Painting Portraits With Values: How to Get a Full Value Range

@thedude posted a portrait study in the Rendering category of the Artwod Community Hub with a simple question: am I on track? They were working specifically on values, which is one of those things you can practise for months and still not know whether you are improving (we have all been there).

Three artists answered, and between them they covered the whole problem. Let's see what the Community breakdown teaches us.

@sig looked at the study and made the key observation: the values in it are accurate, and the painting only uses a narrow slice of the range available. Everything sits in the middle. The darks are not dark enough and the whole image reads as soft and flat.

This happens often with portraits painted on a white background. A white background quietly sets your ceiling. With the brightest thing in the picture sitting behind the head, you unconsciously keep everything else lighter to match, and the shadows never get the depth they need. Not to mention, white backgrounds are not great for your eyesight if you're working digitally, so keep that in mind.

@sig’s advice was to treat the background as part of the value structure rather than as empty space. Give it a value, and it immediately will start doing the work for you. A darker background pushes a light edge of the face forward. A lighter one lets a shadowed jaw sink back. You get depth from the relationship between the figure and what is behind it.

In the paintover, @sig darkened the whole image and deepened the shadows. The result is the same drawing with a far wider range, and it reads as solid rather than washed out.

If you are painting on a white canvas right now, fill the background with a mid grey before you do anything else. It costs you one click and it changes every value judgement you make afterwards.

@rayka07 added the process, which is worth following from the beginning of a painting rather than trying to rescue one at the end.

Start by analysing the reference and splitting it into light and shadow. Two values, nothing else. Where does the light stop? That line across the form is the single most important decision in the painting, and getting it right early means everything that follows has somewhere to sit.

Once the split is clear, develop the shadow side. Core shadow, bounce light, ambient occlusion and cast shadows all live in there, and they are what stop a shadow from reading as a flat grey shape. Then move into the light and place your halftones and highlights. The background sits outside that split. The light and shadow division describes the form, while the background is a separate value you choose. That is why toning it first helps so much: with a mid grey behind the head, you have something to measure both sides against from the first stroke.  If you have trouble with 2 values, check out our Intro to Painting. It will definitely help.

@plupp13255421’s paintover went after readability. 

This is the fix for when a painting starts to feel muddy: muddiness is rarely about colour. It usually means the shapes have broken down, and the eye cannot tell where one form ends and the next begins. Merging those small steps into a single confident shape brings the structure back immediately.

The same thinking went into the hair, which got minimal brushwork, softened edges and skin showing through in places. Hair painted strand by strand turns into a helmet. Hair treated as one soft mass, with the face emerging from it, sits on a head.

Before you render anything, ask what the big shapes are. If you cannot name them, no amount of blending will save the painting.

Every halftone you remove is a deliberate simplification. So the number of halftones you use is what sits between a graphic image and a photographic one. Few values and few halftones give you a bold, poster-like painting with strong design. Many values with smooth transitions give you realism.

Neither one is the advanced version of the other. A two-value painting demands that you know exactly where the light stops, with no gradation to hide a vague decision, and that is a harder call than blending your way across a cheek. What changes is what you are making, and the choice is yours.

Decide which you want before you start, because a painting that drifts halfway between the two reads as unfinished rather than as a style. Try this free Halftones Art Workout - you will understand this concept better with practice. 

What is a value range in painting?

The distance between the darkest dark and the lightest light in your picture. A wide range gives depth and impact, and a narrow one reads soft and washed out. Most flat paintings have accurate values crowded into the middle of the range.

Should I paint a portrait on a white background?

Tone the canvas with a mid value first. A white background caps how dark you are willing to go everywhere else, and it gives you nothing to judge your lights and shadows against. Our Intro to Painting covers setting a painting up properly from the start.

Why does my painting look muddy?

Muddiness is almost always a shape problem. When the values inside a shadow break into many small steps, the eye cannot tell where one form ends and the next begins. Merge them into one clear shape and the structure comes back.

How many values should I use in a portrait?

As many as the look you want. Few values give you a bold, graphic image, and many values with smooth transitions give you realism. Decide before you start, since a painting that lands halfway reads as unfinished. The free Halftones Art Workout is the fastest way to feel the difference.

How do I know where the light stops?

That edge follows the planes of the head, so it is a drawing question before it is a painting one. Our Intro to Portraits road goes through the planes of the face in detail.

Where can I get feedback on a value study?

Upload it to the Artwod Community Hub and say what you were practising, as @thedude did. Naming the skill gets you answers about that skill rather than general comments.