Glazing and underpainting - for the frugal oil painter (grisaille, brunaille, verdaille, etc)

As I learn oil painting, I have become obsessed with glazing over monochromatic underpaintings like grisaille (black and white), brunaille (brown and white), verdaille (green and white) and other variations of this style.

The idea is to use cheaper paints to create a detailed underpainting of values and shapes.  Once dry, we apply extremely thin layers of paint over top.  

Here's my first attempt at a burnt umber underpainting.



This has a couple of benefits.

One.  Painting is hard enough to learn as it is.  This way, we can focus on shapes and values and deal with colour later.

green glaze over raw umber sample

Two.  Paint costs money.  The pretty colours are usually more expensive.  This way uses the cheaper paint to do all the work and tiny amount of fancy paint on top.

If I am going to be totally open with you about this, Three.  It's not a popular style of painting these last few hundred years.  It often seems actively discouraged.  So I want to try it.  

As you can see, I've done a bit of experimenting so far.  Still have a lot more to do and even more to learn.  Documenting it here and maybe others can join in with their thoughts and experiences.  

(for my own experiments, I'm using oil as both layers, although it's fairly common to use acrylics these days - I can imagine it would be quite similar)

Over here, I document my first expierence with oil beading on top of an underpainting.  https://permies.com/t/40/235117/art/Beginners-oil-painting-questions-oil#2848868

Some glazing experiments so far.

I tried some on my own but ended up using way too much paint.

After talking with my painting teacher about the project, he showed me just how little paint is needed on the brush for glazing.  And recommendations on transparent pigments to use for glazing.  He used no medium and it turned out amazing.

So I tried again and the results are staggering.

I have a lot of these little "10x10" pre-primed canvas panels to practice on because the metric system betrayed me.  

Still not quite at a place where I want to start the version I'm going to hang on the wall.  (and I suddenly realize I posted this out of order, imagine this goes before the above disaster)

One of the things about glazing, underpaintings, etc, is that oil paint is by it's nature somewhat transparent.  And it grows more transparent as it cures.  Even with opaque pigments, some light still travels through the layers of paint and bounces off the under layers.

In watercolour, we use this to our advantage and the white of the paper helps the painting glow.

In oils, it's more tricky.  

From what I can figure out, is the colour of the canvas and underlayers influences the final look of the painting.  If we want a warmer tone to the whole painting go with sienna or umber.  If we want cooler, go with black.  If we want the colours to 'vibrate' (whatever that means) use complimentary colours in the underpainting.

Only people online also say the opposite.  That the colour of the underpainting makes zero difference to the final result.

I suspect a lot of the confusion comes from all the different ways of oil painting.  

But inquiring minds need to know, so...



I'm tracing out six gosling samples to test my theory that: the underlayer makes a huge difference to the final painting, especially when transparent glazing is the main method of adding colour.

And since I want to use the cheaper paints on the bottom, I've chosen Burnt Umber, Raw Umber, and... I need four more to choose from.  Any suggestions?  (earth pigments are usually cheaper, but I'm up for trying some different ideas).

First sample is done.  I think he looks rather fetching!



This is done with Raw Umber and and Titanium white. Raw umber is an interesting colour and one I don't know very well.  At moments it looks grey and other times it looks blue.  

I'm also having trouble fighting the opaque and blue nature of Titanium white. Anyway, it's good enough for a sample.  


Another thing I learned was how little paint is required.  I started with my five (seven if you include the extremes) piles like this.



And ended with only this much paint used



So I took the remaining paint (except white as I can use that for the next sample) mixed it all together, then made piles of paint.  Each pile got a different medium like Walnut Alkyd (Which is my prime suspect for the beading problem up thread), some various mixtures of calcite putty, and a control group of just paint.  

It was amazing painting with these different types of medium one after another as they make a huge difference to how the paint behaves.  I had no idea.  Now I want to learn more, but first, more goslings.

I painted so much today, my eyes feel broken.  Like looking through a film negative.

This does look like a baby goose, right?

some more geese in ultramarine blue and burnt umber




The blue hurt my eyes to paint.  But it turned out pretty nice.

I should have stopped there but I kept painting even though my eyes were blurry.  I wasn't so good at just placing the paint, especially in the background, so it got "muddy" which I don't like.  

Working on green today.  



A change and an observation.

When I mix my paint piles normally, I take the extreme values the colours can do (titanium white and the dark) and then mix colours that compare to them.   Hang on a moment, let me grab a Munsell value scale to help find the language I need to explain.



Titanium white is about a 10 on this scale (technically less, but the real world doesn't match theory and there is no true white or 10... but never mind, for today, it's a ten).  And maybe raw umber is a solid 3.  Mars black might get to 9.5 or 10...but that's a later painting.  

So with a range off 3 to 10, I would mix up five piles using those as the "walls" of my value range.  

The middle pile would be in the middle of 3 and 10, so maybe munsell value scale 6.5-7ish.  Then the end values would be at least one value jump less than the extremes.  And the remaining piles would be between those... (keeping in mind that the photo alters how these show up, so I'm using the numbers of what I saw in real life)



This works fine.

however, I wondered what would happen if I, instead, used the Munsell Value Scale to change how I mix my piles.  

For the green, I mixed up my light to be a 9, my next one to be a 7, the middle pile I made middle grey (5) on that scale.  For darks it was trickier as I didn't have as much range.  So I got a 4, a 3.5 and a 3 (as dark as this colour could get in mass tone).  That leaves the pure white as 10 - and I try not to use it if possible.  

It worked better for my brain to do this way, even if I can't really explain it well with words.  Finding the right paint to put in the right spot was more instinctive when I knew which pile was middle grey - knew technically because I measured, not just by instinct (which was pretty close, but not exactly).

---

The observation was that I enjoy painting more and paint a better picture when it's a colour like.  Raw umber and this green both were effortless to paint.  Blue was too cerebral and burnt umber just annoyed me.  

At least a good goose according to my low standards.

It's also fascinating how each goose is so incredibly different although I am working from the same photo reference and tracing.  

So far I've chosen transparent and semi-transparent paints to mix with opaque white.

A big problem I'm having is that when it's just the colour paint, it's hard to get it to stay where I put it.  It doesn't come off the brush and just pushes around on the canvas.  So that the mix with just a tiny bit of white, although looking lighter, actually makes a darker colour on the canvas than the paint without white.

It's weird.  It makes no sense.  It's hard to explain.