Wednesday, May 18, 2016

Colour Theory in Fair Isle Knitting

I've just spent the last several days playing around with stranded colourwork, and as part of that exercise I've reviewed some of the most authoritative writing on the subject. First, Alice Starmore's Book of Fair Isle Knitting, from Taunton Press (I have the 1988 edition). I had the good fortune to take some classes with Alice back in the early 1990s and in my mind she remains the queen of Fair Isle knitting. Just browsing through the pages where she interprets bits of nature from her native Shetland is the best inspiration and learning there is (especially the shells page!) BTW, it was delightful recently to see Alice dyeing wool with lichens in a pot over a fire on the beach in the TV series "Coast". Quite unexpected and a real treat.
The second source I turned to to refresh my understanding of fair isle colour theory was Mary Jane Mucklestone's book, "200 Fair Isle Motifs". She suggests dividing colours into light and dark, and then checking to see that the darkest in the light pile is still lighter than the lightest in the dark. She emphasizes the importance of having a "poison" colour, i.e. a colour that seems a bit odd or out of place, that will spark the others to life.
Finally, I re-read Jared Flood's two excellent blog posts on colourwork theory in stranded yokes. Both he and Mary Jane make the point that value (the relative lightness or darkness of a colour) trumps hue in importance when planning fair isle work. He shows us some examples of skeins in both colour and greyscale, the latter really highlighting how our eyes fail to distinguish contrast when two skeins of similar value are placed together. He also suggests knitting the most important motif in the darkest colour.
So, putting all that theory into practice, what have I come up with?
Here's the original "Fusion" chart. There's a problem that doesn't show up in this digital chart,


or even in this closeup of the actual knitting.



It's only when you see the knitting from a distance that the problem becomes more apparent.


The tips of the large motif (in Bird's Egg) are so close in value to the background colours (Split Pea and Bosc) that from a few feet away they are barely distinguishable. I still like this version of the design, perhaps BECAUSE it looks different up close than far away, but it illustrates the point about value.
Another way to understand the importance of value is to see what happens when a bunch of colours are organized by hue. Some time last weekend I realized that I didn't really love the bright greens and purples I was using with the grey main colour for my own Fusion cardigan. (Don't worry, I still intend to use those bright colours in Isabel's version, with the dark plum as the main colour.) I wanted an earthier set of colours to go with with grey, something that would echo the natural dyes of an earlier era. I chose this set of skeins. From left to right: Sedum, Honey, Bird's Egg, Slate, and Bosc. The Slate is actually much greyer and greener than it appears here on my computer screen.


My natural instinct was to use the blues for the background and the greens for the main motif (with the terracotta as the "poison"). Here's what happened.

Right away, you can see that this sample, at least on paper, suffers from the same problem as the original Fusion chart. The middle section seems to disappear, while the background at the top and bottom tends to dominate. All wrong!
Then, I rearranged the colours by value. That meant making the main motif almost entirely in the dark Slate, with the other colours flickering around in the background.


There's the new chart over on the right. I may extend the Bird's Egg (lighter blue) to encompass a couple of more rows, but over all, I love this. Plus it's really beautiful next to the light grey main colour. The whole thing has a lovely, faded antique quality.
And what about the bright green and purple colourway? I tweaked it a bit by moving the darker purple (Frank's Plum) to the outer edges of the large motif (see the bottom example).


So much better. Lessons learned!