Eight palettes, from the first pair of Japanese inks to the last set of stacks, measured the same way.
Where this landed. The last palette here is not the one in the app. It was replaced by four palettes of a few paints each, with their washes beside them.
They are not alike enough to compare by eye. One is pairs of inks behind arrows, one is nine pigments meant to be mixed in a well, one is a whole tin on a dial, and the last two are stacks you swipe between. So each is measured twice: what the palette can reach in total, and what it can reach without switching, which is what is in hand while painting. Reach counts the pigments and every pair of them mixed, and a slice of the hue circle only counts when the colour that lands in it has some life in it, so a grey that leans blue does not fill the blue slice.
| Palette | Colours | In hand | Whole reach | Reach in hand | Value | Darkest | |
|---|---|---|---|---|---|---|---|
| Complementary01 | 16 | 2 of 8 | 1.9 of 12 | 7 | replaced | ||
| Traditional02 | 16 | 2 of 8 | 1.4 of 12 | 10 | replaced | ||
| Wide gamut03 | 20 | 2 of 10 | 2.1 of 12 | 7 | replaced | ||
| Six parents04 | 9 | 9 | 11.0 of 12 | 7 | dropped | ||
| 静 Quiet05 | 16 | 2 of 8 | 1.1 of 12 | 9 | replaced | ||
| the 24-colour tin06 | 24 | 24 | 12.0 of 12 | 14 | replaced | ||
| Stacks by subject07 | 24 | 5 of 6 | 5.0 of 12 | 14 | replaced | ||
| Common palettes08 | 15 | 5 of 7 | 6.4 of 12 | 14 | replaced |
Whole reach is the twelve slices, filled with the most vivid colour the palette can put there. Reach in hand is the average across its sets or stacks. Value runs from black on the left to paper on the right; darkest is the darkest colour or mix any one set can make, where 0 is black and 100 is paper.
Every one of these palettes can reach nearly the whole hue circle given enough switching: between 10 and 12 of the 12 slices. That measure has never been the problem, and choosing a palette on it would pick the wrong one. What separates them is how much is in hand at one time.
The paired inks gave between 1.1 and 2.1 slices at a time, out of 12. 静 Quiet reached every slice as a whole and 1.1 in hand, the narrowest of the lot: its harmony came from having almost nothing available at once. The tin on the dial is the opposite, 12 of 12 in hand, and the trouble there was choosing rather than reaching. Stacks sit between: 5 of 12 in hand, which is about what a painter lays out.
Two things worth noticing. The common palettes hold 14 paints to the subject stacks' 24, yet reach the same 5 slices in hand on average, so the extra ten paints buy range across stacks rather than range inside one. And the Japanese inks go darker than the tin's: sumi sits at value 7 where Ivory Black sits at 14, which is why ink work looked deeper before the tin arrived.
Eight pairs of traditional Japanese inks, each pair a colour and the colour across the wheel from it, so any pair could make a grey. Living inside one pair kept a painting harmonised, but two pigments at a time meant most of the wheel sat behind the arrows.
One row: an ink and its opposite, three dilutions of each, with arrows to change pair.
16 colours, 8 sets of 6 swatches. Whole palette reaches 12 of 12 slices; the best single set reaches 3.
Quieter than Complementary and better for ink painting, at the cost of the greys that only opposites make cleanly.
The same row, with the pairs chosen for how they were used together rather than for opposition.
16 colours, 8 sets of 6 swatches. Whole palette reaches 10 of 12 slices; the best single set reaches 2.
Two more pairs to cover the gaps, which is more of the wheel and more arrow presses to get at it.
The same row with ten pairs instead of eight.
20 colours, 10 sets of 6 swatches. Whole palette reaches 11 of 12 slices; the best single set reaches 3.
Built for a well that never worked. The well was tried three times, as an XY pad, as a stir field and as the real engine in a second dish, and all three were dropped: the targets were too small for a fingertip and the mixes came out muddy. Without a well, nine parents are nine colours that are not the ones you want.
Nine pigments, water, and a mixing well. Nothing on it is a finished colour.
9 colours, all in hand at once. Whole palette reaches 11 of 12 slices; the best single set reaches 11.
An answer to colours that fought each other on the paper: every pair sits close in value and temperature, so wet edges blend instead of arguing. It was the last of the paired-ink palettes.
The row again, with pairs that share ground rather than oppose each other.
16 colours, 8 sets of 6 swatches. Whole palette reaches 12 of 12 slices; the best single set reaches 3.
Real paints with published pigment codes, so what the app mixes can be checked against what the tin does. Everything was in reach without switching, which is why it went on a dial. The problem was not reach but choice: 24 targets on one arc, and no reason to prefer any of them.
All 24 on a dial, then on a single arc running warm to cold, with brush size on the rim.
24 colours, all in hand at once. Whole palette reaches 12 of 12 slices; the best single set reaches 12.
The same 24 paints, cut into six stacks a painter would lay out for one kind of subject. Every stack holds a dark, a warm and a cool, so each one can finish a painting on its own, and between them they use every paint in the tin.
One card of five colours, swiped sideways to change stack.
24 colours, 6 sets of 5 pigments. Whole palette reaches 12 of 12 slices; the best single set reaches 7.
Palettes painters actually use, fitted to this tin. They are not balanced against each other and do not cover the tin: Zorn is three colours and no blue, Ink alone is one. The question they answer is whether a known palette beats one built to score well.
The same card, with each palette at its real size and water in a slot of its own.
15 colours, 7 sets of 1 to 6 pigments. Whole palette reaches 12 of 12 slices; the best single set reaches 10.