Before I started this series, I should probably have planned how many parts there would be. That would be sensible and clear, wouldn't it? But somehow I couldn't quite pin myself down… Perhaps it's a reaction to all the constraints elsewhere in life. Anyway, I thought I should be open with you: I am just following the thread where it takes me! 

This is part four out of some number bigger than four. I'll be adding this content to the Resources Hub in due course, but in the meantime links to previous emails are at the bottom of the page.

Last time we looked at a few huck drafts and identified some of the distinctive patterns in the liftplan which tell us what this structure is doing. This week we'll consider turning a huck liftplan into a tie-up.

The liftplan-to-tie-up conversion is one of the more challenging aspects of putting a weave into action. This is mainly because there is no guarantee that it can be done at all. 

On an 8-shaft loom there are 254 distinct shaft combinations we can choose to lift, and a liftplan may contain any number of these. If there are too many combinations in the design, and no way to reduce them, then the design needs a rethink. (Spoiler alert: the draft I'm sharing today poses no such problems. We'll work up to that!)

But why are we wrestling with this at all, when we can just get computer software to do this work? Well - besides being prepared for all eventualities, including life without the laptop - because it's the thinking behind the process that is important. It's about deepening our understanding of the way that weaving works.

I love a wiggly line, don't you? This is a draft for 3-end huck lace on 8 shafts. My original design in this style was on 16 shafts, but I wanted to see whether I could produce an 8-shaft version. This was what I came up with.

Let's say, though, that this image was shared on Pinterest. Can I convert what I see here into a feasible tie-up for my floor loom?

The process is essentially one of elimination: we need to whittle the draft right down to its barest essentials. The first step is to isolate a single repeat of the design. 

In this draft the effect of a curve is created by two vertical areas of lace connected by two diagonal areas of lace. Reading from the top, the drawdown begins with one of the verticals; then the pattern moves to the left; there's a second vertical area; and finally the pattern moves back to the right. I have cut the draft short just before the first vertical section appears again.

That gives us this liftplan to work with:

This liftplan consists of all the picks necessary to weave the design. There are 54 in total, but the crucial point for us is that they are not all unique. We need to remove all the duplicate picks, and what remains will be the distinct combinations of shafts that are required for this pattern.

We've already noted the distinctive rhythm of huck half-units in the liftplan, so we could break it up into threes to take a closer look. I'm going to take a somewhat brutal shortcut, however. 

Let's look at the first set of three picks, which alternate odd tabby-pattern pick-odd tabby. We have already found our first duplicate pick! We only need two out of the first three rows in the liftplan - I have outlined them in orange so you can see which rows I am talking about.

The second three picks have the same rhythm but now we see even tabby-pattern pick-even tabby. Again, we only need two out of three - outlined in green below.

And now that we have nailed both tabby picks, all we need from the rest of the liftplan are the pattern picks. We can immediately eliminate two-thirds of what remains.

Now that we have got rid of the tabby, the pattern picks themselves are much more visible. Our final elimination step is to remove any duplicates.

Are you surprised by how many duplicates there are? We have two types of repetition in this design. In each of the straight vertical areas we are directly repeating one pair of half-units; and the two diagonal areas we have the same sequence but in opposite directions.

We have whittled 54 picks down to ten unique combinations of shafts. This little stub of a liftplan is all that is left.

Turning turtle

For the process I have illustrated so far, I find it easier to work with pattern rather than numbers. That is, I have used a format of liftplan that simply involves black and white squares rather than explicitly numbering the shafts. Remember, though, that the leftmost column corresponds to shaft 1 and the rightmost to shaft 8.

To see this minimal set of lifts as a tie-up, then, I just need to tip it onto its back.

The top left corner of the liftplan has become the bottom left corner of the tie-up. Substituting numbers for black squares, we can see which shafts need to be tied to go up.

Assuming that I have ten treadles at my disposal, I could rush straight to the loom. However, the sequence would make for an awkward arrangement. As written here, the leftmost treadle would be the odd tabby, then we have a pattern treadle, and then the even tabby. 

You'll have your own preferences, but mine is typically to put the tabby treadles in the middle. Then I assign the pattern treadles which alternate with the odd tabby on the same side as the even tabby treadle, and vice versa. That will give me a left-right rhythm with my feet.

So that's the story of a liftplan becoming a tie-up. 

In fact (in a whisper) you can shortcut this whole process even further by clicking on either of the full-colour drafts I've shared and downloading a wif in the corresponding format. If you've managed to read all the way to end of this email, you deserve it!