This post is the culmination of an exercise in which I copied a sargent portrait drawing. The first postwas primarily concerned with laying out and measuring, picking out some anchor points and getting them established. Thesecond post continued with the laying out in two dimensions, establishing the main shapes. In thethird post I developed the two dimensionsal scaffolding for the drawing into three dimensions, building a conceptualthree dimensional wire frame of the main forms. In this post, I’m adding value, light and dark, and finishing the drawing.
The value of a plane is primarily determined by two things:
- Its local value. Whether it’s light or dark in itself, e.g. a black or a white cube.
- The amount of light falling on it. A white plane in full shadow can appear darker than a black plane in full light.
The surface quality of the plane and how much light it reflects will also have a bearing. A shiny surface will reflect more lightthan a matt one. But for the purpose of portraits, that’s a largely a second order effect.
Now I’m at the stage of adding value tomy copy of Sargent’s portrait of Lady Sarah Spencer, I need to decide how to proceed.
At this stage quite a bit of time has gone into this drawing already, although it doesn’t look like much. What I should have nowthough, is a fairly solid foundation on which to build. It’s kind of like decorating: The preparation takes most of the time. If youdon’t do it right, the last stage is much more difficult, takes longer, and doesn’t look as good when it’s done.
I must confess to a certain amount of nerves at this point. What if I screw it up after all that work laying out? And where doI start? Since I’m concentrating on the form here, I’ve decided to start by ignoring local values and to concentrate on the effect of thelight on the planes of Lady Sarah Spencer’s head. So I’m going to ignore the dark local of her hair, and of her eyes, and thebackground. To begin with,I’m going to use value to start to define the form more clearly.
Broadly speaking, shadows can be thought of as belonging to two distinct types:
Cast Shadow: A shadow cast by one form onto another.
Edge Shadow: A shadow created as a form turns away from the light.
Here, I’ve started with a cast shadow, the shadow cast by Lady Spencer’s nose onto her right cheek. It seemed like a decentplace to start to me.
What this shadow does is state very clearly the direction of the light. It’s pretty obvious from the direction of theshadow that the light source is slightly above, to the right, and slightly forward of Lady Spencer’s head.
Here I’ve started adding edge shadow, shadows formed as the planes turn away from the light. I’m trying to keep the main formsand the direction of the light upper-most in my mind as I do this.
The diagonal light on the egg-shape of the head is producing more shadow on the left (as we look at it)and towards the bottom of the head. Within the shadow side of the overall egg shape I’m thinking about three main forms: The ballof the eye in its socket, the ball of the cheek, and the vertical barrel of the mouth.
Of course, I’m referring constantly to the original drawing to show me where the shadows will lie, and how relatively dark andlight they are. I’m still doing a copy. But I’m trying to keep an appreciation of the forms that the shadows describe as I work,I’m trying to feel the form. The Plasticine model I made earlier of the eye and nose area is really helping me now. Andsince I’ve already established the relative positions of things on the head, I don’t have to worry about positioning at this stage.I can just think about how the light is creating shadows on the forms.
Moving over to the other side of the head, the ball of her left eye is like a microcosm of the head as a whole, with shadowsfalling on the left and underneath. There’s a deep shadow under the chin, where very little light is falling. Her left cheek is alsofalling away from the light a little, but more light on this side of her face means a much lighter shadow.
As subtle as that shadow on her left cheek is, it describes oneof the main planes of the head and is crucial to describing the overall form. At this stage I can seethe main forms of the head established much more clearly in three dimensions.
It’s perhaps worth mentioning again here that I haven’t addressed any local value yet. All the areas that I’ve introduced value to are ofroughly the same local value, the local value of Lady Spencer’s skin. Since I’m pretty happy that the form of the head is established,and I’m over my nerves and well into the value stage (and also since she’s looking a little odd with no eyes) I think it’s time to startintroducing some local value.
The first thing I’ve done here is quickly scribble in some background tone and fill in an overall value for her hair. The changein the feel of the drawing is pronounced though. Adding local value into the hair in particular has given us something to relate the value ofthe skin to. Adding the background tone has created the effect of space around her head.
However, it’s not all good news because I’ve also introduced an imbalance in the values which is making the drawing read oddly.The first area of value I put in, the shadow cast by her nose, should reallybe as dark as her hair. There’s no light falling in that cast shadow, so it should be as dark as the darkest parts of the drawing.
I’ve made work for myself here since I’m going to have to readjust all the values on her face now. It’s perhaps not such a big thing, I canadjust as I go along without too much trouble. But, looking back through my process now, it probably would have been a good idea toestablish my full value range right at the start by putting in at least one area of full dark as a guide, at the lowest value Iintend to go to. If I’d made that first value the cast shadow of the nose, as dark as I could get it, then everything else would have beenadjusted in relation to it and I’d have used my full range right the way through.
To an extent, I’m translating the values of the original into a more narrow range since I’m drawing with pencil and the original is aprint of a charcoal drawing, with a wider value range than I can achieve. That process of translation would have been easier to do I think ifI’d established my full range right at the start.
The value imbalance can be seen more clearly now I think. I’ve already found that I’m having to darken the background value to the leftof her face to bring it into line with the value of her hair.
Although I’ve put some value into the eyes to make her look a little less like a zombie, the eyes are still lighter than my darkestdark (especially her right eye) so they look a little odd, they don’t quite ‘live.’ They just don’t read right in her face yet.
More problematic perhaps is that the planes of her face, which looked pretty good earlier, now look distorted, the shapes aren’t readingright. I notice it particularly around her mouth, it looks like her mouth area and chin are jutting forwards a little.
Two things are contributing to this distortion. Firstly, I’ve drawn a line running down diagonally right to left from just beneath herleft nostril to her top lip. I added that line when I started to add value to that area at the second stage I’ve posted here. It’s amistake. It has the effect of making her mouth read as if it is protruding. In the original, it recedes. That line justisn’t there in the original.
Herein lies one of the problems of working with schematic models, like the planes of the head model. I think what’s happened here is thatI’ve imposed my internal schematic head model onto the drawing, rather than drawing what’s actually there. It’s true that my schematic model is helpingme in many ways. I certainly wouldn’t dismiss it because of problems like this. But what it actually represents, perhaps, is just a more detailedschematic to override the very simple ’round ball with two dots for eyes and a line for a mouth’ head schematic that most people have.For me, the verdict is still at least a little out on this. I think the crux is that the schematic head modelshould be used as a guide to help us interpret what we see, and to help us look for the right things. The danger is that we unconsciouslyimpose it upon what see, we forget to look at what’s there. It’s probably a good idea to bear this potential pitfallin mind when working with plane models. The planes of the head model is too useful to throw out entirely, but perhaps has it’s owndrawbacks.
Apart from the direction of the plane, the other problem I have in the area around her mouth is that the shadow value is too lightthere in relation to the rest of the drawing. This gives the effect of more light falling on the shadowside of her chin, which could only happen if it was jutting forwards and turned slightly to the light. We read values in relation to eachother, never as absolutes. The value imbalance together with the mistake in the direction of the plane isbeginning to destroy the form.
This shows quite clearly, I think, why relative, not absolute value is so important. It doesn’t matter a damn what absolute value you give to agiven area of a drawing, but it must relate coherently to the whole. You can draw (or paint) in a light or a dark key, with a narrow or awide range, but the relationships between the values must read right or the piece won’t work.
I’m going to get into trouble now if I’m not careful. I can see that something is off so I think it’s time to stand back and comparethe drawing with the original.
Once again I’m reminded of one of the most useful features of working sight size: Immediate, like-for-like comparison of the copy with theoriginal. The effect is so powerful that it even shows up in small images on a web site like this. Try it: Start by looking at the mouth areaon the original, then flick your eyes quickly, backwards andforwards, between the original on the left and the copy on the right. If you see what I see, the effect will be of her mouth area appearingto push forward in the copy, to become convex instead of concave. The whole of her face appears slightly flattened to me too, due to theshadows on the right side of her face not being deep enough to push it back into the background.
I’ve got to get my form and value relationships right now if this copy is going to work. Fixing the direction of the form is easy nowI’ve seen it. To fix the value imbalance, I’ve got a choice here I think. I could lighten mydarkest values to the same values as the shadows on her face, or I could darken the shadows on her face down to my darkest dark. Since I’ve gota lot of range left in my pencils that I’m not using (I’ve ben a bit tenuous with them so far) I’m going to go darker.
After a bit of work, I’ve made some improvements to the value relationships and fixed the area between her nose and mouth to an extent.Try the eye flick test now, her mouth has moved back into her head but not far enough yet. But already the forms are starting to read morelike the forms in the original drawing. My value range is still much morenarrow than the original, but the relationships are improving. At some point around this stage, I decided to go as dark as my pencils wouldlet me and use my full range. Of course, that means I’m going to have to continue tweaking my value relationships across the entiredrawing as I go on, but nonetheless I fearlessly whip out my 6B graphite pencil and dive in.
That’s a bit more like it. I still have some mouth problems, mostly value related, but the eyes and overall shape of the head arelooking fairly good to me now. As my copygets closer to the original, as it becomes more finished, drawing largely becomes a case of looking for inaccuracies and correcting them. I’mclose enough to the original now to know that it’s going to be a fairly easy down hill run from here. The hard work is done and I cancontinue tweaking and correcting as long as I feel like it.
At this point though, the main focus of this exercise has really been completed. I’ve practised the three main points I mentioned back in thefirst post of this series: Accurate judging of size and shape in two dimensions; appreciation of the form inthree dimensions; adding light in the form of value relationships. I did work on this drawing a little more though, for enjoyment’s sake as muchas anything else. Here it is next to the original on my easel, at the point at which I left it:
It’s not an accurate copy by any means, but it was never my intention to make a mark for mark copy. Along the way I’ve learned a little about theanatomy of the eye, I’ve done a bit of modelling in Plasticine and I’ve clarified for myself a little more the difference between accurate judgingof shape and position and conceptualising form in three dimensions. This is just a small step along a much longer road. ButI do think I’ve picked up some useful things during this exercise, especially when taken together with my other Sargentform based and more complete value studies.
This part of the journey really started when I began to look into form drawing and in particular the reductionof the head down to its main planes as recommended by Loomis in Drawing the Head and Hands, almost two years ago now. Perhaps themost valuable lesson from this particular tangent has been what a difference physically modelling forms in three dimensions has made to mydrawing. I conceptualise things differently now when I’m drawing. I see value as much more an interaction between planes and light than Idid. Most importantly perhaps, I think differently when I’m drawing or painting. I’m thinking about the ‘why’ as much as the ‘what.’
I do believe now that deepening our appreciation of form and keeping it uppermost in our minds as we draw can have a subtle but quiteprofound effect on the end result. Although I was working from a two dimensional model here, I chose the Sargent drawing becausehis form is so strong and clear. Somewhere in the process of translating three dimensions into two, I think he’s done something more thancopy the optical effect, something more than translate with an ‘innocent eye.’ Much more. I believe he’s simplified and emphasised,the better to deepen the illusion of three dimensions.
Sargent has already done the really hard work for me. He’s already picked out the salient planes, the ones that describe the form of thisparticular head most clearly. He’s done all the hard work of setting up the light and balancing the values. That’s what makes this drawingsuch a great model for this exercise.
I think it would be a big mistake to try and do the same thing using a photograph as source material. Whilst itwould still be a two dimensional reference, photographs have the effect of flattening the subject. I think it’s something to do with theway they average out the values. I’d strongly advise against working from photographs at all if you’re interested in deepening yourappreciation of form. You can practice two dimensional judging of shape fine with a photograph. But one of the hallmarks of work done fromphotographs, I believe, is a flattening of the form. I’ve seen this too often for it to be coincidence, and it’s evident even inotherwise accomplished work I think. Quite often the subject will have a curious ‘cut out’ appearance against the background. The better theartist becomes at copying the photograph the more pronounced the effect will be. Photo realist work is particularly lacking a feeling ofdepth and form, I find. The illusion of depth has to be created, it doesn’t just happen.
I would argue that it’s much more productive to put in the hard work to learn how to translate three dimensions into two, and to learn fromhow previous masters of form have done it, than to take what is, let’s face it, the convenient – I would even say the lazy – route. You won’tdo yourself any favours. Artists who work well and convincingly from photos (and they do exist, it would be stupid to deny it) are, withoutexception, perfectly capable of doing the same from life. If you’re learning like me, don’t hamstring your development by taking the easyroute. Learn what defines form and how to translate it from three into two dimensions. Learn to feel the form. If my experience so faris anything to go by, your work will be the stronger for it.
For sake of completeness, I’ll finish with a bigger photo of the copy on it’s own:
It’s obvious to me that my copy has little of the subtlety and clarity of Sargent’s original. But it has taught me something about renderingform which I can use whether I’m working from a flat copy or from life. For my next post, I’m going to show how I went about trying to transfersome of what I’ve learned from my planes and form practice into a sight size self portrait. The process was essentially the same, and I dobelieve it produced a piece of drawing considerably advanced from the self portraits I was turning out before all this practice with planesand form.
The Keys to Colour - Free 6 step email course
Learn how to:
- mix any colour accurately
- see the value of colours
- lighten or darken a colour without messing it up
- paint with subtle, natural colour