Jyoti asked (in email) whether the progression of dye mixes is linear. The answer is no, it’s not at all linear. Some colors are “strong” and some are “weak”. In my observation colors with darker values are “strong” and colors with lighter values are “weak”, but I don’t know if that’s true generally.
Anyway, the first six (!) colors of the 29 are 90+% fuchsia (in fact the first three are 99+% fuchsia) and less than 10% intense blue. After that it progresses more or less linearly in units of 5% (athough I did do one 10% jump somewhere in the purple) through 100% intense blue. I then mixed the intense blue with turquoise in a series of 8 steps to get the transition to turquoise (not depicted in the skeins photo).
I’ve found that “pure” colors are very hard to mix without getting that hyperbolic trail-off, particularly when mixed with a darker color. The eye is very good at discerning the difference between a color with high intensity and a slightly muddier, darker color. So it was a real struggle getting a smooth transition from fuchsia to purple. (Just as well, though: it resulted in a broader fuchsia stripe, which is what I was after anyway.)
The dyes I am using are Cibacron (Sabracron) F, a fiber-reactive dye from Pro Chemical and Dye. Unfortunately Intense Blue has been discontinued by the manufacturer and is no longer available – it’s an absolutely gorgeous color and great for mixes. I bought about a pound of it before they ran out – wish I could have afforded more at the time! But that will keep me going for awhile.
Jyoti also asked how I was planning to use knitted blanks. I’m going to dye them, then unravel them and use them as weft. I could conceivably do this with warp as well, but it would be more complicated since I’d have to reskein it and steam to relax out the squiggles from having been knitted. With weft I’m told you can just wind it straight onto the bobbin.
Finally, Peg asked whether playing around in Photoshop produced an accurate result. The answer is that I tested it out the technique on my Liquid Fire shawl and the prediction was quite accurate. On Black Fire (Liquid Fire with a black weft) it did not work quite so well, but using a “dissove” (spatter) pattern gave a pretty accurate result. I’m not sure how it will work with this shawl. I already know that the gradient function doesn’t quite give the right colors for the shawl – they came out lighter and with less purple than the actual warp – but I think it will be close enough to see.
Someone else (non-weaving friend of mine) asked me if all this wasn’t a HUGE amount of work to put into a shawl, or even a series of four shawls? My answer is that yes, if it were work, it would be a huge amount of labor. But I don’t see it that way. To me it’s an extended exploration – of color, of weave structure, of new techniques. I’m enjoying all the parts of the process, and learning lots, so it doesn’t “parse” as work – just a lot of play. And I’m enjoying playing.
I have now settled on at least two patterns for the four shawls (it might only be three once I finish all my sampling). One of them is still in my head, but it’s basically a bunch of four pointed ovals repeating against a rippling background. The other will be plain goose-eyes (I think that’s the right term for the structure – rosepath threading and treadling with a straight twill tie-up), which will give me a nice four-shaft pattern that will display both warp color progressions (on the front) and weft (on the back). This will be useful if I do write up the technique for WeaveZine. It will result in less color blending but the results should still be interesting (and pretty!).
I have also selected two color progressions for the weft- one will be fuchsia to turquoise, and one will be either black to royal blue or royal blue to red (it depends on what my samples yield). I will also do one in a pure black weft. Cashmere-silk most likely – I have a lot of the stuff lying around, and it’s too weak to use except as weft. Which means I have to dye it and skeinwind it by hand, since it’s too weak to use the conewinder with. Foo. (That part is work.)
Back to designing patterns and doing simulations…I am hoping to get more warp on tonight.