degree or another. 

Squinting is always a good way to simplify what you're seeing - by limiting the amount of light that the eye receives, you limit the amount of information as well, and we see only the simplified picture. If you were to squint at my color wheel, I think you can see that the values do not change within any slice of the pie.

…that is, within reason. I don't claim my pie is perfect, but the aberrations are (should be) within a few percentage points at most.

Below is a good illustration - I converted the color wheel to grayscale in Photoshop to eliminate hue information. All we see now is value.  See what I mean?

I did the same with a couple of color wheels that I found online. Here's one in which the colors move toward black as it moves toward the center.

And another one in which the color moves toward white as it moves toward the center. It's got a hole in the middle, but you can see what I mean. 

In all three examples, all I did was convert the color file to grayscale. There's a huge difference in what information is presented, isn't there?  I really don't find the second and third color wheels helpful. I mean the second one shows what happens to a pure hue if you mix it with black, and the third one shows what happens when you mix it with white. That's not really useful information for a painter. I don't want to know what happens to red if I mix it with black or white. What I want is to be able to show how chroma affects color. I want to show that value and chroma are two different functions.
That changing the chroma of a color doesn't change its value. (Changing the value of a color does affect its chroma. Chew on that one! hint-the second and the third color wheel does show us this fact)

Anyway, I hope these examples clear up the confusion about chroma and value. Remember to squint!

Friday, January 9, 2015

Color Systems - The Center of the Pie

Another Working Day, 9 x 12 inches, oil on linen

Happy New Year! I hope you had a great Holiday~ We stayed home this year and spent a nice, slow-paced Christmas and New Years with good friends. Very relaxing, which was good, but so relaxing that I did not work at all for two weeks. (gasp!)

But it is now the new year, and I'm starting to get my brushes in gear. I actually wanted to do this post sooner, but I couldn't. You see, I needed a color wheel as a visual aid, and I did a search on Google, but to my surprise did not find the kind I was looking for. I found a surprising variety of color wheels, but many were just plain useless, and some were very confusing. It's no wonder students get confused when learning about basic color theory.

Anyway, since I couldn't find the one I needed, I made one. This is why this post is late! But anyway, my color wheel has 12 slices of the pie. It's your typical Y-O-R-V-B-G wheel with the in-betweeners like Yellow-Orange and Blue-Violet represented as slices as well. Now that I look at it, it still needs refining, but I think you get the idea.

The important thing about this color wheel...and what I could not find on my search, is this; the hues lose saturation as it moves toward the center, but their values remain the same.  This is important because when we talk about manipulating color, we need to be able to isolate the specific aspect of the color we are messing with; hue, value, and saturation. If say, the red of the apple is too saturated, we need to be able to change its saturation without changing its value. 

Most color wheels show the colors moving toward white or black or mid gray as it moves toward the center.  Which doesn't help us understand saturation because with these wheels, value and saturation are not treated as separate things. The grayest part of the yellow slice of the pie should not be the same value as the grayest part of the red slice of the pie, see. 

OK, so with the single-color themed approach that I talked about in the last post, we are essentially working with a narrow slice of the pie. May be a few slices at most. The idea is to limit the range of available color so that everything within that range is closely related to one another, thus ensuring harmony. 

We can apply this limited range idea in a different way. Rather than using a narrow slice of the pie, we can limit the range by using the colors in the middle of the color wheel. The pure hues are unrelated to one another unless they're analogous, but as we move toward the center, they have more and more in common, so the colors become more harmonious. 

Granted, we give up the brightness of pure hues, but this is a strategy for keeping colors from going out of control while still having a full spectrum at our disposal. If the problem is over-saturated colors that lack harmony, then desaturating to gain harmony seems like a logical solution, doesn't it?

If we go farther in, the colors become grayer and grayer, but look! the entire spectrum is still represented. (unlike the narrow slice of pie approach)

I should explain that this color wheel doesn't address value changes within a hue. To really express the full range of color, (hue value, and saturation), we need a three-dimensional model a la Munsell Color Tree. I'm not suggesting that we paint with just the values we see in my color chart; I'm just showing saturation changes without bringing in value changes. I hope that makes sense.

In practice, I use this approach most often when painting outdoors, responding to natural light and the colors I see in front of me. I don't particularly try to be literal with my colors, but I do tend to work with a full spectrum (albeit limited in saturation). 

Coastal Farm, oil on linen

I've noticed that most of the time, when a painter is responding to the colors they see - as opposed to values and shapes - in picking out what to paint, he tends to identify the hue first, go to its purest form out of the tube, then gray it down by adding complements or other muted colors (like black or brown).

If you start your mixing from the pure and move toward muted, you are always comparing your mixture against its more saturated version, which tricks your eye into seeing your mixture as more muted than it actually is.  When you become experienced at color mixing, there's no problem, but often in the beginning, the student struggles to control harmony because he's not getting the colors muted enough (closer to the center of the pie) for them to be harmonious. Isolated colors are just that; isolated colors. Colors need something in common for them to be "related", and if you stay too far to the edge of the pie (the more saturated area), it's not easy to relate the colors to one another.

Autumn Around the Corner, 11 x 14 inches, oil on linen

What if you approached the target color from the other direction? Rather than identifying the hue and graying it down to reach your target color, why not start at neutral gray, and add to it the hue until you have the desired saturation?

This actually works very well. In fact, I used to paint that way all the time, when I felt my colors were out of control.  This way, I can saturate the mixture a little at a time, and it's always going to look more saturated than it actually is, because I'm comparing it to neutral gray.

Because you are starting every mixture from gray, harmony is virtually guaranteed. (Every color has the same gray-the colors used to mix that gray-in them)

There are a couple of problems that arises. One is  that the painting sometimes look too gray and blah. The solution is simply to allow more saturation in strategic areas. These can be accents, or just push the saturation in one or two of the colors, not all of them.

Wanderlust, 11 x 14 inches, oil on linen

Another problem is controlling value. If you simply start at a neutral gray at mid value (or some other arbitrary point on the value scale) and start adding a bright tube green to try and arrive at a target color that is a muted green, chances are the value is going to be wrong.  When I was trying to figure out this color stuff, I came up with a solution that worked pretty well. 

Since I tried always to compose a picture in three or four major value groups, why not have three or four grays from which to mix colors? That way I can get the value in the ball park very quickly, and once I've identified the hue direction, I only had to worry about the saturation. From there, fine tuning the value was not that difficult. 

While I was at it, I thought, hey, since I'm trying to paint a rich full spectrum colored painting by mixing all colors from grays, why not mix the initial grays from pure colors (rather than mixing them from black and white)?  And so I mixed all my starter grays from cad yellow, cad red, ultramarine and white. I just varied the amount of white I mixed into the puddles to achieve my three or four starter grays. 

[Eventually I only needed two gray puddles to begin with; one for the lit areas, and one for the shadow areas. The value variations within the two categories were mixed as I went along, and I found that I didn't need three or four gray puddles any longer since separating light and shadow values were more important than separating values within the categories. (Note: this doesn't work as well in diffused light situations) ]

The primaries you use to get these grays have an impact on the colors you mix from them. the basic cads + ultramarine is a good one, but it's not the only combination. In fact, you can use all kinds of different mixes of pure colors, as long as you end up with neutral grays. I encourage you to find what works for you - everyone has a color bias, and preference, so a strict recipe is not recommended. Just try different yellows, reds, and blues. You can even use secondaries to mix your starter grays. How about Viridian, Cad Orange, and Cobalt Violet?

Garnet Lake Morning, 12 x 9 inches, oil on linen

This painting above, and all the ones before it on this post are painted in this system. You can easily identify one or two relatively intense colors in each of the painting, (though they're still not pure hues) and if you hide them with something, the rest of the image is pretty muted.  Another way of looking at that is, you only need one or two bright colors to make a painting not look drab. 

There are a couple of other factors that contribute to a sense of colorfulness in these images. one is a strong value structure; value contrasts increase impact, which gives the impression that the colors are strong. (whatever that means) The other thing is crisp edges. This is important when painting with subtle colors. If you soften or lose an edge between two areas with subtle colors, you diminish the colors' identities significantly. You can easily lose them completely. That's not to say you can't have soft edges, but you have to be aware what a soft edge does to adjacent color notes, and use your hard and soft edges strategically. If you always put down the paint in such a way that you're not mushing each stroke, you'll avoid this problem of losing subtle colors on canvas.

Afternoon on the Farm,  9 x 12 inches, oil on linen

This last one uses a little bit wider gamut than the others, but still done the same way. the green is the only thing that's saturated. The other clearly identifiable hues are not as intense, although they may not look very gray. 

In the very light end of the value range, as in the sky, you can get very clean colors without going out of harmony because everything is so close to white. If you think of white as essentially a very light neutral gray, then you can see that even if you try to paint with intense colors, you're not that far from the center of the color wheel. The white, in this case, is the common denominator. 

I encourage you to try this center-of-the pie system. Even if you're not a gray painter, it'll may give you another tool to control saturation in your colors.

Tuesday, December 16, 2014

Color Systems: Single-Color Structure

Working with a single color theme naturally puts the painting in a tonalist category, I'd say almost by definition because you have to do most of your composing with shifting values. 

It's simpler than using a lot of colors, but the flip side of that is you cannot hide behind splashy colors. You really have to have very good value control to pull this off. 

The idea is not complicated. You basically pick one hue, and paint more or less monochromatically, and strategically adding a little bit of color shifts to make the painting seem less monochromatic.

That's about as close to a formula as I can get! (there's that F word~)  But let me talk about each painting and see if I can pull out some of the things I did which deviated from the strictly monochrome structure, because that's where we can add a little bit of complexity and interest into an otherwise very limited color space. 

The top image of the delta, obviously, has a violet themed structure. It never gets very saturated, which help to maintain a quiet, somber mood. If you look closely, the very distant mountains(?) and the sky are slightly different in temperature - the sky has a tiny bit of red in it, which differentiates itself from the cooler hills. This is subtle, but not a tricky color shift because violet is made from blue and red. Adding a little bit of red warms up the color, and adding blue cools it, and we don't have to worry about the new mixture being out of harmony. 

The darkest land mass in the front has some Transparent Oxide Red in it. Which, if you think about it, is still red. The TOR is used, then, to control the saturation of the violet so that we don't have a screaming purple. Another way to control saturation is an addition of a low chroma blue, instead of (or together with) Ultramarine or another intense blue. For this I probably used Ivory Black as a low chroma blue.

One more thing about controlling saturation. (Because you know, I'm a little shy about using loud colors) Ultramarine is already a violet-leaning blue. Alizarin is a violet-leaning red. They are both very intense colors, so if you mix them, you get a very intense violet. Great, if that's what you're looking for. But if you want a little less intensity, you can try Ultramarine plus an orange-leaning red. The orange being complementary to blue, the resulting violet is much more muted than if you mixed Ultramarine and Alizarin. Because nothing I paint requires screaming violets, I like to use the mixture that's already a little muted even before I gray it down further with Black or TOR.

 Green. The color shifts toward yellow a bit as the values get lighter. It doesn't have to, but that's what I chose to do to deviate from a strictly green painting. I also used TOR in the underpainting and the darks of the foliage interior. Red is complementary to green, so it helps to gray down the green if you mix them. If you juxtapose them without mushing green and red, you start to get simultaneous contrast, a little bit of which helps to break up the monotony. 

Mixing Ultramarine and yellow ochre, the resulting green can't get too saturated even if you want it to, so that's a good way of limiting your intensity. You can always add Cads later if you need to punch up an area.

I tried to get some color variation in the ground plane, mixing the same three colors plus white (Ultramarine, TOR, and Yellow Ochre) in different amounts to get different, yet very closely related notes. 

Peachy! Or red orange. The single color theme sometimes isn't strictly single colored. Sometimes it's better described as a "narrow slice of pie". The pie refers to the familiar color wheel, and narrower the slice, the more specific the hue. If you cut a fat slice of pie, you're basically using analogous colors - neighboring hues such as red and orange, orange and yellow, etc.

As long as the slice of pie isn't too big, it still works the same way. In this painting, if you ignore the violet in the distant tree masses, we basically have an orange themed painting, but the lither colors (sky) has more yellow in it, and the darker colors leans more to the red slice of the pie. We are not seeing bright yellows, oranges and reds because saturation is kept in check. In this case, I'm reserving the saturated (relatively speaking) colors for the lighter range of the value scale. In the shadows, I drop not only the value but the saturation as well.  Can you use a saturated dark red in there? Sure you can. But you have to ask yourself, what's making it so bright in the shadows? The color of the sky affecting the dark areas, where it's not even facing the sky? Might make sense if someone was having a bonfire at the base of the trees. The point is, without a good reason, pushing color becomes a purely subjective decision, and the more you do it, the more you deviate from a structure that makes logical sense. What's wrong with that? Nothing, but if the painting ends up not describing a convincing light / atmosphere environment, that will be the price for your expression.

The little bit of violet in the back trees is a deviation from the slice of the pie. But not by much. The violet leans heavily toward the reds I used, and the yellow's in there too, to tamper the intensity and ensure harmony. The saturation and value are kept in check so that the violet, even though it's different from the rest of the painting, doesn't stick out.

Yellow. The lighter end of the scale is obviously yellow, and the darker end - I needed to go very dark - becomes a very grayed down dark warm color. Grayed down because at that value, we can't tell a yellow from a red. But kept warm (it's a reddish brown, very close to black) so that it harmonizes with the yellows. To ensure this, I used TOR and black both in the very dark areas, and in the very light areas. The yellow sky isn't very intense after all; the impact comes not from the yellow color, but the strong value contrasts.

This painting is a little more complex than the previous ones in that color deviation from the single-color structure includes introduction of local colors. The green of the trees, for example.  But notice that only the trees in close to the viewer is green, and the far ones just become darker version of the yellow /brown structure.

And, the greens you do see are not super green. They're more like green versions of the foundation color. I figure out the value that these greens need to be, and nudge the yellow/brown in the direction of green by slowly adding green into it. It helps to use the same yellow (ochre, in this case) to make that green. This way, I can maintain close harmony and the look of a very tonal painting.

The same thing is done with the violet grays of the pavement. They're just slightly violet version of the yellow/brown that I started with, and nudged the colors a little bit at a time till I got what I wanted.

The big exception is the bright red I used for the tail lights. Why does it work? Because 1)they're accents, used very sparingly. and 2) they are their own light sources. Because a tail light is a light source in itself, it can have its own color, especially if they're close to the viewer and are being less affected by the colored atmosphere that we have established.

Peachy again, with some local colors used as accents - the green of the palms, the reds of the tracks, and a few small spot colors on the figures at the bottom. But again, these colors are used very sparingly, and are nudges versions of the foundation colors. The red can be pushed without going out of harmony because it's part of the peachy DNA.

Blue! The dark areas become almost black, but still has a lot of blue in them. In the distance, I have a few different color shifts - some violet, which is a closely analogous color to the main blue, and green in the trees, which is also a closely analogous color, but on the other side of the slice of the blue pie. They're both nudged versions of the blue.

The sunlight in the distance is a pale yellow, but it's not very saturated at all. It does have blue in the mix, along with a bit of black, but the biggest common denominator between that pale yellow and the blue is white. You can see that the blues, especially surrounding the pale yellow sun light isn't very saturated, and the values are closer to the yellow.

To further integrate the yellow into the otherwise blue painting, I brought it in to the big field of passive area near the bottom of the painting. The Double yellow line helps to tie them together, also, but that's getting into the accent for expression's sake domain that I mentioned earlier.

And the red tail lights again. I can get away with using bright reds because they're light sources, and used as bling.

This last one is built on a muted red-violet structure. The color deviations are either analogous (pavement), nudged local color (yellow bus) or light sources (tail lights). All other variations happen within the slice of the pie.

It's a simple system, but when you put it together with slight variations, you can end up with a painting that doesn't feel monochromatic, yet very tightly harmonized. I love working this way because it's logical, yet allows for a lot of subjective variations, and I can push and pull between simple monochromatic structure and complex combinations. Tonal paintings are very effective in creating the kind of mood I want to express, and building on a single color structure keeps things from getting out of control.

Boy that was long winded! Thanks for reading till the end! Happy Holidays!!!

Wednesday, December 10, 2014

Color Palettes: The Brown Palette

One of the most common questions I get asked is, "what colors do you use?" so I thought I'd talk a little bit about my colors in a series of posts.

I work with a handful of different color "systems", depending on what I'm trying to do. But whatever "system" I'm using, I typically have the same set of colors on my palette - I may add one or two others as needed, and I don't always use all the colors that I squeeze out onto my palette.

The basic colors are as follows–they're all Gamblin paints, unless otherwise noted;


  • Permanent Red (Rembrandt)
  • Alizarin Permanent
  • Transparent Earth Red


  • Ultramarine Blue
  • Cerulean Blue
  • Ivory Black


  • Cadmium Lemon
  • Cadmium Yellow Deep
  • Yellow Ochre

Titanium White

Basically, it's a variation on the primaries palette. I mix my greens, oranges, and violets instead of using tubed secondary colors. There is a warm, a cool, and a low-chroma version of each of the primaries.

Transparent Earth Red is Gamblin's name for Transparent Oxide Red. Each brand has its own name for this one.

I consider Ivory Black to be a blue. A very greyed down blue, but a blue nonetheless.

Sometimes I use Cobalt Blue instead of Ultramarine, Prussian Blue instead of Cerulean, and Paynes Gray instead of Ivory Black.

Sometimes I use Indian Yellow instead of Cad Deep.

Sometimes I use Asphaltum instead of Transparent Earth Red.

OK so those are the colors on my palette, most of the time. Now let's talk about the brown palette, which is what I used for this painting. This is your basic earth tone palette that the pre-impressionist guys used; Velazquez, Duveneck, et al. Mind you, I don't know exactly which pigments the masters used, but the system is a simple one. For the core colors, I use Ivory Black, Transparent Earth Red, and Yellow Ochre for the three primaries, plus White. I don't have some of the classic earth tones like umbers and siennas.  Nothing wrong with umbers and siennas – after all, they were good enough for the Old Masters – but mine is just an earth tone version of the simple primaries palette. I believe that the modern Transparent Oxide Red is a synthetic color as opposed to having been made with natural iron oxides, so they (the TOR) have much more intensity and are cleaner (less muddy, both visually and literally) Most of my painting I'm posting today is done with just these four colors.

The few bright colors used as accents - the green jacket the woman is wearing and the dark blue-green of the seat back has some Prussian Blue in it.

I may have used a tiny bit of Permanent Red for the man's jacket and the server's ear...and there's a spot of red on the table, and again on the bow-tie guy's cheek. But just about everything else is painted with Black, Trans. Earth Red, Yellow Ochre, and White.

This "brown" palette system works very well for old-school tonal paintings like this, especially interior scenes where there isn't very much ambient light.  Without much ambient or bounced light, the shadows become very dark, and these dark shadows are painted very thinly and transparently.

I don't particularly think that transparent shadows work very well if it's lighter in value or if you can actually see lots of color and detail in that area. There are stylistic considerations of course, but for "traditional" representational painting, I tend to reserve transparent shadows for very dark areas, and this brown palette interior genre is full of them.

You can see that the shadows in this painting are so dark they're practically black. You can also see that these dark areas connect with one another, and there are no details or color information in these areas.

There are just a few areas where you can actually see anything in the shadows - the server's apron has some shadow patterns which are lighter than the dark receding shadows so that they're visible. It's only because the apron's local value was so light to begin with that I thought I should keep it visible even in the shadow areas.

In this type of set up, you don't have a lot of colorful impact, and it would be a mistake to try to impose color contrasts into it–the brown palette is not very good at accommodating impressionist temperature shifts. You can try it, but I think you'll find that the more you do it, the less convincing the light and shadow relationship will become.

The brown palette is really good for–surprise!–brown paintings. Seems obvious, but I see students trying to combine this tonal palette with high key color temperature shifts all the time. In fact, I've tried to do it (despite my instructors telling me not to) for years before I finally came to the conclusion that may be my instructors were right.

So we can't rely on color contrasts to provide impact. But we can rely on, and get away with, value contrasts! In fact, you have a much wider value range to work with than when you're working with lots of color. You can't easily get away with huge value ranges when working with color temperature shifts, because, quite simply, the saturation of colors diminish to nothing when you approach the extremes of value (black and white).

If I were painting this scene with a more impressionist approach, you can be sure that the server's black vest and white shirt / apron would not be painted in these values; they'd be much closer in value, and taking advantage of the saturation ranges available in the mid values.

The bright(ish) colors I do use in this painting are just spot colors, or accents. They're used sparingly, and if saturation is emphasized, it's still the local color that's pushed, not the color of the light source(s).  Consequently, even if you do see light and shadow on a brighter colored area, temperature shifts therein is minimized or nonexistent. I might even say that temperature shifts are almost irrelevant in this context.

Did I already say that the dark shadows are painted transparently? OK, yes. The opposite is true of the lit areas, which are all painted opaquely. That's kind of a simple rule of thumb. But what about the shadow areas which are still visible, like the shadows on the apron? That's painted opaquely too, but not as thickly as the lit area. Plus, I dragged some transparent paint over it (glazing) after it was dry, so that it relates better to the rest of the dark shadows.

Another exception is the background, where it's a little lighter (upper right corner). That was part of the underpainting where I took a paper towel and wiped off the dark paint. I left it like that because it seemed to work as is. Thin paint doesn't jump out like thick opaque applications, so in this case it worked well even though the area is not a dark receding shadow.

Next I'll talk about the single-color-themed tonalism.

Friday, December 5, 2014

A Start And A Finish

A Hint of Jasmin, 15 x 30 inches, oil on linen

Another recent favorite, A Hint of Jasmine (Click on image to enlarge) went through a lot of changes as well. The most obvious change is the color of the dress; it started out as a red dress, just like in the painting in the last post.

Somewhere along the way, the dress turned white. It's not because I wanted to express the idea of purity, or some other notion about this particular subject. The decision was a visual one. You see, I was having difficulty integrating that stark red into the rest of the painting. It seemed too isolated. By making it a white dress, I was able to lean the shadow areas toward violet, making them relate much more closely to the background.

At one point I had the background very dark. It was very dramatic, but also sinister in a way, so I brought light into it. The considerable back and forth resulted in more a involved abstract surface. Compared to the red dress stage, which was very early on, you can see the finished version has many more layers of pushing paint around.

In fact in the early stage, I was pushing paint around to find the light and shadow pattern in the folds of the drapery and the sheet. I was looking for shapes and values based on reality, whereas in the later stage, I'm going against it, in my effort to integrate and obscure.

The same comparison can be observed in the strokes that I used to paint the dress, her legs, arm, and hand; in the red stage the strokes describe form, or try to, anyway. The folds in the dress are painted fairly directly. It's straightforward. In the legs, the most visible strokes express the core shadows as the forms turn from light to shadow, meaning at this point, I'm sticking to "rules" of representational painting. In the later version, there are many notes that have nothing to do with describing form. They intrude and interrupt the conventional representation of form, becoming less about the figure lying there and more about the expression of the artist's (that's me!) identity.

It's very tricky to not lose sight of the subject completely, though, and I find it a struggle to maintain balance. I won this battle, but I lose many, too. One of these days, I hope to win more than I lose.

Wednesday, December 3, 2014

Lost Edges

Second Thoughts, 18 x 18 inches, oil on linen

This is one of my favorite paintings I've done recently. It took forever, too–though it may not look like it.

The background changed over a dozen times. I tried a very dark background, and a very light background, and a bunch of in-between values, and I flip-flopped a lot.

The figure itself (except the head) came together pretty quickly, mainly because I stayed within a "normal" color scheme. That is to say, the local colors of skin and dress are more or less what it was, and I did not emphasize the color of the light source nor use a subjective color theme.

I wanted her hands to be suggested and not rendered, so that took many tries of pushing paint around to make it look like I (almost) accidentally slapped paint on the right spot.  As nothing else in the painting is rendered tightly, the hands needed to fit that context too. But it's really hard to get the drawing right for a convincing gesture without resorting to noodling out each finger. I was very happy with the way they came out.

The head gave me a lot of trouble, too. I wanted to suggest anonymity and obscurity, as opposed to independent identity, which meant that I needed to paint it without much definition of actual features. 

Cropping the head where I did, is another way of not defining the identity of the figure. It has to do with creating mystery, and that means withholding information. If you know my work, you may have noticed that unless I'm specifically doing a head study, I don't paint facial features. I throw the face in the shadow, turn the head away from the viewer, abstracted it, or like in this case, chop it off and do all of the above.

This painting utilizes lost edges quite a bit. There are a lot of areas where one shape encroaches onto another, resulting in abstraction and simplification. If something doesn't need to be clearly defined, why define it? If there's a good reason – like separation of shapes creates two nice shapes rather than one boring shape, or may be it creates more tension, or it's a chance to use another color, or the painting just doesn't make sense without that separation– but there needs to be a reason for it. 

Wherever you see me lose an edge, you can be sure I've tried separating the shapes too. I may have gone back and forth between losing and keeping that edge several times. That's typical of how I work. I mean I just can't tell which is better until I see both ways. And even then, I may change my mind later as surrounding areas change as the painting develops.

One way to make losing edges less scary is to make sure your color harmony is working. In this painting, everything in light has yellows and reds in it, and everything in shadow leans toward violet. Even the brownish areas of the skin in shadow, and the dark red areas of the dress in shadow have blues in them. When two adjacent areas are not only close in value but also in color, connecting them to make one shape is a lot easier to pull off than if the colors are very different.

Two adjacent shapes being close in value is almost a given for a lost edge between them. They don't have to be exactly the same, but if they're very different, you can't connect them without having some sort of a gradation, which is more like a soft edge and not a lost edge.

The boldest lost edge in this painting is where I pulled in the background color into the girl's back. You can see I pulled the color also over the chest and the neck area. Though I didn't lose edges there, having the same colors there ties the whole area together, and helps to make sense of the foreign color in the area where we expect to see a normal skin color or that of the dress. 

The real key to pulling that off is to make it look absolutely intentional. If you're unsure or timid, it shows, and it'll look like you tried to fake it. You have to put it down like you mean it, even if you don't know what you're doing. As I always tell my students, paint like you're lying to a child.  Santa Claus? But of course he's real!

One more thing on edges. Very sharp edges are like accents, especially in a context where much of the painting is loose and brushy. I tried to use sharp edges strategically. Most of them, if paired with value contrast, help to lead the eye to the hands. Edges which are sharp but where values are close don't stand out so they don't demand attention from the viewer, but they do contribute to the sense of decisiveness and intention to the painting. But whether they're used as accents or not, they need to be used in conjunction with soft and lost edges to be meaningful. I'm only talking in the context of my paintings, of course. I can't make a broad statement like that and claim it to be true for all paintings.

Wednesday, November 19, 2014

Sketching People


Last weekend I tagged along with a geology class from CSU Sacramento on a field trip to the South Fork of the Yuba River. It was a really great location with lots of interesting rock formations and swimming holes. 

Actually it was my son (still a high school student)  and a few of his Science Olympiad teammates who were invited to tag along with the college kids, and I was just the driver. But I had been wanting to check out the Yuba River to scout out painting locations, so this was a great opportunity. 

Basically, while the class looked at boulders I sat around and sketched them in my sketchbook.

Most of the time, they didn't stay motionless for more than a few moments, so these are really quick drawings. My only aim was to try and capture the gesture, or the attitude of the posture in each sketch.

I used a regular ball point pen and my sketchbook is a Moleskine. 

I don't do as many of these sketches as I used to, and  I was newly reminded just how enjoyable doing these are, and how fundamental these kinds of sketches are in defining the identity in my art. I think the skills required in capturing the quick gesture is at the core of everything I do; not just when I'm painting the figure, but also trees, rocks, even cars and nonrepresentational abstract marks. 

Some thoughts on doing quick captures:

Don't worry about doing good drawings. Just do a ton of them, one after another. You have to let go of expectations and allow yourself to do bad sketches.

It's about the gesture. That means, you need not and should not even think about likenesses. Instead focus on communicating what the figure is doing. And this goes for large overall gesture of the whole body, as well as individual curves and straights. What is that shoulder doing? What kind of curve is that shin? 

If you try to follow the precise outline of an arm (or a head or a leg or back or whatever), you're only thinking about shape. Or worse, just the outline of the shape. Shape is going to happen anyway when you draw both sides of the torso (or the arm or the leg or...) so don't worry about the exactness of the shape. Instead worry about what that shape is doing. Is it stretching? contracting? curving gracefully? supporting weight? flowing from one point to another? Overlapping? Obscuring? Turning around? Twisting? Cutting something off? Lining up with something else? What's it doing?

These are the questions that go through my head as I sketch. Likeness will happen on its own when you get the gesture right. It's almost like magic.

Try to see the whole figure at once. If you only look at the tip of your pen or pencil, you're not relating the line you're drawing to anything else. Consequently you will have a hard time with proportions. 

When you are seeing the whole figure, big proportional errors are really easy to spot. And when you spot them, you can fix them. If you can't spot them (because you're not seeing the whole figure), you have no hope of fixing them.

With these quick sketches, there's no time for measuring, so seeing the whole figure becomes even more important.

It's easier to see the whole figure when you draw smaller.

Drawing around the form - using cross-contour lines- is just as important as drawing the outside contours. And these cross contours should never be treated as an afterthought to add volume to an otherwise flat drawing. They should be used as you build form on the paper. 

Sometimes these lines are clearly visible as hem of a shirt or pants, cuffs of a shirt, brim of a hat, hairline, belt, stitches in clothing, etc. Other times, you just have to put them in to show volume. Think about how you would draw a sphere (a three dimensional form) rather than a circle (a flat shape with no volume). If you could only use line and not rely on shading, how would you do it?

Straight lines are easier than curved lines. If some part of the body has a very straight "attitude", emphasize it by using a straight line. Sounds obvious, but if you look closely, rarely you see an actual straight line on the body so anything you represent as a straight line on paper is you imposing your perception onto your drawing. 

Treat curves similarly. Accuracy of the curve is not important. It's how you interpret and impose the attitude of that curve onto paper that makes your drawing come alive with your intent.

When drawing one side of a form, look at what the other side is doing. Does one side echo the other side's attitude? Does it oppose the other side? Does it play call and response? Trying to verbally describe how the two sides of a form relate to one other forces you to consider the intent of your line.

When you draw a line, see if it has a rhythmic quality that flow into another part of the figure. Emphasizing this "flow" will have a profound impact on the gestural quality, and also on composition as whole. This is what we call continuity of rhythm and if you look at old master's drawings, you'll see it used all over the place. And you'll start to notice that this flow is not accidental, but it's actually the intent of the artist. 

Anyway, get sketching'!! You can do this just about anywhere, and it takes no time at all to do one or two figures while waiting to pick up your kids at school. And you don't even need your paints. I don't know about you, but I really have no excuse for not doing them. (Guilt motivates me. So there.)

Tuesday, November 18, 2014

Rainy Night, Chicago

Rainy Night, Chicago, 18 x 48 inches, oil on linen

Here's a painting that I did recently. I was in Chicago for a few days this summer, and took a bunch of  photos while walking around aimlessly. I've started a series of cityscapes based on this trip, and this is  one of the first ones. 

I hadn't done a lot of nocturnes up to now, so this was a new challenge for me. It turned out to be quite an educational experience. 

One of my goals, as usual, was to push abstraction. In this particular case, all we really see are bunch of lights and reflections on the wet pavement, with a few recognizable "things" to help define the context. I wasn't sure if it would be enough to carry a composition and a depiction of a convincing environment, but I think it worked out better than I expected.

I worked on this over several sessions, so there's a lot of wet on dry layering going on, as well as wet into wet mushing around. I like to combine the two to get interesting interaction between notes and edges.

The crosswalk going across the painting was towards the end of the painting. On a dried surface, I used a straightedge to draw in the line with a pencil, (it's actually done in three or four sections, to give it a very slight curve) which I used as a guide to paint the lines. Then I worked the paint to integrate the lines into the surrounding areas, so that they wouldn't look pasted on. Variations in value and color were done at the same time.

The various lights and their halations(?) were done wet into wet, generally. I would first paint the area without the lights, and then drop the halation in, working the wet paint to soften its edges, followed by the bright "center" with slightly firmer edges. I tried to make sure each of these light centers had unique shapes, just so they didn't look stamped in.

The halation effect is the light sources illuminating the moisture in the air, in this case rain drops and drizzle. It's essentially the same thing as sunsets being orange, just on a very small scale.  They are in effect, transition areas between the surrounding darkness and the bright light center. If my light centers are opaque and the darkness around them are transparent, how should I paint these transition areas? If I acknowledge that these transitional areas are rain drops and drizzle being illuminated–that is to say, they are lit things,  I can apply the rule of thumb; if it's lit, paint it opaque. 

Obviously there are many ways to paint with oils and opacity and value can be thought of independently. You can paint these areas transparently, if one so choses. I just like to have a logical answer to my questions, and besides, I've tried it a bunch of different ways and opaque halation always looks better than transparent. In my paintings, anyway.

Oh, and if you'll notice the yellow lights next to the big light post on the left side of the painting, they are affecting the value and color of the pole even though the lights are behind the pole? What's up with that? That's diffraction, where light bends around the object as it passes by it. In practice we see this in painting trees in landscapes. Say a tree is painted against the sky, which is much lighter than the tree mass, the small branches against the sky are painted much lighter than the big trunk, even though the local value of the twiggy branch is the same as that of the big trunk. Why? because the light coming through the branches bends around them so some of the light spills in front. Similarly, if we are looking through a window at a much brighter world outside, the small lines of window panes need to be painted lighter than the window frame for the same reason in order to look right. 

If you push this effect a little bit, what you end up with is a suggestion of a more atmospheric view, a very effective device in creating mood. 

This painting is at Sloane Merrill Gallery in Boston. If you're in the area, drop on by and check it out!

Friday, November 14, 2014

A Looser Start

The last post was an example of a tight start (d)evolving into a more abstract expression. I think I mentioned that I don't always work that way. In fact, more often than not I start my paintings quite loosely.

This one I'm showing today is a smallish (7 x 14) study for a larger painting.  The first image is the start of the painting. As usual, I'm using Claessens No.66 oil primed linen, and using a small brush, I drew out the placements of the major elements.

The brown color is Asphaltum, which is just a mixture of bone black and mars red, I think. It's a warm transparent reddish brown that's not as red as transparent oxide red, and cleaner than burnt sienna.

Sometimes I use a straight Ivory black for this stage, and often I mix transparent oxide red and ultramarine. But any grayish brownish transparent dark color will do.

The composition is basically an arrangement of dark geometric shapes in a light valued field, so it didn't require tricky drawing or anything. It's just a pattern.

The really tricky thing, I thought, was the firescape, its shadows, and the traffic light all stacked on top of one another. I wanted this area to be a somewhat abstract jumble at first read, and may be make sense upon further study. In other words, I wanted expressive mark making to obscure some of the literal depiction.

Same thing with the far left dark shape representing another traffic light post. It's a lot of painting, scraping, repainting, pushing and pulling. The identification of these abstracted elements rely heavily on the context within which they exist. It's not dissimilar to how we see the world around us. We can only focus on just a small area of our field of vision, and everything in the periphery is blurred / abstracted, (especially if we are moving) and yet we are able to identify the various objects in our field of vision and recognize how our environment is laid out.

If something is out of context, we notice it. On the other hand, if something is perfectly in context, we don't need it to be defined so clearly for it to be recognizable.

By starting the painting very loosely, I'm able to establish the visual context very early. I can get a rough idea of just how much definition is needed for the environment to make sense. Once that's established, the degree of "tightening up" is not for recognizably of things, because we can already tell what it is. That squarish blob is a window, for example. We know this without having to render window panes. 

So how do I decide how much further to go? For me, the question is one of balance and expression. I pick a few elements to describe relatively tightly, to provide a focal point and an anchor of sorts for my visual context. And I decide which areas can be so abstracted that they can't be recognized without context. Everything in between, is... everything in between. 

I play with super sharp edges against goopy paint, thick areas against thin, textural against smooth. I try to have fun just pushing paint around, always checking to see if it fits my context. If it doesn't, does it still work? Sometimes it does, other times, it needs to be reined in. But if I can remember just how little information was needed to define the environment in the initial loose lay-in, I can keep myself from over rendering. In theory, anyway. 

In practice, it's still hard to stay loose and expressive. Painting loosely (yet drawn well) didn't come naturally to me, and it doesn't to most painters. I'm an analytical guy (I think I have two left brains) so even learning to painting loosely had to be somehow logical.  

It's starting to look intuitive, finally, so I'm pretty excited about that.

Wednesday, October 29, 2014

She Disintegrates

Upon reading the last post, a reader asked me whether it would make sense to first paint the figure with a lot of information and then proceed to break it up, or would it be better to start with lots of abstraction and tighten up only where necessary.

Good question. if you ever find out, let me know!

Seriously, I don't know which works better. I do both ways, depending on whim, mostly. I do like my images to be driven by good drawing, so at some point –either at the very beginning or much farther into the process– I like to have the entire figure drawn convincingly. 

Most of the time I don't really get excited about rendering realistically, nor do I think it's relevant in what I do, but once in a while a certain level of realism creeps in. I suspect it's when I'm feeling a bit insecure that I start noodling, not knowing where to take it or how to express myself. 

I don't fight the impulse, for I know that if I keep going, sooner or later I'll have satisfied my doubts about painting "realistically" and then I'll get bored with it. Consequently abstraction and expression are inevitable. 

The painting I'm posting today is one such example. The very first stage, at the top of the page, was painted with a live model in a couple of hours. I don't remember why exactly I decided to paint her this way, but I had a rather rendered painting at the end of the session.

I have to stress that I don't normally paint this tightly. Only once in a while, just for kicks.

The seated pose didn't work too well, because I didn't do a good job placing the figure on the 12 x 9 panel, and her knees and her finger tips came too close to the edge, which bugged the hell out of me. 

My solution was to change her pose. As the seated figure was just one session with a model, this posed a bit of a challenge. The model's gone, and I have no reference photos or drawings. Can I change the pose without any refs? I didn't know, but I decided to try it. I mean what have I got to lose? It was already a loser, so no risk there. 

The new pose turned out ok. I only changed the lower part, and her arm, so it wasn't too drastic. The lighting was simple, so there was no need to make up a complex shadow pattern either.

At this point, I simplified the background and rendered the figure in a pre-impressionist glazey style. Still in the realist mindset, but not concerned about subtle skin tones. This is a more or less a monochromatic tonal rendition.

It's a simpler representation than rendering subtle warms and cools of the flesh, because all I'm doing is modulating value without dealing with temperature shifts. 

After letting the painting sit around for several weeks, I came back to it and started to introduce some abstraction. I was playing around with background patterns, (changed many times) and repainted the figure using opaque, patchy strokes.  The patchy shapes don't necessarily have anything to do with the form it sits on. I'm still controlling the values carefully, but I'm also intentionally not responding to the form with my strokes. It's harder to do than it sounds, especially if you've been trained to mind the form with stroke directions all these years.

But disconnecting my strokes from following the form, I found, is a significant way to move away from realism, while maintaining realism with values. Does that make sense?

This is where I am right now. I'm still fiddling with background, trying different colors. The figure is breaking up more and more, and the rate of change, if you will, is becoming faster as I become more and more comfortable with the idea of abstracting this particular figure.

It's a funny notion, that I have to become comfortable with abstraction each and every time I start a new painting. It's like going through the same journey of insecurity, tentative attempts, loss of control, and embracing risk over and over again.

I suppose I'm attracted to this maddening roller-coaster ride, and that's why I keep coming back to it.

This painting is not finished. I'm still fiddling with it. The next step is to try some bigger strokes in the background. As it is, the notes in the background seems too fussy.

After that, I'll reassess and see what else jumps out at me.