Colour harmony, limited palettes and realism

I’ve recently started making a very conscious effort to restrict my colour palette (in individual pieces and sketch kits, my overall watercolour studio palette is still a 50+ pigment smorgasbord,  my coloured pencil collection is even larger, and I love it).

Mixing Surface, Studio Watercolor Palette and corner of Gladiolus painting
My usual studio setup – 52+ pigments in my studio palette, about 3-5 paints make it onto my mixing surface

The goal is both to practice more conscious colour mixing, and to train myself to make more deliberate and informed choices about colour and light interaction.  These are more broadly described with the term colour harmony and cohesion, but like many abstract concepts in art, have a strong scientific basis.

Colour harmony is one of these “fuzzy” subjective topics in art that I really didn’t “get”  for a long time.  Many artists extole the virtues of a limited palette for creating a “cohesive” look within their paintings, but each artist had different and conflicting suggestions for essential pigments.  In addition, doesn’t the real world contain all of the visible colours?  In a world with all the colours, what is a “cohesive” look anyway? Wouldn’t it be better to have every colour at your disposal?

Reality is quite a bit more complex.  The colours we actually see are affected by the colour of light, as well as reflected colour from surrounding objects.

Our eyes play tricks on us – for instance, we know a ripe banana is a bright, middle yellow, but in a dark room, in a fruit bowl surrounded by dark red fruit, the “correct” colour to paint a ripe banana for realism might be a neutralised orange, such as an ochre or Quinacridone Burnt Orange.  It took me a while to understand is that in this case, it wouldn’t just be “cohesion” or “harmony” that would be disrupted by painting a bright yellow banana, it would actually not be a realistic depiction of the banana.

Our eyes are actually even tricksier than this  – humans actually only have “good” colour accuity in a very narrow cone.  We do have a fair bit of peripheral vision outside this cone.  Our brain fills in the edges of our vision with colours to match what we see in front of us, then edits our memories to correct any inconsistencies as we look around.  Part of the reason a painting with a very bright pop of colour all the way to one side, which doesn’t appear elsewhere, is considered a “bad” or “uncohesive” composition is because this is not how we actually ever see the world.

Magenta and Two Greens
Two Greens and a Magenta – An effective colour palette for tomato plants?

In a way, choosing a very constrained colour palette is an effective way to mimic real lighting/environmental conditions as we might see them.  Take, for example, the limited colour palette above, where I chose a magenta (PR122) and two greens – a cool almost-turquoise PG7 (Winsor Green BS) and a warmer PY129 (Azo Green, or Green Gold).  This palette produces a surprising gamut of slightly neutralized warm colours from reds to rusty oranges and plums, as well as some eye-poppingly bright greens.  It might be a great colour palette for rendering berries or tomatoes shades by sunlit leaves – in fact, I am planning to paint these very pieces using primarily these colours.

Screenshot 2017-07-22 20.50.54
Instagram Feed by Iraville

And then, of course, there is the matter of using colour selection to draw attention to a specific element within a composition, or to evoke a mood within a collection or body of work.  I’ve recently started taking online courses with the School of Visual Storytelling (SVSLearn),  and was struck by the strong focus on manipulating colour and light to help with storytelling.

An illustrator who I think has a really stellar grasp of limited palettes for communicating mood and lighting is Ira Sluyterman van Langerweyde (Iraville).  Her studio palette rivals or exceeds mine, but she regularly uses a very limited, and fairly consistent subset to create a cohesive, whimsical, warm and natural-looking collection of playful illustrations.

In the field of botanical and scientific illustration, we are somewhat shielded from all of these concerns.  Botanical illustrators draw close-up, isolated subjects under bright, white light, specifically in order to communicate details of the individual subject without influence and distraction from surrounding scenery.  We don’t have to worry about the colour of ambient light, because we choose white light.  We don’t have to worry as much about competing colours with limited subjects, and can manipulate our composition without worrying about impacts on surrounding elements.  We use neutral backgrounds and are specifically discouraged from using cast shadows.

Nonetheless, I have found that the most successful scientific and botanical artists and illustrators, and the ones I admire the most make both deliberate and subconscious choices about lighting, colour, and subject selection which are very helpful in creating eyecatching, striking compositions.

This  can play out within individual pieces, where artists choose to highlight the unique characteristics of their subject – in my lichen on elm bark image in my header image, I really loved the stark, graphic white look of the cup fungi, and chose my colour balance in a way that would highlight that graphic, almost comic look, choosing to neutralize most of the greenish-gray crustose lichen and tree bark, but pump up the colour in the pops of green, rust and ochre.

Lichen and Cup Fungi on Elm Branch by Lee Angold
Cup Fungi and Lichen on Elm Branch – Coloured Pencil and Micron Pen by Lee Angold

On a broader scale, even realist artists develop a colour “brand” or “colour signature” across their body of work.  As an example, Anna Mason’s brightly lit, vibrant fruits and florals have a very different look from Jess Shepherd’s  dramatic and earthy leaves and vegetables, although they both make large-scale, realistic botanical paintings.

Muted Primaries
Muted, earthy triad – Spanish Gold Ochre, Salmon and Mayan Blue

Inspired by Iraville and other illustrators I admire who I have noticed use similar colour palettes, I thought it would be fun to create an earthy, natural triad to try sketching with or using as a base for some earthier subjects (obviously, this will never be an appropriate palette for very bright citrus or purple flowers  I used my new paints from Pruche along with Mayan Blue from Rublev, which is bluer than it appears in this image but still very dark and muted.  I generally prefer much brighter, clearer colours, especially greens, but I think this is an interesting jumping off point, in particular urban sketching and simple studies.

I’ve been spending a lot of time poring over James Gurney’s book Color and Light, which is fantastically detailed with loads of visual examples about different lighting situation, how they affect colours within a painting, and how they can be used to communicate mood and focus within a painting.  I particularly enjoyed the chapter on gamut mapping. In future posts, I’ll be exploring pieces created with a limited palette, and sharing the gamut mapping for each of them.

Rethinking the colour wheel: Mixing primaries from secondaries

Originally, I was going to give this blog post a much more clickbaity title, “There’s no such thing as primary colours”.

PG7 to PV23
Green + Violet = Blue ?!?

In this post, I will discuss why everything we’ve been taught about primary colours is an oversimplification to the point of being wrong. You can actually mix so-called “primary” colours from secondaries.  And on the flipside, no matter which set of 3+ colours you assign as primary, I can find you a shade that cannot be mixed from them

As a first example, take a look at my spectrum mixing chart to the left, mixing Winsor green with Winsor Violet.  According to every lesson on colour mixing I’ve ever seen, these are secondary colours.

If your primary school was anything like mine, you were taught fairly early on about the concept of primary colours.  Sometime in kindergarden, or maybe it was first grade, we were taught that there three primary colours that can’t be mixed, from which all others can be mixed, and shown a beautiful colour chart.

The primary, secondary, and tertiary color wheels.
Yellow, Red, Blue Primary mixes

In this first class, we were also taught that these three colours are Yellow, Blue and Red,  and then asked to mix a pie chart with the secondary and tertiary colours.

I dutifully completed this exercise, and like most of you, was dissapointed my result looked nothing like the example above.  Most of the mixes looked grey or brown.The flustered teacher comforted upset kids, assuring us it was not our fault the mixes came out muddy.

CMY colour mixing, from John Muir Laws

Some years later, perhaps in middle school, the same exercise was brought up again.  We were told the “primary” colours we had previously been taught didn’t work well because they weren’t the real primary colours, cyan, yellow and magenta.  With cyan, magenta and yellow,  we really could mix every other colour. Perhaps this was even paired with a lecture about additive and subtractive colour.

 

Once again we were asked to mix colours.  This time, the mixes did come out somewhat cleaner, but I was still wasn’t quite satisfied.  I could not, for the life of me, mix a pure, non-neutralized ultramarine blue, or the brightest greens.

Finally in high school and beyond, we were taught the concept of split primaries, where instead of 3 primary colours you have six (a warm and cool yellow, a violet-blue and a turquoise, a red and a magenta), but by this point, I had lost all faith in the concept of primary colours.

Where does the concept of primary colours come from, anyway?  In pure physical terms, colours are just a continuum of wavelengths of light in rainbow order (ROYGBIV) There are no colours with special “primary”  distinction.

There are, however, wavelengths of light that we distinguish better than others.   Our eyes have 3 colour receptors (red, blue and green) which each record light within a band of wavelengths and an intensity receptor. Our brains interpolate intermediate colour in the overlaps between these receptors (so what we perceive as yellow is a colour in the very narrow band between the extremely far apart green and red receptors where both are fired.    When both blue and red are fired we interpolate violet and magenta- magenta doesn’t even have an associated wavelength – it is the halfway between red at one end of the visible spectrum, and violet all the way at the other.

By wrapping that line of wavelengths around on itself, we create a “colour wheel”.  In actual fact, a wonky horseshoe/triangle might be a more accurate representation in 2D space of the colours we really see.  However circles are more satisfying and easy to understand than wonky horseshoes, so colour wheels it is.

On this colour wheel, you can pick any two colours, they don’t have to be primary colours,  and mix them together to get colours on the line between them.  Mix a yellow and a magenta together, and you will get an orange (albeit a less saturated orange than the brightest orange you could see).  Likewise, mix a violet and a green together, and you will get a blue, albeit a muted, dusky one.  There are many resources, such as the one below by Bruce McEvoy of Handprint, which map common pigments onto a colour wheel.

You can easily map out the area you can mix with any given palette of 3 or more colours by drawing lines between the pigments you are using.  The area enclosed is the colour gamut you can mix.  No three colours can actually mix all of the visible colours, however, they may be enough to render the subjects and lighting you are interested in representing.

So what makes primary colours special?

Our most accurate “primary colours”, in terms of paint (as opposed to emitted light) are in fact the midpoints between our three colour receptors placed on the colour wheel.  In other words, what we commonly consider primary colours: magenta, cyan and yellow, are actually the 3 colours we see least well.

While it seems backwards, when explained in words, that our primary colours are the ones that we don’t see very well, there is a good reason for this.

In the image to the right, you can see the results of the XKCD colour survey.  People were asked to name colours on a computer screen generated from RGB codes.  As you can see, the range of different colours we perceive as green and blue are huge compared to what we label yellow.

It is therefore very easy to mix  an acceptable “green”  from yellow and cyan or blue.  Mixing the “primary” colours, however, is a little bit trickier, as the target is much narrower

Amazonite to PV55Cyan, or blue, (as there are no true cyan pigments artists usually substitute a blue or teal), is probably the easiest of the primaries to mix from secondary colours. Any bluish-green with purple mixes a satisfactory range of teals and blues.

You’ve seen my PG7 and PV23 example above.  In case you thought that was a fluke or something to do with the specific pigments, here’s another example, mixed with Amazonite Genuine (DS) and Quinacridone Purple (PV55).

But this really does work with any colour to some degree.  Let’s take the example of the red/magenta area.  You can, of course, mix a red (in the grade school “primary” sense of the word) from magenta or rose and orange or yellow.  We’ve all done this.

Orange to violet.png

 

 

 

However, you can also mix an orange with a violet to get magenta.  You won’t get a clear, pure PR122 Purple Magenta colour, of course (just as you won’t get a clear, vibrant green mixing a far apart yellow and blue), but you will get a magenta colour saturated enough to work as a “primary”  in a muted palette.  The spectrum charts on the left are a little overexposed – my scanner does weird things to oranges, but you get the idea.

YellowsFinally, the hardest primary colours to mix (and get my scanner to show properly) are yellows.  Anything but the purest, most saturated yellows read as brown, olive or orange to the average observer.

To mix a yellow, my first attempt actually included two “secondary” colours actually in the yellow colour index.  PY129 (Azo Green, or green gold, a very yellowish green) and PY110 (Isoindoline or Indian Yellow, more of an orange at full saturation, mix to make closer to a primary yellow.

My scanner really shifted the masstone of my Azo Green, which truly does read as a green or chartreuse in real life, but the mixes in this spectrum are fairly close on my monitor to their real life colours.

In colour mixing, think beyond “primary colours”.  Sometimes the blue you need is most easily reached by mixing green and violet.  Sometimes the “secondary colour”  you need cannot be mixed from the “primary” colours you have available.

 

 

 

 

 

 

 

 

 

Watercolour Palette Tour 2017 – Teaser

Whew, I haven’t blogged in a really long time.  I’m eager to get back to it.

I’ve made a number of changes to my palettes since my last colourholics post last year.    Recently, I set up a “complete” studio palette in  a tin of half-pans of all of the watercolour paints I regularly use  (it claims to hold 48 – I haven’t filled up the 52 I crammed in there yet and I think I could find a way to fit a few more in the brush well) , and repurposed one of my old 24-half pan tins to hold my 14 most commonly used pigments in full pans.

Soon after creating my 14-colour “greatest hits” set, I received some freebie dot cards from Schmincke.  One of the samples was  a 12-colour dot card.  These are the colours they include in their filled version of the same tin.  I couldn’t resist painting out the comparison.  The results are a reminder of how far I’ve drifted in only a couple of years, when I used to use sets much like this one.

The top row in the dot card below are the colours in the Schmincke 12 colour set.  The ones in the bottom row and middle are the 14 colours I’ve recently put in the same tin.

Paint Palette
Comparison of Schmincke 12 paint set vs. my top 14.

The colours in the Schmincke 12-colour basic set are:

Lemon Yellow 215 (PY3)
Cadmium Yellow Light 224 (PY35)
Cadmium Red Light 349 (PR108)
Permanent Carmine 353 (PV19)
Ultramarine Finest 494 (PB29)
Prussian Blue 492 (PB27)
Phthalo Green 519 (PG7)
Permanent Green Olive 534 (PO62, PG7)
Yellow Ochre 655 (PY42)
English Venetian Red 649 (PR101)
Sepia Brown 663 (PBr7, PBk9, PB15:1)
Ivory Black 780 (PBk9)

My set contains the following (the starred items are the most permanent fixtures)

Nickel Azo Yellow (Py150, M. Graham)*
Quinacridone Gold (PO49, Daniel Smith)
Translucent Orange (P071, Schmincke)
Quinacridone Rust (P048, M. Graham)*
Quinacridone Rose (PV19, M. Graham)*
Purple Magenta (PV122, Schmincke)*
Neutral Tint (PV19, PG7 M. Graham)
Indanthrone Blue (PB60, Daniel Smith)*
Prussian Blue (PB27, M. Graham)*
Phthalo Turquoise (PB16, Winsor and Newton)*
Amazonite Genuine (n/a, Daniel Smith Primatek)*
Serpentine Genuine (n/a, Daniel Smith Primatek)*
Burnt Sienna (PBr7, M. Graham)
Raw Umber (PBr7, Daniel Smith)*

I find this comparison a rather hilarious reminder of how far I’ve come.  I much prefer my colours (of course).  They are overall much more transparent, more saturated and darkly valued, and are far more chromatic in the magenta and turquoise ranges.  I love turquoise.  Over the course of the past 2 years, I have very effectively gamut shifted from a beginner set like this one to something I find more attractive and fun to work with.

The colours in the Schmincke 12-paint set (and most 12-colour sets like it) remind me quite a bit of the basic box of crayola crayons or pencils.  This is not a criticism of Schmincke paints, as this is the impression I get from most basic colour sets across manufacturers.  The highest chroma “primary colours”  included are the yellow, blue and red we are familiar with from kindergarden – a highlighter yellow (Py3), an opaqueish orangey red (PR108), and a very violet blue (PB29).

The magenta tone in the Schmincke set (PV19) is still quite a warm rose, and relatively muted – there is no way to use this set to mix nice violets and magentas. It’s a shame, as Schmincke manufactures an amazing pure magenta (PR122 Purple Magenta) which would really round this set out.  Even a less muted, cooler variant of PV19 (Schmincke has a few) could be used.  Similarly, the greenish blue spot is occupied by PB27 Prussian Blue.  I actually quite like Prussian blue (and have included M. Graham PB27 in my own set), but the Schmincke version  is particularly muted, and I would like to see a high chroma  turquoise-slanted blue, such as a PB15:3 Phthalo Blue or  PB16 Phthalo Turquoise (my personal favourite) in that spot.

As with most beginner sets, the Schmincke set has a black watercolour as well as a mixed dark containing black.  As most artists, the first thing I do when I buy a set is chuck the black/white and replace it with a personal favourite.  The earth tones – a yellow ochre and “English Venetian red” a brownish brick red made from PR101, are a somewhat surprising choice.  I actually quite like the venetian red but I think I would prefer a PBr7 in this set, to mix neutrals with the Ultramarine Finest.  I don’t like the yellow ochre, and would prefer something livelier and more transparent.

I also got some larger dot cards from Schmincke with my order, which I will review in a later post.  I’ve had a lot of fun experimenting with different watercolours, and it’s renewed my confidence and love for the pigments I’ve chosen for my own palettes.  I look forward to sharing my full palette tour and colour swatches with you soon.

 

 

DIY Christmas Tiny Palettes

As holiday gifts, several people on my list are receiving lovingly handmade tiny watercolour palettes, filled with a selection of artist quality paints.  I’m also making a couple different versions for myself.

IMG_4097

 

My palettes are a shameless rip-off.  Last year, while putting together my urban sketching kit, I came across the absolutely amazing Expeditionary Art Pocket Palette by Maria Corryel Martin.

This palette has a number of advantages over other travel palettes.  it is extremely compact, fitting 14 colours into a business card holder. The pans are shallow but have more surface area than a standard half pan, making this a very usable palette that holds a surprising amount of paint.  In addition, the pans are made of tin and attach magnetically to the palette box, which means if you purchase extra pans, you can swap colours around at will.

Most cleverly (Maria Corryel Martin is a visionary genius), the Pocket Palette is assembled entirely out of commercially available products.  It is a snazzier version of the popular DIY watercolour kits made from Altoids tins, blister packs of gum, dollar store makeup palettes, etc.  Immediately, my gears started churning thinking of all the variations I could make myself. I’ll show you how to DIY your own palettes.

NOTE:  If you just want 1 or 2 basic palettes, support a brilliant artist by buying  from Expeditionary Art.  DIY is only cheaper if you plan to make several palettes or customize them.

What you’ll need:

How to assemble:

  • First, create your mixing surface by masking off the rest of the cardholder/box and using spray enamel on the inside lid.  Allow to dry according to package directions.
  • Squeeze your desired tube paints into tin makeup pans, and allow to dry.
  • Once your mixing surface is dry, stick the magnetic sheet to the inside bottom of your cardholder/box
  • Arrange your paint pans on the magnet surface.

IMG_4101.png

  • Paint colour swatches on watercolour paper for reference.  I used the makeup pans as a template.
  • Done!  Easy-peasy.

IMG_4098

The pictures in this post are of a palette I made and filled for my sister. For my sister’s watercolour kit,  I stuck fairly close to the original Expeditionary Art design. I ordered a business card holder engraved with her name from Vistaprint. My studio business card is included for scale reference.My enamel paint did not arrive in time, so my sister’s palette won’t get a mixing surface until later.

The chart below shows the colours I included in her set (she also received 5 empty pans to swap out with her own tubes).

IMG_4095.png

 

Most colours are from my usual studio tube set.  However, I also purchased a QoR High Chroma Set just for travel kits.  Two of my current favourite pigments are Green Gold/Azo Green (Py129) and Cobalt Teal (PB50) but these are both extra sticky liquid in my favourite M. Graham, and thus unsuitable for small travel pans.  QoR’s High Chroma set includes lovely versions of these pigments, as well as a deep rosy PR122 Quinacridone Magenta and a very rich PV23 Dioxazine Purple, which I also included in this palette.

In a later post, I will share some of the other small magnetic travel palettes I am working on for myself and others.

IMG_4101