Trees on Points

Here’s the first rough block-in of this painting. It’s been downsized to 12×16″ from the planned 16×20″, because I found I didn’t have a 16×20″ panel and didn’t want to wait until I could get to the art supply store! The slightly longer format will let me hint at some of the trees on the extreme right, which is actually a good thing because it explains the cast shadow on the “main” tree.

The colour in the photo is a bit washed-out, but not bad. You can see the raw sienna ground and underpainting peeking through in a lot of spots; I’m hoping it will give the final painting a nice warm glow.

Getting to Work!

If you’ve been following along on this blog, you may have noticed that I’ve been slacking. True, I have been studying a lot, and I’ve learned a lot. But I’ve haven’t actually been painting any landscapes. That was intentional; I wanted to wait until I had specific lessons to implement, specific ways to avoid repeating the same old mistakes. Now it’s time to get to work.

For the first attempt (and also because it’s winter and cold out there!) I’m working from a photograph – the same photo I’ve been using all along. I’m going to go back over the stages of planning that I’ve already covered, so it will be boring for you, I’m afraid. But I’m hoping that the repetition will help to burn the lessons into my brain! As I go, I’m going to create a sort of checklist to help me in future paintings.

So, back to the beginning: 

Site Selection

I described this issue in an earlier post. But the short version is, “A potential site must contain the types of visual cues painters rely on to differentiate forms and suggest depth.” (Mitchell Albala)

Since I am particularly prone to diving in without thinking through the preliminaries, I’ve rearranged the factors that Albala listed. Now they form an acronym that I’m calling The FLOP Test. It will (I hope!) be easy to remember and run through mentally before I start to paint. If the site doesn’t have at least one of these factors, the picture will FLOP!

Foreground, middle ground and background

Light/shade to create a sense of volume

Overlap and scale

Perspective

Does this picture pass the FLOP test? Yes, I think so. It does have a clear fore-, middle- and background; the points of land overlap somewhat; and there is some light/shade volume (although from what I remember the shadow on the big tree was a cast shadow; the sun was from the right.) There’s aerial perspective, though not linear.

Composition and Format

Format

I experimented with various crops and formats for this picture earlier, and decided that I liked a horizontal rectangular format. I’m going to use a 16×20″ board for the finished painting, which lends itself nicely to 8×10″ or 4×5″ studies.

Although I already did this experimentation digitally, I felt it was important for my learning process to repeat it using thumbnail sketches:

They’re nothing more than scribbles; they only took a minute or so each. And they don’t really look much like the photo. But even after all my digital manipulating of the photo, I learned a surprising amount from these sketches – where the values are, how the shapes fit together. They also confirmed my earlier preference for a horizontal rectangle.

As for the other aspects of compositon, the photo to the left shows the golden mean divisions, the rule of thirds, and the “eyes” of the diagonals. The sunlit tip of the front point falls at one of the sweet spots.  Two of the eyes fall at interesting points on the trees, and a third falls in the dark shadows below the tree. My eye keeps moving to those shadows, perhaps because that’s where the diagonal tree line on the left and the line of the tree trunks on the right meet, or perhaps just because I like the places where land and water meet! The trees on the left balance the darker, larger tree on the right nicely.

Simplification

In a previous post, I explored simplifying the values of this photo. Now I want to look at another aspect of simplification: what to leave out. It’s a bit hard to tell in the photo, but the shore curves around on the right to where the photograph was taken. Hence that little bush sticking up rather randomly in the foreground. Should I include the shoreline so the bush makes sense, creating a more solid foreground? Or should I eliminate the shore and bush completely? I experimented very roughly with leaving it it out and smoothing the water, and I think I prefer it that way. I think it would feel crowded if I extended the shoreline, and would change the emphasis from that little point of land that I want to focus on.

What do you think? Will this make a good painting?

Colour Charts

“The point is to know how to use the colours, the choice of which is, when all’s said and done, a matter of habit.” Claude Monet 1905

It sounds so simple, doesn’t it – “know how to use the colours“? But it’s not.

Step one in learning to use the colours is to paint colour charts. Charts are good for me. I know they are. They’re recommended by every painting instructor out there. And it’s true; I learn something every time I do one. But they’re not exactly exciting to do, so I don’t do them as much as I should. That’s something I must work on changing.

There are so many ways to do charts, but I started with three comparing different palettes.

The first one is the classic “rainbow” colours, known to every schoolchild by the acronym ROY G. B(I)V: Red (Permanent alizarin), Orange (Winsor (pyrrole) orange), Yellow (Azo), Green (Winsor (phthalo) green), Blue (French Ultramarine) and Violet (Ultramarine purple). (There are other pigments I could have chosen for each colour, and every combination would have yielded slightly different results. But this will give the idea.) The warm colour mixtures are clear, high-chroma colours. So are the yellow-greens. The blue-greens are a bit muddier, and so is the red-violet/magenta range.

The reason is clear if we plot the colours on a colour circle – the kind you see in most computer software. I’ve put a dot roughly where each pigment falls. The lines connecting the dots represent the possible mixtures of  those two pigments. Because the blue and green are so far apart, the line connecting them is much closer to the neutral centre of the circle than, say, the line connecting yellow and orange. This means the mixtures of blue and green will be less saturated, closer to neutral. Similarly for the red-violet line.

In general, mixing pigments that are close together on the circle gives higher chroma result than mixing pigments that are farther apart. Conversely, one way to get a low-chroma result is to start with pigments that are farther apart.

This chart is a split-primary palette, with a warm and a cool version of each of the traditional primary colours. I used Winsor Lemon, Cadmium Yellow, Scarlet Lake, Permanent Alizarin, French Ultramarine, and Winsor Blue. Again, there are other possible choices for each. The idea with this palette is that mixing, say, a greenish yellow with a greenish blue, you’ll end up with a higher chroma green than if you use a warmer yellow with a more violet blue. It works – in each case, the expected combination gives the most saturated secondary colour. The weakest area of this palette is the violet/magenta range, which turned out more like reddish brown. The greens are higher chroma than in the first chart, but still rather low chroma.

Here’s the colour circle for this chart. Again, the lines connecting points represent the possible mixtures of the two colours. (Any point within the “mixing outline” can be achieved using more than two colours, but no mixture will give a colour that falls outside the lines.) With the addition of the phthalo blue, the mixing line for the greens runs a bit farther out from the centre of the circle, so the mixtures are slightly higher chroma. Basically, though, the overall shape of the area of possible mixes is similar to the first palette – it’s still missing the outer (high chroma) portion of the greens and violet/magentas.

The third palette I tried is what James Gurney calls the YURMBY wheel. I used Azo Yellow, Scarlet Lake, Quin Magenta, Ultramarine Blue, Winsor Blue, and Winsor Green. This chart definitely has the highest chroma greens – especially blue-greens – and the best violets (excluding the tube violet in the first chart). It’s also got some very nice neutrals, including a near black when the magenta and green are mixed.

And here’s the reason for those nice high-chroma results: the six pigments are spaced evenly around the circle, so the mixing lines fall much closer to the edge of the circle than they did in the other two palettes, especially in the green and magenta/violet ranges. This means that the mixtures can be much higher chroma. And of course, it’s still possible to mix low-chroma colours by choosing a pair of non-adjacent colours, or by adding a third pigment to the mix.

Simplification and Massing

This is the fourth in a series of posts as I work through Mitchell Albala’s book Landscape Painting. The first three are site selection, format, and value.

In the early stages of learning to draw and paint, we’re told, “draw what you see”. At that stage it’s good advice, meaning “don’t just draw the symbol your brain uses for that object”. The problem is that at a slightly later stage in learning to paint, it’s awfully tempting to want to paint everything we see, including each and every one of the 200,000 leaves on that big tree there!

Of course, painting all the detail can’t be done. The trick is to simplify, to mass the values and colours into shapes that represent the major planes in the scene, and to find the degree of simplification that is the perfect balance between too much detail and not enough. But it’s difficult to know where to begin. Mitchell Albala says,

“Painting or drawing a shape is not difficult, but seeing a shape through layers of surface detail and complexity requires a practiced shift in perception”

The usual advice for seeing a simplified scene is to squint. Squinting reduces the amount of light entering the eye. With less light, the rods of the retina, which see light/dark, take over from the cones, which see mainly colour. Details disappear, similar values blend together, and basic shapes begin to appear. This blurred photo isn’t exactly the same as squinting, but it gives the idea. The details of the leaves, the waterlilies, etc. all disappear and bigger shapes of similar colour and value emerge.

Albala also suggests a series of three exercises to help see the planes in a scene, one of which is a four value sketch, similar to the digital image at the right. It has a similar effect, in that it forces you to make decisions that simplify a wide range of values into just four, helping to create the big shapes.

I discovered a great example of using squinting and value to simplify on George Perdue’s blog. As a bonus, the photo he used was taken at Scotsdale Farm, which is a lovely spot less than an hour away from here! I like his work; I’ll be spending quite a bit of time exploring his blog.

Perspective of Inclines

I think this may be the last of the perspective topics. I’ve already looked at eye level, perspective centre, ellipses, and three-point persepctive. Now it’s inclines.

An incline in just a sloping surface – a ramp, a sloping roof, a road running uphill. Stairs are a variation on an incline.The simplest form of incline is just a rectangular box cut in half diagonally, as in the diagram.

If you extend the edges of an uphill incline, they meet at a vanishing point above the horizon. The edges of a downhill incline meet at a vanishing point below the horizon. But what I found most interesting is that the vanishing point of the incline falls on a vertical line through the vanishing points of the original rectangular box that frames it. Each object can have a different set of vanishing points, but the incline VPs always fall directly above or below the “box” VPs.

Hills are a common feature in landscapes, so it’s important to understand their vanishing points. The fact that the VP is above the horizon is one of the main clues that the road is uphill rather than flat. You can see this in the photo at the left (from Wikimedia Commons). But to really drive the point home, I’d suggest heading over to Google. Do an image search for “straight road” and  ”uphill road”. Within the first few hits, you’ll see some photos of straight-to-the-horizon roads. Take a few minutes to find their vanishing points; you’ll quickly be convinced! (I couldn’t find a really good photo of a straight-downhill road, but if you have one, check its VPs too.)

Value

Value is the topic of the third chapter that I’m studying in Landscape Painting by Mitchell Albala (it’s not the third chapter in the book; I’m studying them out of order to fit my own process).

Value before color.

That phrase should be carved on my easel in big letters as a reminder. Albala goes on to say, “Value can perform without color, but color always has value as one of its attributes.” A black and white photograph is enough to understand the subject. I’m old enough to remember black and white television, and it seemed quite normal. Colour is a bonus.

A few years ago I took an introductory colour class at a nearby art school. The instructor was excellent; she used so many historical examples to explain and demonstrate each concept. The one that has stuck most strongly in my mind is the story of Superman. You know, the guy in the blue and red costume? Except that his original costume was grey and brown. Why? Because he was originally filmed in black and white, and the values of red and blue are too similar. The grayscale photo of the modern Superman below conveys the meaning, but it’s the value contrast in the old photo that makes it interesting. Value before colour.

I took the photo I’ve been using to study site selection and format and converted it to grayscale up there at the top of the post. It has a pretty good range of values, although it could perhaps use a few more darks on the left.

I also experimented with another photo, over there to the right. Albala had said in the chapter on site selection that a “wall of trees” is usually a bad choice of site because it doesn’t have enough value contrast, so I looked at a crop that is mostly the trees on the far shore. I think he’s right, it is less interesting. “Where a plane changes, there is a corresponding change in value.” This whole scene is essentially one big plane of trees, although there are of course smaller changes within it. So the whole scene tends to be of similar value.

Then there’s John F Carlson’s idea that a plane’s value varies according to its angle. This may not be the best photo to use because of the reflections in the water, but it seems to hold true. I took samples from the sky, water, angled foliage and vertical foliage, and the values get darker for each:

As a final experiment, I converted both photos into just two values, black and white. The “wall of trees” is a rather confusing jumble, but the other makes a much better picture, with fewer major value shapes.

Golden Ratio, Rule of Thirds, and Rabatment

The Golden Ratio in Art

The golden ratio has been known and used since ancient times. Its most common use by artists is to locate their centre of interest. The lines joining the golden ratio points of opposite edges of the canvas intersect at a spot that is considered to be the most visually pleasing, the “sweet spot”. Seurat and Turner, in the paintings below, have used this method to place their centres of interest:

Bathers at Asnières Seurat	The Fighting Temeraire Turner

The placement of the golden ratio intersections varies according to the proportions of the canvas.  In the first format below, the golden sections divide the square canvas almost in thirds.  In the second format (a 1 x 2 ratio, for example a 12 x 24 canvas) the lines fall closer to the centre:

The Rule of Thirds

Most standard-sized canvases have a length:width ratio of between 1.2 and 1.4 to 1 (somewhat shorter than the “ideal” golden ratio of 1.618 : 1).  A 12 x 16 canvas, for instance, is 1.33 : 1.   On these formats, the golden sections (shown below in red) fall very close to the lines dividing the sides into thirds (shown in blue).  This has led to a simplification of the golden ratio principle, known as the Rule of Thirds, which approximates the “sweet spot” by dividing each edge of the canvas into thirds.

Rabatment

Another interesting option for placing the centre of interest is the rabatment. The rabatment is the square that results when the short side of a rectangle is rotated onto its long side, as shown by the green line at the right. It’s not as well known as the Rule of Thirds and the Golden Mean, so I wanted to experiment with it a bit. In each of the examples below, I’ve marked the various methods as follows:

red = diagonals, with the “eyes of the rectangle” (halfway points) as dots
green = rabatment
purple = rule of thirds
blue = golden sections

This is the reference photo for a painting I did a couple of years ago. I cropped it to precisely a golden rectangle, so the length is 1.618 times the height. Not surprisingly, given the way the golden rectangle is constructed, the rabatment falls exactly on the golden section.  The rule of thirds “sweet spots” are nearby (again, not surprisingly, since 0.667 is close to 0.618), and the eyes of the rectangle are just a bit further out, since they are based on quarters rather than thirds.

Van Gogh’s Sower with Setting Sun has a more standard length/width ratio of 1.26 (80.5 x 64 cm).  In this case, the golden and rule of thirds “sweet spots are again close together and the “eyes” are a bit further out; this will always be the case since these methods are based on constant proportions.  This time, though, the rabatment falls almost exactly on the eyes of the rectangle – again, not surprising, since they fall at the quarter marks of the canvas, and the height of the canvas is 0.79 times the length (very close to ¾).  The figure of the sower falls on the rabatment. (He’s breaking another of the traditional composition “rules” by walking out of the picture, but rules are meant to be broken – or at least bent a bit!)

Finally, Thomson’s Jack Pine is an almost square format (127.9 × 139.8 cm), with a length/width ratio of only 1.09.  In this case, the rabatment is way off by itself near the edge of the canvas, pretty well useless for composition planning. But look where the trunk of the tree falls, and the outer ends of those long branches!

Among modern artists, Alex Colville is perhaps the best known for his deliberate and extensive use of the Golden Mean and other geometric methods in planning the layout of his paintings. He’s not a landscape artist, so I’m not going to let myself get distracted by studying him right now. But he’s fascinating, so if you aren’t familiar with his work, I’d suggest looking him up.

The Golden Ratio

So, once you’ve decided on where to paint, and what format to choose, it’s time to start thinking about how to place the elements of the picture on the canvas. And one of the methods artists use to divide up the canvas is geometry. As a former math geek, I have to admit that I really enjoy this topic…

Geometry has been important in art since ancient times; it appeared in ancient Egyptian, Indian, Greek, and Roman art. During the Renaissance, math was seen as the rational key to God’s perfect creation. So geometry was the artist’s way of translating creation into a painting.

So, for example, in della Francesca’s The Baptism of Christ (at right, from Composition, by Sarah Kent) Heaven is represented by a circle, and earth by a rectangle. The spot where they meet – the centre of the circle – is marked by the dove, representing the Holy Spirit. Jesus is placed on the vertical median, and the tree and John are on the verticals that divide the rectangle into thirds.

Because it symbolized the Christian Trinity (God, Jesus, and the Holy Spirit), the triangle (especially the equilateral triangle) was often used to compose religious paintings. It was particularly commonly used in paintings of the Madonna and Child (as in Raphael’s Madonna of the Meadow, at left), and of the Crucifixion.

The most famous mathematical compositional tool, though, is the Golden Ratio, also known as the Golden Mean or the Golden Section. It is based on the number Φ (phi), which is approximately 1.618. Phi is an incredibly cool number (well, to math geeks, at least).

Essentially, the Golden Ratio is the point at which a line can be divided so that the ratio of the long segment to the whole is the same as that of the short segment to the long.  So:

1 : 1.618 = .618 : 1

And did you notice that the short segment’s length is Φ-1?  Even better, Φ² works out to 2.618, or Φ+1.  It has other fascinating properties, too, like being the number toward which the Fibonacci series converges.

A rectangle with sides of length 1 and Φ, called the Golden Rectangle. is shown at right. If you divide off a 1 x 1 square at one end, you’re left with a small rectangle with sides 0.618 and 1 – another Golden Rectangle! And if you divide off a square at the end of that rectangle, you’re left with an even smaller Golden rectangle. And so on. If you then draw quarter circles in all those squares, you get a spiral that duplicates many found in nature:

Totally awesome. No wonder in the Middle Ages it was known as the Divine Proportion, embodying God’s perfect logic!

Ahem. Perhaps I’m not such a former math geek, after all!

In the next post, I’ll look at using the Golden Mean in composition.

Three-Point Perspective

In one-point and two-point perspective, horizontal lines converge toward vanishing points on the horizon, but vertical lines are considered to be parallel. That makes sense, because those lines are usually an equal (or nearly equal) distance from your eyes along their whole length.

But imagine for a moment that you’re in an airplane, looking down at a tall building. The sides of the building are no longer an equal distance from you along their entire length. Instead, they’re receding toward a vanishing point (called the nadir) somewhere down below, as in the diagram of Escher’s Tower of Babel at the left. This viewpoint is sometimes called bird’s eye view.

Vertical lines can also converge at a point above the viewer, called the zenith. Imagine yourself lying on the sidewalk looking straight up at a tall building. From that point of view, the side of the building take on the same position relative to you as, say, a road would if you were standing on it – it recedes away toward a vanishing point at your current eye level. Lying on a sidewalk to study perspective is a bit extreme, but standing and looking up creates a similar effect. This is what happens in the photo of the Canadian mint at the right. This viewpoint is sometimes called worm’s eye view.

Three-point perspective is more common in comics, where it’s often used to an exaggerated degree, than it is in landscape art. However, I think it’s useful to study it anyway because it helps in understanding the concept of multiple vanishing points which are not necessarily on the “normal” horizon. And that understanding will be very useful in the next perspective topic – inclines!

Tom Thomson

Spring
Tom Thomson, 1916
National Gallery of Canada

When I searched for images to illustrate my post on formats, I realized that I don’t know much about art history in general, and landscape history in particular. So I’m starting an ongoing series of posts about famous (or possibly not-so-famous) artists. I’ll start with artists who painted the landscape I know best, and work out from there. And since I am Canadian, that means starting with Tom Thomson and the Group of Seven, because they are pretty much synonymous with Canadian landscape art in most people’s minds.

Canada’s famous Group of Seven almost certainly would have been the Group of Eight if Tom Thomson hadn’t died (under mysterious circumstances) in 1917, before the Group was officially formed. He knew and worked with the other members, and was the first to paint the northern Ontario landscape that they became so well know for. So I’ll start with him.

Tom Thomson was born in 1877 and grew up near the small town of Leith, Ontario, on southern Georgian Bay. He attended two different business colleges (one, in Seattle, owned by his brother), which helped develop his skills as a commercial artist. His early career was rather unsettled, but in 1909 he joined the Toronto design firm Grip, Ltd., working with Franz Johnston under the head designer, J.E.H. MacDonald. In 1911, they were joined by Arthur Lismer & Franklin Carmichael, and in 1912 by Frederick Varley. At a 1911 exhibtion of MacDonald’s sketches, he met Lawren Harris. The story of the Group of Seven was set to unfold.

The Grip artists often organized weekend painting trips near Toronto. Thomson’s work at that point was quite conventional, like the watercolour landscape to the right which he painted in 1910. His turning point came in 1912 when he made two trips into northern Ontario, first to Algonquin Park, then canoeing in the Mississagi Forest Reserve west of Sudbury.

His sketches from those trips, like Pine Stump and Rocks at left, were still conventional and dark, showing none of the strong colour and brushwork he would  later develop, but his friends felt that they caught the character of the northern landscape and encouraged him to continue. In 1913, his first major painting A Northern Lake was purchased by the Ontario government. Also that year, Dr. James MacCallum offered to support him for a year to allow him to paint, and introduced him to A. Y. Jackson.

Unable to enlist during WW I for health reasons, Thomson spent a great deal of time in Algonquin Park from 1914 until his mysterious death in 1917. He sometimes worked as a guide and fire ranger as well as painting small sketches. He was often joined by his friends, whose more formal art training and knowledge helped him develop his style. During the winters he returned to Toronto and worked the sketches up into finished paintings. From 1915, he lived in “Thomson’s shack”, seen at right in its current setting at the McMichael Gallery.

Thomson’s work over these years developed quickly. He applied paint thickly, in an impressionistic style. At times he experimented with pointilism and abstraction. He became fascinated with recording the changes in the landscape over the course of the year; in 1917 he painted 62 daily sketches recording the progress of spring. Some of his work is shown below, in roughly chronological order.

Sunset, Algonquin Park 1914 (oil study on board):

The 1915 painting Northern River is below on the right; the gouache study for it is on the left. Looking through trees across water to the far shore is a frequent motif for Thomson.

Thaw in the Woods (study on board) 1916

Thomson’s two most famous works were both painted in the winter of 1916-17. The West Wind is now in the Art Gallery of Ontario in Toronto, and The Jack Pine is in the National Gallery in Ottawa. The sketch for The Jack Pine is shown beside it. It’s interesting to compare the two; the energy of the sketch gives way to a stylized calmness in the painting.

Finally, this sketch, Early Spring, was done in 1917, only a few months before Thomson died:

More about Tom Thomson:

• Thomson works at the National Gallery
• Thomson works at the Art Gallery of Ontario (the AGO’s collection isn’t searchable, but some of his works are included in this page of highlights of the Canadian collection.)
• Tom Thomson Art Gallery
• The McMichael Collection
• Wikipedia