Rembrandt Vs Turner: Who was right about values?

In the blue corner we have Rembrandt van Rijn. The very archetype of the Old Master, storyteller extraordinaire, magician of expression, character and light.

In the red corner we have Turner, progenitor of the impressionists, giant of British art and painter of the unpaintable whose primary subject matter was light.

Two painters, two very different approaches to value.

The pre-match warm up

Rembrandt is limbering up. He wears a large, worn, felt hat. He looks tired, as if the world has taken its toll on him. But don’t be fooled. He still knows how to handle a brush and can paint light with the best of them, his dark canvases punctuated with brilliant, intense patches of focused light.

And here’s Turner, squaring up to his easel, full of bluster, bristling as he holds forth, gesticulating wildly with his long brushes. “This is how you paint light,” he enthuses, as he fills his canvas with shifting, shimmering radiance.

These two painters had very different approaches to painting light. Both were masters of it.

Rembrandt’s approach can be called “top down“. It produces a darker overall painting with areas of intense, direct light.

Turner’s, and that of most of the impressionists who came after him, is “bottom up“. It makes for a lighter picture overall, with dark accents – the opposite of Rembrandt’s approach:

Both are methods of manipulating values which can yield very different, but equally beautiful results.

The descriptions of these two approaches come from The Practice and Science of Drawing (free download) by Harold Speed – probably the best book on drawing I know.

Why you should care

Value is one of the most important aspects of realistic drawing and painting. Done well, value gives us form, light and depth. Understand values, and your work will come alive.

Learning about these approaches to value will free us from attempting to copy exactly what we see (which is often impossible with value in any case) and put us in the driving seat.

It will also be extremely good practice.

Rembrandt and Turner, head to head

The best way to answer questions about painting is to get your brushes out and find the answers for yourself. And the best way to test something out is to simplify it as much as possible.

So to test this, I painted a white cube in a shadow box, in natural, indoor light. Nice and simple.

I soon found that it wasn’t as simple as I thought.

I already had a method of accurately judging values using chips based on the Munsell value scale and a little home-made isolator, like this:

So I started by attempting to paint exactly the values I saw on my white cube. But accurately replicating the values I saw didn’t give me a very convincing painting.

The values you see aren’t always the ones you should paint

First, here’s the set up: a white cube in a mid grey shadow box in indoor light:

It’s a fairly simple set up. You can easily make a shadow box like this with some cardboard and a grey cloth. It’s lined with grey to cut down reflected light, and control the light falling on the cube, to clarify and simplify the values.

As I said, I was trying to match exactly the values I saw. But I pretty quickly realised that the perceived value of my canvas (which was white) was lower than the lightest plane of the cube, so I couldn’t match it.

Ah, first problem.

That’s because a cube is three dimensional, and my canvas is two dimensional, so even though my canvas is white, I have less light falling on it than the lightest plane of the cube, which is angled more towards the light:

 

I also found that measuring the full range of values I could see, from the lightest plane of the cube to the cast shadow on the cloth, was just about the full range of paint, from titanium white to ivory black.

So if I painted the cast shadow black and the lightest plane titanium white, I could just about match all the values.

But as I looked at the cast shadow, I could see that it wasn’t black. It was actually a dark grey, about a value 2.5 in the Munsell value scale. A black cast shadow might not be the effect we want, right? Also that would leave us with no breathing space for highlights on the cube . And imagine if we added a black cube. How would we paint the shadow side of the cube if we’d already used our darkest value for the cast shadow? We’d have run out of room in our darks.

So Speed’s two approaches to coping with this problem, top down and bottom up, might help us to find a way round this.

Now, Speed also talks about another approach, used mostly by naturalist painters like Bastien-Lepage. This approach relies on compressing the values evenly, trying to preserve the proportions of the value relationships, but in a smaller range.

Seconds away…round one!

Let’s see what happens when we try to paint a white cube using each of these approaches.

1. Compressing the values

First, as a kind of control, here’s the Bastien-Lepage method, squeezing the relationships between the values, trying to find a fairly even compression. The cast shadow is accurate, about a Munsell 2.5. The light plane of the cube is white. It works ok I think.

Bastien-Lepage’s cube

2. Top down, Rembrandt’s cube

For this one, I started at the top and worked down, and kept the value relationships accurate until I ran out of room in the value range.

But I kept the cast shadow a visually accurate value 2.5. That meant I had to compress the values in the dark end of the scale. After all, if this was a still life and I had a white and a black object, I’d have to compress the values, since you can’t hit that range with paint (this is pretty easy to verify with a white and black cube and a little colour isolator, like the one I used above.)

Rembrandt’s cube

3. Bottom up, Turner’s cube:

For this one, I started with the actual value of the cast shadow, 2.5, and worked up till I ran out of room in the lights. That meant I had to do a lot of compression in the high end, and everything is shifted up the value scale.

Turner’s cube

So which works best?

That depends on the effect you want.

The most interesting parts of the paintings of Rembrandt’s and Turner’s cube are the shadow plane and the background.

They’re much darker on Rembrandt’s cube, giving that focused, sharp area of light in a darker overall picture. Turner’s cube appears full of light. The shadow plane of the cube being much lighter gives an impression of bouncing, reflected light in the shadow, giving the effect of more light in the picture overall.

Interesting. But who wants to paint cubes?

I know, I know, they don’t make the most interesting of pictures. We’re generally too busy trying to create beautiful finished pieces to spend time on things like this.

But I think this kind of practice is important. It will actually result in you progressing much faster. Because once you have value schemes like this in your head, from painting them a few times, you can use them at will depending on what effect you want to create.

And the winner is…

So who wins our celebrity Old Master Value-painting death match? Rembrandt or Turner? Or perhaps Bastien-Lepage?

The answer, of course, is that none of them were right about values, and all of them were, depending on what you’re effect trying for.

We can be thankful that we live in a time when we have so much information available to us, and so much history behind us that we can call on.

We don’t need to be partisan about anything. We can Google up some information, get our brushes out find our own answers.

So that’s what we should do.

Thanks for reading,

Paul

Thanks to James Gurney for inspiring this blog post by publishing a series of blog posts on his own blog on Speed’s book, and reminding me of this experiment. I hope you found it useful.