Drying shift is the tendency of watercolor paints to become lighter and less vibrant as they dry. This is a bane of my existence, increasing the difficulty of painting with enough value range. You can paint something out that looks high-contrast but have it all dry pale and weak-looking. Why does this happen, and what can we do?
Note: in this post, I will use the following color theory terms:
- Value: How light or dark a paint is
- Chroma: How bright or muted a paint is (high chroma = brighter); aka vibrance or saturation
- Hue: The actual color of a paint (e.g. red, orange, yellow, blue, etc.)
Which paints have a drying shift?
Trying to learn more about why drying shift happens and to which paints, I found this article by Lee Muir Haman useful, as well as this detailed explanation by Bruce MacEvoy. From them, I learned some interesting facts I had not previously known:
- Mostly, colors lose chroma rather than value, meaning they get duller/less satured. Subjectively, though, this can make them appear lighter. Some colors also change hue.
- Muted colors are not necessarily prone to more drying shift, but can appear to be since they have less chroma to lose. (You might not notice if Opera Pink loses a bit of its excess brightness, but you will notice if Perylene Maroon goes from moderately muted to very muted.)
- Colors with a small particle size are more likely to display drying shift. This explains why high-granulation paints tend not to have too much drying shift, and also why many of my favorites do: I tend to love smooth, transparent paints. If I’m going to continue to love them, I have to accept drying shift as part of the deal.
The MacEvoy post has a list of paints in which he actually measured drying shift by taking measurements of hue, value, and chroma while wet vs dry. This lends credence to some of my own, anecdotal observations. Here are some of the worst shifters from his list, starting with the most total shift, and my thoughts:
- Black pigments – Lose value, but not chroma (what would that even mean for black?) But the losing value is a big problem for a paint whose only role is as a dark value-setter. You put down a flat-looking black paint and it dries light and patchy. You can get the deep black you want if your paint is thick enough, but how do you know? Not-thick-enough and thick-enough look the same wet.
- Prussian Blue – If you know how Prussian Blue looks on the page – muted, moody – it can be a shock to see it wet, when it nearly looks like Phthalo Blue. However, to me, the loss of value is the real problem. I always want to use Prussian Blue as a dark color. I prefer it in gouache, where it doesn’t seem to lose as much value.
- Phthalo Blue Red Shade (PB15:1) – Loses chroma and changes hue, from a green bias to a middle blue. It goes down looking for all the world like Phthalo Blue GS but dries looking like a very moderate middle blue. Lightens a bit too.
- Naphthamide Maroon (PR171) – Like Perylene Maroon and Violet, which it resembles, this muted maroon lightens and loses chroma.
- Indanthrone Blue (PB60) – MacEvoy found this lightened by 50%! It should be noted that he used a Winsor & Newton version, which is one of the greener ones. My perception is that Daniel Smith’s more muted, violet-biased version does not lighten as much. The DS version does lighten but it doesn’t seem remarkable to me; when I tried the Da Vinci version, a green one like WN’s, I found that it lightened a frustrating amount, to the point where I stopped using it.
- Dioxazine Violet (PV23) – Lightens and loses chroma. My preference is for Daniel Smith’s version of this pigment, Carbazole Violet, not because it does any better but because it starts a bit more low-chroma than most other versions, so the chroma loss doesn’t seem as pronounced or surprising. The loss of value does bug me a bit though.
- Chromium Oxide Green (PG17) – Loses chroma, and gains value (becomes darker rather than lighter).
- Phthalo Turquoise (PB16) – Loses a significant amount of chroma. Darkness loss is present, but not as noticeable. Weird… I don’t really think of this as being a problematic color.
- Perylene Maroon (PR179) – Loses a bit of value and a lot of chroma. I like the wet color, which is a rich deep claret red, but when it dries it’s very dull and washed-out. I would have thought of this as one of my first answers for “what colors have a big drying shift?”
- Perylene Red (PR178) – Doesn’t lose lightness (if anything, it dries darker), but does lose a lot of chroma. It goes on very bright, so IMO the loss of chroma mostly makes it look more handsome.
I noticed that most blues had significant drying shifts. The major exception is Cerulean Blue (PB35), which ranked among the highest for stay-the-same-ness. Other winners were many forms of bright yellow. I personally rarely notice a drying shift problem with the middle of the pack, including most Quinacridone and Pyrrol colors, though there are finicky exceptions (like Quinacridone Burnt Orange, Quinacridone Violet, and Perylene Red).
Some more that annoy me, not listed by MacEvoy:
- Other Perylenes: Like Perylene Maroon, Perylene Violet and Perylene Green go on dark and bold, but lose a lot of value as they dry. They do not feel reliable to me as dark value setters.
- Deep Scarlet (PR175): It paints out a rich dark red but always seems to dry weak.
- Transparent Orange (PO71 or PO107): This is mostly a chroma issue. Goes on bright orange, but dries to a more muted, pumpkin color.
Strategies for Working Around Drying Shift
So what to do if a paint you otherwise like has a significant, annoying drying shift? Here are some strategies.
Use a different paint instead (one that doesn’t have as much drying shift).
This is the obvious solution if you keep running into an annoying drying shift with a particular paint, and it’s the path I’ve chosen for several colors (especially those I didn’t like much anyway). However, so many of the paints that I otherwise love have a drying shift that I want more strategies to work around it. If you only choose paints with minimal drying shift, you are limited to a particular set of mostly-granulating paints, with your only blue being Cerulean.
Change your mindset.
Commit to memory the way the color looks when dry, and try your best to think of that as “the way the color looks,” not the way it looks when it’s wet and it first comes off your brush.
It’s funny, I feel I have successfully done that for certain colors (such as Phthalo Blue RS and Chromium Oxide Green). Although they are on the “most drying shift” list, they don’t bother me much because I know the way they look off the brush is false.
For many other colors, I find it difficult to remember. I consistently expect Perylene Violet to keep its off-the-brush juicy grape color, and forget how brownish it looks when dry.
Use the paint in a different role.
A paint that loses chroma will bother you most if your main reason for using that color is its brightness. Similarly, a paint’s tendency to lighten will annoy you most when you are using it to set dark values. Knowing which paints shift and in what ways can help to use them for roles where they’ll shine, instead of falling short.
Phthalo Blue RS is one of the most notorious changers, yet it doesn’t typically bother me in the ways that I tend to use the paint. I mostly use Phthalo Blue RS as a sky color. I personally have a tendency to make skies too dark, bright, and green-biased, so Phthalo Blue RS’s drying shift compensates for all that. The ways in which it changes work with the way that I use the color, instead of against it.
The times I’ve been most frustrated with drying shifts are times when I tried to use a color with a significant value drying shift to set my darks, e.g. Perylene Green. When the color I used for my darkest darks dries very light, it throws everything out of whack.
Mix multiple paints to get darker darks.
Generally, I’ve found that most individual single paints are pretty bad at getting dark enough on their own, but mixes of two or more colors can get much darker. If you want any area of your painting to be truly dark, don’t rely on a single dark paint, use a mix of two or more complementary colors. For best results, at least one color should be transparent (since opaque colors, especially light ones, can only get so dark).
Get more rigorous about water control.
Overdilution is the #1 reason why my paintings turn out too light. Either I put down too much water on the page for a wet-on-wet technique, or I put too much water in the paint mix, or both. Drying shift exacerbates this when a paint has a similar wet appearance in masstone and midtone, so you can’t tell by eye how much pigment you have in your mix, until it dries and you see that what you thought was masstone was actually midtone.
I’m getting better at noticing when this is happening by judging paint consistency by its feel and how much movement it has on the page rather than simply by how dark the wet paint appears, but some paints are more difficult to judge than others.
Some practical tips for improving water control (without relying on the appearance of the paint):
- Activate pan paints by dropping or misting a little water on them a few minutes before painting. Trying to take pigment directly off a dry cake of paint will always result in a weak wash.
- Wet paper evenly for a wet-on-wet wash; brush out any puddles before you paint in them.
- Use a smaller brush if you find yourself dropping too much water on your strokes, because smaller brushes hold less water in the belly. You can also pull water out of a loaded brush by pressing a paper towel just to the thick part of the brush.
- Wipe your brush dry after rinsing it, or even after intentionally going in for more water. It takes the tiniest amount of water to make your palette puddle appear a lot more diluted.
- Use a flat mixing palette, rather than one with wells, to force yourself to use thicker paint to make your palette puddles and prevent water from collecting in one place.
- Dry off your mixing palette thoroughly after rinsing it, to avoid invisible water puddles or droplets getting into your mixes.
Let the wet painting be ugly.
This is the logical extension of the idea of compensating for drying shift on certain paints you know to be affected, and the aspect I find hardest to swallow. Watercolor looks different wet and dry, so you have to pick which one you want to optimize for. One of the hardest truths I’ve had to learn about watercolor is that if it looks perfect wet, it’s probably going to look worse dry… which means that if you want it to look good dry, it has to look worse wet!
The most zen thing would be to decide you’re after an experience rather than a product, so who cares how it looks dry. I’m not there.
Instead, right now I’m interested in learning the tricky skill of figuring out in which ways to make it look bad wet, that will lead to it looking better dry.
Some things I’ve figured out:
- Don’t mix too much. In wet-on-wet skies or other large areas, random brushstroke shapes and unmixed/barely-mixed color combinations will soften, spread, and mingle naturally in drying. You don’t have to spread and mix them yourself with your brush; if you do, they may become too uniform. That just-barely-mixed look is watercolor magic.
- Go for high-contrast. You will lose contrast as the painting dries. Besides, too much contrast is very rarely a problem.
- When in doubt, stop. If you get confused or don’t like the way the painting looks wet, just stop and let it dry completely. You can’t problem-solve until you know what the actual problems are and you won’t know until the painting is dry.
This is a very difficult skill to learn, and one I’m working on!