eye level

Deepest apologies to Sheridan Smith.

I was sketching along to one of Court Jones’ caricature lessons over on Proko.com using a reference photo which had good lighting and strong features — just what you need to practice caricature. It was only after I’d finished drawing that I noticed the eye positions. I’ve had this problem before, and I frequently use the ‘fresh eyes’ approach to take a break and see any inaccuracies in the drawing. But I definitely have an inability to see how the complete picture is working as a whole as I’m drawing all the smaller elements. Maybe there’s a term for this — ‘drawing blindness’, perhaps.

One solution to this is to build up the portrait systematically using a basic construction of the big shapes before going to the smaller forms within them, and then, finally, move on to the details. See how Cesar Santos puts in the construction lines at the beginning of this painting demo.

It also helps to be aware of the many ways in which a portrait can go wrong. Sometimes it’s not a big mistake, like an eye in completely the wrong position, but an accumulation of smaller errors which might seem insignificant in themselves but which add up to a face which somehow just doesn’t look right.

Here are a few examples of the ways a portrait can miss the mark, taken from the pile of charcoal quicksketches on the endless computer paper.

The perspective lines for the eyes don’t match the rest of the features, making the far eye look too small.

Often the far eye in a portrait does appear smaller as it’s farther away, but the rest of the face needs to match, and any parallel construction lines need to share the same vanishing point. The difference in size in the features seen in three-quarter view will depend on how close you are to the subject — if you’re up close, there will be a more noticeable difference.

If the axis lines for the features and the centreline of the face don’t match, the face looks skewed. This can make an eye look too high, or a mouth slanted, or just give the sense that the features don’t quite hang together.

If the face is holding an expression which isn’t symmetrical (or the features themselves aren’t symmetrical) then axis lines can only provide a starting point. Is the mouth shifted to the side because of the expression (blue face) or because of inaccurate drawing (red face)?

Is Stan smiling with a sideways smile? Hard to tell when you can’t see both sides of the face. It’s often helpful to draw construction lines through the form even when you can’t see the hidden side, as though the head is made out of glass.

Stan Prokopenko has made a number of really useful videos on head construction. Here’s one which uses an intuitive approach to constructing the basic head structures.

An off-centre view combined with an asymmetrical expression can make centrelines hard to find. The corrected version in the animation above shows the more accurate centreline as taken from the reference photo. That said, the drawing has other elongations and distortions which confuse the final result and make the face appear slightly flattened.

There are so many factors which will influence the final appearance a portrait and so many ways to get there: Loomis methods, Reilly methods, Asaro heads, skull anatomy, caricature, tonal block-ins. (See Jeff Watts demonstrating some of the methods in this video.) I think there’s something valuable to be taken from all of them, and the aim is to have them all available as tools that can be used intuitively.


stumped and shaded

Experimenting with textures and effects using charcoal on Bristol board to show a view in Bath Abbey Cemetery, trying to capture the contrast between the areas bleached by sunlight and the gloom under the shade of the yew trees.

I wanted the shade under the trees to be deep black to get the maximum contrast with the paper white. I tried graphite on a slightly textured heavyweight cartridge paper but ended up with a grey shine on the darkest areas. The delicacy of graphite combined with the deep darks of charcoal would be ideal, but in areas of darker shading I found the soft graphite fills the grain of the paper and forms a plate-like barrier to the charcoal and the vine stick wouldn’t grab, as you can see in the trunk of the tree below.

I made another attempt on Bristol board (picture at top of the post). The smooth surface of Bristol isn’t usually the first choice for charcoal which needs some tooth to grab the grains, but it did take the charcoal, even if much of it turned to dust on the surface which occasionally had to be brushed or blown away. When blended with a paper stump the vine charcoal turned a light grey and almost has the appearance of an ink wash.

A kneaded eraser was useful for lightening areas and seemed to work more effectively than a hard pen eraser. However, I built up many of the leaf shapes and branch effects by adding extra shading to the negative spaces rather than picking out the lighter areas with an eraser. The stump would lighten the tone and flatten any texture, and then the dark background shapes could be drawn in to this grey midtone. This drawing in reverse allowed multiple layers of depth to be built up and was much more precise (and satisfying) than using an eraser.

The white of the Bristol paper appears cold, almost blue, compared to some papers, but in reality it’s probably closer to a neutral white and its coldness only appears in contrast to warmer surroundings, such as a matboard or the newsprint backing paper.

Other papers will have different handling with different results. Here are some sketches using graphite or charcoal (Mary in graphite) on a sheet of Fabriano Disegno drawing paper which has a texture laid in to the surface. The graphite and charcoal catch on the paper grain and sit on the top if lightly applied but blending with a soft brush or a stump can achieve a more even tone, as seen in the face and jacket of the figure on the right.

Zoom in or click the image to see the paper grain.


Seago copy

A copy in oils on panel of a painting by Edward Seago.

Seago was a prolific painter of landscapes. Stapleton Kearns has written extensively about his methods on his blog, which is well worth a read.

I started with a charcoal sketch to get the feel of the composition. I think the composition is instantly appealing, with those unusual tree angles and the light-dark balance. The eye is naturally drawn around the scene and off towards the distant trees and sky.

Seago’s style is very loose and painterly. He painted very quickly with brushy, textural strokes, and it was important to remember this when making this copy. It’s a style which can’t be reproduced by fussing around or tentative licking with the brush. This was especially true for the foreground snow where the textures of the grasses could only be achieved by a confident, definite stroke. A palette knife was useful for many of the effects.

I let the first pass of paint dry before drawing the trees over the top, though Seago probably did this wet-into-wet.

I might have oversaturated the colours in this copy. Seago’s colours are more muted and greyed down in the original painting, though two separate online references have different colour casts. Another painting of a similar scene [The reed bed — Winter] has different colours again.

These master copies really are useful for trying out different approaches to painting. The different styles force you in to new ways of seeing and translating. And while examining the surface of the painting can reveal valuable insights into the artist’s methods, so much more can be learned from ‘acting out’ a bravura brush stroke.


after Orazio

A not-too-slavish copy of Head of a Woman by Orazio Gentileschi. This copy is about a third the size of the original, and was painted with a limited palette of just ultramarine blue, transparent oxide red and titanium white. I first heard about transparent oxide red (PR101) from Terry Miura who uses it instead of umbers, and I think it’s becoming one of my favourite staple paints, not least because it mixes well with ultramarine to create a surprising range of flesh tones. The only downside is that it’s so tempting to leave the glowing orange when really I should be muting the colour down.

I haven’t done an overlay check of proportions with this one as that wasn’t really the exercise (though I did play around in Photoshop with a photo of the half-completed painting to identify an overly bulging eye). The challenge was to see if I could capture anything of that arresting and inquisitive gaze.


from life and imagination

This started out as an afternoon painting session in the garden. It was a sunny day, I wanted to paint in oils — and I didn’t want to rely on another photo. So I set up a small box easel and began to paint a self-seeded poppy growing out of a plastic pot. There was only one chance to do this as after a few hours the last of the petals fell off.

At the end of the session I wasn’t particularly impressed by what I’d done. So I experimented with thumbnail acrylic sketches and digital overlays to see if I could transform what I had into something more interesting.

The result (top picture) looks somewhat Victorian, and the short-lived suburban poppy has been hammed up with the drama of the fading light into a vanitas. (At least, that’s the story I’m telling myself.)



Apologies for making this post flash like the Blackpool illuminations, but I’m using a modern version of a technique used since olden times to check accuracy and proportions. Without a master draughtsman looking over my shoulder to tell me my ears are in the wrong place I’m reliant on overlaying my drawings on top of the original photo reference to see where I went wrong. Leonardo made practice drawings on glass plates to compare to the original, but I’m pretty sure he would have made use of today’s digital tools.

Usually, I take a photo of the sketch into Photoshop and simply add it as a semi-transparent layer above the original photo. Then it’s a matter of trying to align the sketch and the reference by rotating and scaling until (hopefully) the features line up. By quickly switching the layer on and off, you can compare the two.

The technique illustrated in the photos above is a little more complex as it uses a number of filters to find the edges (shown in red or white) of the features in the original photo. Usually, I don’t go to all that trouble when checking a sketch and instead use the transparent layer technique mentioned above.

By doing this regularly I have begun to identify some of my repeated mistakes: I make the backs of heads too small, brows are too high on the skull, everything is made too wide, chins tend to disappear, hands are too short.

Once these weaknesses have been identified they can be consciously corrected. I’ve found the quickest way to improve is to concentrate my practice on the things I find the hardest to draw.

[Models courtesy of New Masters Academy.]


hand study copy

These are copies painted in oil in a sketchbook based on hand studies by Bouguereau. The thick sketchbook paper will take oil paint if first covered with a layer or two of acrylic gesso. Any leftover paint can be spread on top of this to create an oil ground for the next sketch, but it’s not necessary. This was another way to use the intimidating Moleskine, even though it seems a bit odd to put oil paint in a sketchbook.

As ever, the challenges were to match the values and hues while at the same time trying to model three-dimensional form, all by using buttery oil paint. The hues tended to be somewhere around orange—slightly more red or purple in some areas, more yellow in others—but all of of the hues were quite greyed-out or desaturated (low chroma). Bouguereau himself somehow managed to make flesh tones look natural and almost translucent (see this link for a discussion on this topic) yet his colours are not a dead grey.

On the dimly lit shelf, my version looked a lot more smooth and natural than when brought out into the light. Bright light shows how rough and sketchy the paint blending really is, as you can see if you zoom in on the photos above. Studio lighting is something I really need to sort out, as painting in the dark often fools me into thinking I’ve done a better job than I have.

I noticed that too, when painting the puppy portrait. The painting viewed in natural indoor light on an overcast day looks very different to the same painting in direct sunlight, as you’d expect, but it’s not just the colours that become more saturated, the edges become harder and more noticeable.

Click for larger image.

Anyway, it’s all good practice. To see some extraordinary oil paintings in sketchbooks, have a look at the sketcbook tour videos by Cesar Santos.


space saver

A typical mid-session mess.

For various reasons, I usually paint in the second smallest room in the house. My ‘studio’ working surfaces are a few book shelves (covered, predictably, with books) and a small table which is home to a computer, modem, keyboard, mouse mat, graphics tablet and a cup of tea. In short, space is short.

Out of necessity I’ve found various ways to cram the messy paraphanaelia of oil painting into as little space as possible. There’s enough space on the table for a glass picture frame (my palette) which when not in use can be balanced precariously on a section of my CD collection (the bad ones).

I’ve found that using glass as a palette is the easiest way to mix paint and the easiest to clean. I turned the lining cardboard to the front to create a neutral midtone. At the end of the session any extra paint is gathered up with a palette knife and transferred to the bespoke paint stick which was cut to fit in the frame. Large amounts of unsullied tube paint can be kept in separate blobs, but everything else is mixed up into a pile of mud. The stick is then wrapped in junk-mail clingfilm.

After the bulk of the extra paint has been saved, a few sheets of kitchen towel and a small blob of hand soap will complete the clean-up, no solvents necessary. In fact, I rarely use solvents for any part of the painting process, so even in this small room there is only the faint smell of linseed oil.

I don’t bother cleaning the brushes. Instead I get the worst of the paint out on a kitchen towel then dip the end of the brush in a mixture of safflower oil and a dozen drops of clove oil. This combination is very slow drying and will keep the brushes supple and in good condition for at least a couple of weeks before needing another dip. At the next painting session, just wipe off the excess brush dip and carry on painting.

In this small space, the easel is put up wherever it will fit. The lighting… well, that’s a problem I haven’t really solved yet. I want to get a consistent soft-box setup so I can draw the curtains on the sun beaming uncontrollably through the west-facing window. But that will be the subject of another post…


pet portraits

A friend wondered if I could do a portrait of her dog. I wondered too, but it sounded like an interesting project. I didn’t realise that pet portraits were a thing until I went online and saw page after page of search results for pet portrait artists.

I’ve never met Cooper, a boxer mastiff cross. Everything was based on a collection of photos taken at different angles, including profile shots and various poses, that let me get some idea of the basic shape of his head.

Working in charcoal seemed the most flexible and natural method for me at the moment, but I really wanted to do an oil painting, just to see what I would come up with. I ended up doing two pieces, one in charcoal of Cooper as he is now (which didn’t quite work—I need to make another attempt), and this oil painting of him as a puppy.

For the puppy painting, I started with some warm-up charcoal and pencil drawings, then used a Polychromos pencil to do a more detailed sketch. I like the Polchromos pencils as they build up gently into darker tones, even though they don’t erase as easily as charcoal.

The pencil sketch was scanned and copied on to office paper and the back covered with old charcoal dust that I keep in a jar. The outlines and tones were transferred onto an A4 canvas mounted on aluminium composite by drawing over the sketch with a ballpoint pen.

The canvas ground was acrylic gesso covered with a layer of grey mud oil paint left over from a previous painting. Acrylic gesso can be very absorbant and suck the oil out of the upper layers, but this mud preparation acts as a barrier to the thirsty gesso and effectively creates an oil ground that is very pleasant to paint on.

I did about three passes on separate days to build up the painting, with a final pass to make some corrections to the background. I let the painting become touch dry before the next painting session.

I used ivory black in this painting and it’s not a pigment that I’m used to. I bought a tube to mix with titanium white and burnt umber to make neutral greys for reducing the chroma of colour mixes (following Paul Foxton’s method), but it was tempting to use this black instead of my usual mix of ultramarine blue and burnt umber for some of the darkest areas. A problem I’ve found with ivory black is that it sinks in really noticeably. It looks like a deep, silky black when wet but dries a lighter matt, almost like gouache. I don’t know if the oil is sinking in to lower layers or if the pigment itself is soaking up all the oil.

I oiled out the painting at the beginning of each pass, but eventually that thin linseed oil layer also sinks in. I was betting everything on a final varnish using Gamvar Gloss to even things out, which it seems to have done.

He looks cute in this picture. He’s grown up to be the size of a small horse.


sinking in

To varnish or not to varnish. Some painters don’t like the glare and shine of gloss varnish, preferring a matt or satin finish, or not varnishing at all. There are matt varnishes which contain a small amount of wax to take away some of the glare, but for paintings with dark areas this can seem like the worst of all worlds: there is still a sheen of glare (though not as much as with gloss) but the deep tones become cloudy due to the wax. Maybe it works well with high-key paintings.

Varnish is there to protect the surface (it’s a removable, replaceable clear layer) but it also brings out the depth and vibrancy of the original wet paint. It also evens out any patchy, sunken-in areas, which is where I find it most useful. Some pigments sink in more than others (ivory black and burnt umber seem to soak up all the oil) and mediums added to the mix can alter the final gloss. Thick impasto can have a glossy sheen next to paint thinly applied with a brush.

It’s a minefield of confusion which I attempted to clear with systematic testing, but like many of these experiments I ended up throwing various combinations of paint and varnish on the canvas in the vague hope that something useful would come out of it.

Here’s a painting of a rose (this was originally going to be a post about painting a rose from life, with the various challenges that presented) with a background that contains large areas of ivory black. It looked deep and dark when wet, but as you can see from the photo, it sinks in really noticeably to a dull matt finish. So much of the oil is soaked up that the paint doesn’t even feel particularly well bound to the canvas, as you can see from the top-right edge where the paint was lifted during oiling out.

I was going to go in to detail about all the various combinations of paint and the effects of thickness of application, oiling out, layering, and so on. But you’re spared all that and here’s a picture instead.

Click for a larger image.

The conclusion I reached: just apply a layer of gloss varnish at the end and all the problems are solved. Except for one: whilst sunken-in areas regain their depth, any areas that are already very glossy become even more glossy, so there’s still some unevenness. The portrait at the top of this post had glossy areas where I’d used stand oil in the paint mix. Adjacent areas were relatively matt. The gloss varnish evened things out a little but not completely. I tried applying a number of coats but they all seemed to dry with the same result. In the end, in a spirit of experimention and desparation, I poured a pool of varnish over the non-glossy areas, hoping for an evenly-glossy final painting. It didn’t work, but I quite like the photo.

Things to try in future experiments: lightly abrading the very glossy areas to take down the sheen before applying the final varnish.

Here’s Mark Carder with a video on how to varnish a painting. However, if you’re using Gamvar (as I did) the application needs to be thinner. See here for Gamblin’s varnishing instructions.