By working diligently (okay, obsessively 🙂 ), I have managed to get the first blank knitted, dyed, unraveled, and on the loom.  Here is a shot of the pirns – aren’t they pretty?

And here is the beginning of the weaving:

My only critique (so far) is that some of the color changes lo0k more abrupt than I’d like.  I will keep an eye on this as I weave it off, to see if I need to modify my painting technique for the next blank.

While I spent most of this week biting my nails and waiting by the phone (job interview nerves), I have wound and painted a new seven-yard sample warp, one which I hope will help me solve some of the process puzzles around making a painted warp (almost) precisely reproducible.

The problem is this: I need ten identical, or nearly-identical, painted warp/knitted-blank weft panels to make the jacket.  (Front, side front, back, side back, + sleeve = 5 panels x 2 sides = 10 panels.)  However, painted warps are notoriously difficult to reproduce precisely, especially when weaving multiple painted warp panels in a row.  Painted warps, in fact, tend to belong squarely to the improvisational weaver, not to inveterate planners like me.  Combine it with a knitted blank, which also tends to be imprecise, and making the panels match becomes that much harder.

So.  First, it is true that no two painted-warp panels will ever match precisely, because the dye flows differently, the warp shifts a little, etc.  That’s OK.  Mostly what I want is for the color changes to occur in about the same spot all the way up and down the garment, so they won’t be visually jarring.  That’s precise enough for me.

Second, the puzzle of how to make ten panels of painted warp match each other – especially difficult since warp bouts shift during painting, and while being beamed onto the loom.  The inaccuracies add up, and can easily amount to as much as 6 inches over the course of a 7-yard warp.  This is pretty simple to solve, though.  All you need is registration marks (so you can line up the bouts correctly when weaving a new panel) and a way to adjust the bouts individually.  The first is easily done by wrapping a tight choke tie around each warp bout at the start of each panel’s length of warp.  The choke tie resists the dye and results in a highly visible white spot on each bout; then all you have to do is line up the white spots, and you know the bouts are correctly arranged.  Adjusting the bouts is easy if you simply throw them off the back of the loom and weight them individually, rather than beaming onto a warp beam – then you can easily pull them forward/back as needed.  (I will use a trapeze for this, as I did with Kodachrome, so I can weave longer before having to adjust the weights.)

To make each panel match the others, simply paint next to a measuring tape – three inches of this color followed by three inches of that color, etc.  I used this technique quite effectively for my Kodachrome jacket.

Next is the puzzle of the knitted blank, which can be separated into three parts:

• How do I paint a knitted warp so the color changes are consistent across 10 knitted blanks?
• How do I know how long to make each section of color in the knitted blank? In this case the color changes in the knitted blank weft, once woven up, need to correspond to the color changes in the painted warp. This is decidedly nontrivial.
• How do I know where on the warp to start weaving the knitted blank weft, so that the color changes line up correctly?

The first question is easy to answer: if you place marker rows in the knitted blank (as I showed a few blog posts ago), it’s easy to paint each section a different color and get precise color changes that way.

The second question is also relatively easy to answer.  Basically, you need to figure out how much weft yarn goes into an inch of fabric, and knit up your blank with markers every 1″ worth of weft.  To do this:

1. Weave a short sample on the warp, and count the number of picks per inch.
2. Unweave part of it – say, about 10 picks’ worth – and measure the unwoven weft to determine exactly how long an average weft shot is.
3. This, in turn, tells you how much weft is required to weave one inch of fabric:  (Length of weft shot ) x (picks per inch) = (length of weft required to weave one inch of fabric).
4. Knit a short sample on the knitting machine at a width you think is appropriate.  Unravel several rows and measure the yarn.  If you unravel four or five rows, you should get a pretty good idea of how many inches there are in a row of knitting.  Adjust the number of stitches up and down until you’re happy with the length of yarn in a row.  (For my sanity’s sake, I like to make one row of stitches equal in length to one or two shots of weft; that makes calculations easy.)
5. Figure out how many rows are required to make one inch’s worth of weft.
6. Knit up a short sample blank, putting in a marker row after every inch’s worth of weft.
7. WEAVE A SAMPLE to make sure that all those wonderful calculations worked out more or less accurately, before committing to an entire knitted blank.  There are a lot of variables and opportunities for the yarn to shrink, especially when dealing with elastic fibers.  An inch’s worth of weft  should produce roughly an inch’s worth of fabric – plus or minus about 5%.
8. Knit your blanks, dye them identically, and start weaving!

(I know that doesn’t sound like a very simple solution, but it’s the simplest I know of.)

The third question was the stickiest one for me.  The best solution I could think of was measuring: after all the registration marks are lined up, I will start weaving the knitted-blank weft at exactly 4″ past the registration mark.  This may not be perfect, but at least it will be consistent across each panel.  If the knitted blanks are painted consistently, and the warps are lined up and also painted fairly consistently, the result should be a relatively consistent panel.

One could ask whether I’m going too far out of my way to make sure that the panels match.  After all, part of the interest in painted warps is in the complexity of the natural color changes; why not let the panels be somewhat different from each other?  Won’t that just add to the color complexity/texture/interest of the garment?

Well, perhaps.  But there will be plenty of color complexity/texture/interest in the garment even if the panels all match.  I am not trying for complete uniformity – that is basically impossible because of the many variables involved – but I want to keep the variation between panels subtle, not jarring.  And to do this, I need greater control over the warp and weft colors than is typically attempted.  Thus, the calculations and the lengthy sampling process.

The painted warp is in the oven baking (to set the dyes) as I type; tomorrow morning I’ll rinse it out.  While it’s drying, I’ll wind and beam on the metallic gold warp, which goes onto the sectional beam.  Between that, a guild meeting, and a craft group meeting, I’ll be surprised if I get anything else done tomorrow.  But I think that will be enough!