“Paint what you see, not what you know”.
I’m sure you’ve heard that before.
And most of the time, it’s pretty good advice. But it’s not the whole picture. Why? Because you can’t always trust what you see.
What we see is a result of what our perceptual systems show us, processed through our previous experience and our knowledge. It’s filtered. What we see is affected by what we know. You can’t turn that off. It’s the reason visual illusions like this work.
So what we know can cause us to interpret what we see incorrectly.
Thankfully, the converse is true: what we know can help us to interpret what we see better. It can help us be fooled less easily – as long as we’ve taken the time to gather some better understanding.
A good example of this is modelling factors
You may have come across the modeling factors of form before. It’s a very useful bit of knowledge that helps with making objects in drawings or paintings look three dimensional.
You may have seen diagram like this with more factors. I’ve missed out the ones to do with the cast shadow and also the terminator – the point where the half tone turns into shadow. Why I’ve done that will become clear shortly.
But you get the idea. The modelling factors describe the way value changes across an object depending on how the light falls on it. And they’re very useful to know because they can help you interpret what you see. They can help you make more realistic drawings and paintings.
However, as shown here they describe value only. They don’t tell you anything about how colour changes across a form.
How to model form with colour
If you’ve come across Munsell before, you’re probably aware of the three aspects of colour – hue, value and chroma.
If that’s new to you, or you want a quick refresher, click here and I’ll email you a link to a video with a quick overview.
OK, good. So we know what Munsell is. But how does that help us when we’re at the easel?
Well, let’s assume we have a sphere of a single local colour, sitting in a mid-grey shadow box.
That’s a Munsell chip I’m holding there, of exactly the same local colour as the sphere. I know because I mixed up some paint to exactly match the colour of the Munsell chip, and then painted the sphere with it.
So I know exactly what hue, value and chroma the local colour of that sphere is. That’s going to be very important. You’ll see why shortly.
By using a simple colour checker, along with the Munsell chips, we can isolate the colours we see for each modelling factor, like this:
In the picture above, the colour checker is showing me the colour of the reflected light in the shadow. Now I can mix it exactly.
I can use the same method to nail down the five colours I need for each modelling factors of our blue sphere. They’re all (or almost all) on this page of the Munsell big book:
And here they are, with their respective modelling factors:
All those chips are the same hue, they only vary in chroma and value. The one exception here is the colour of the core shadow, which was slightly warmer – the hue moving around the Munsell hue wheel towards green a little. That’s the only hue change in all these colours.
And that’s worth noting. Almost all the colours across the surface of this sphere are of the same hue. That’s why I’m using Munsell for this, and why it’s so useful that I know the exact hue of the local colour of the sphere (5PB, or 5 Purple Blue, in Munsell terms).
Because that’s how I know that the hue is staying the same across most of the modelling factors of this sphere. The Munsell chips tell me. They’re helping me to understand how the colours change from light to shadow.
The value changes the most, the chroma changes quite a bit (dropping in the highlight and in the shadow) and the hue hardly changes at all. So if you want to get better at painting stuff that looks real, don’t obsess over changes in hue. Look very closely at value and chroma.
Here are those five colours mixed and ready on the palette:
Now I can paint the sphere 🙂
Painting the Sphere
This is actually a pretty simple process, once the colours have been nailed down and I know, from the modelling factors, where they need to go.
In fact, the hardest part of painting has been done already – judging and mixing the main colours. I just pulled that part out and did it separately, instead of on the fly as I’m painting.
Step 1: Background
First, I paint the background, ground and cast shadow with Munsell neutrals.
Step 2: Core shadow and reflected light
I just block these in, without any blending.
Step 3: Half tone, centre light and highlight
Again, these are just blocked in, without any blending or fancy business. I usually paint the half tones and light more thickly and with more texture. It helps create the illusion of form.
Step 4: Blending and adjusting
I spend quite a bit of time at this stage, observing and adjusting. Although the process up to this point might seem formulaic, you still need to paint when you get to this part. The more developed your observational and painting skills, the better the final result.
Here’s a direct comparison of the actual sphere and my study, pretty much from the viewpoint I painted it from.
How this works in the real world
The good thing about this is that the understanding you gain from this kind of practice can be translated directly to real world objects.
People often struggle with yellows, especially in the shadows. So I used exactly this procedure to paint a study of a lemon.
Now, lemons are very high chroma, so they’re more difficult. But with an approach like this, you can find out exactly what colours you need, just the same.
Here’s the palette, with the tube colours I used to mix all of those colours – cadmium yellow, cadmium orange and burnt umber.
A few things worth mentioning about this:
- The core shadow moved slightly around the hue wheel, as with the blue sphere. It got slightly warmer, which is why there are two Munsell chips there – it was between the two.
- All the other chips are the same hue
- I used two colours for the half tones, to help with the modelling
- For the centre light, the challenge was to keep the chroma high enough. I used cadmium yellow, with a small amount of cadmium orange mixed in to change the hue, but otherwise unmodified. And I painted it thickly and then made sure not to fiddle with it too much, to make sure I got the maximum chroma out of it.
Here’s the finished study, next to the lemon:
Modelling form with Munsell
So there’s a few very useful points you can take from this:
- Understanding colour from a point of view of hue, value and chroma, helps you paint more realistically.
- If the object you’re painting has a single local, then the hue will hardly change across the form. The value and chroma will change much more, so they show form turning much more than hue.
- Understanding the ‘modeling factors’ and where they fall on a simple object like a sphere helps you clarify what you’re seeing on more complex objects, too.
- You can use this knowledge to inform what you paint when you come to paint a real world object – like an clementine. Or a lemon, say.
Of course it takes practice to really understand colour in this way, it takes a mental shift. I’ve been working with this approach for a while, trying to understand form and colour better.
And once you understand colour in this way you can play with it. If you know how to model a form of a local blue, you can apply those modelling factors to anything – even something you don’t have physically have in front of you.
How about a blue lemon?
That bottom study was painted from what I’d gained from the two previous studies, without looking at the lemon at all. The colours for the modelling factors in blue come from the blue sphere. The form and the colours for the modelling factors of the yellow stripe come from the lemon study.
Although that’s a bit extreme, I hope it illustrates something: that understanding colour gives you the freedom to change things in your pictures, depending on the effect you want to create. It releases you from your visual impression, it allows you to create more – if you want to.
But most importantly – and this is where it really counts – it helps you to understand what you’re seeing and paint it much more accurately – and that means more realistically.
So if you struggle with colour and want to be able to paint more realistically, you can use this method to figure out exactly what colours you’re seeing, to mix them accurately, and then paint with them.
There will be times, of course, when what you’re trying to paint is outside the range of what paint can replicate. But even then, having this understanding of colour and light will help you to do that more effectively, too.
Knowledge really is power.
On Wednesday 24th August 2016, I gave a free webinar where I demonstrated this method of finding the right colours for the modelling factors live. It was a fun webinar (despite some sound problems towards the end!) Here’s the replay:
Best wishes and thanks for reading,
Free Value Tutorials
Subscribe: Join over 6000 other artists and get free updates. I'll also email you THREE FREE value tutorials that will bring your pictures to life.