I’ve decided to go ahead and knit to the shoulder of the first side panel for the waving lace sweater. Now that I have all of the correct yarn measurements, it’s looking like I should be able to make it at the current gauge with the yarn that I have. I’ll double check my yarn usage when I get to the shoulder, but I think it’s going to work out ok.

Now that the knitting is back on track, I am turning my attention to the pattern writing for this section of the sweater. I’ve been finding that it’s easiest to break the pattern writing up into pieces that advance at the same time as the knitting. I’ll knit one piece, write its pattern, and then start knitting the next. That way I have a solid draft of the pattern written by the time I finish knitting the sweater, and in most cases it’s already semi-test knit.

Since I love to talk about process (and, judging by the comments, you do too), I thought I’d talk through some of the things I’m thinking about while writing the pattern. If there were such a thing as a Handbook of Pattern Writing, these are the kinds of things that I think it should contain…all the details you don’t think about until you get to the pattern writing stage and realize that you still have a lot of figuring left to do.

Incidentally, posts like this are precisely the reason that I will most likely self-publish all of my patterns. There is far too much “secret knitting” in the standard publishing world for my taste. I figure that if you’re the sort of person that can look at a schematic and knit the sweater, then you’re probably also the sort of person that can look at a picture of the sweater and knit the sweater. Either way, you don’t really need the pattern, and I think it’s far more interesting to talk about design than it is to sell patterns, anyway. So without further ado, let’s poke around under the hood, shall we?

For both this and the lace ribs sweater, I’m finding that pattern grading (creating more than one size garment) is one of the most time consuming parts to write. Sizing up a fitted pattern isn’t a simple matter of scaling up the design; different directions increase at slightly different rates, so there are always adjustments to be made here and there to make sure  that the final piece will actually fit a human shape. This is proving to be a real challenge for the waving lace sweater, mostly because the lace motif is so large.

Right now, it’s looking like I am going to have to either write two different versions of the pattern to accommodate the different sizes, or publish a pattern where different sizes have a slightly different neckline.

The drawing below should give you a rough idea of how I’m shaping this pattern. It’s very simple, really: it just follows the curves of the lace. This inclination toward natural shaping was the first thing that inspired this sweater, and it’s one of my favorite parts of the design. When shaping for the waist I leave out half of one repeat of the lace pattern, then add it back with a little extra at the bust, then decrease following the lace at the underarm and the neckline.

So far, so good. I am knitting the version on the left; it starts out with 3 lace repeats and goes smoothly from there. The problem comes in when I want to change the size.

Each lace repeat is about 2-2.5 inches wide (with almost no blocking). Generally, garment sizes are spaced about 4″ apart. That means that I would need to add half a repeat to each sweater front and one full repeat to the sweater back in order to go up one size. At first, that didn’t seem like a big problem.

And then I drew it out.

You can see the problem in the drawing on the right. Its underarm seam and side shaping is exactly the same as the one on the left, but the neck shaping has to start either half a repeat earlier or half a repeat later if it’s going to follow the lace. That will raise/lower the neckline by about 2″, which is a really dramatic difference.

It seems that I have three (ok…four) choices. The first option is to use only full lace repeats and publish a pattern that is “missing” certain sizes, and space the sizes at 8″ apart. That’s certainly the easiest option, and we saw in the last post that you can get a pretty big difference in size just by blocking the lace differently (it just so happens that the difference in the two blocking methods is about 2″ per front panel, which works out to 8″ across the whole sweater, which is the same size difference as adding a full lace repeat).

The second option would be to publish both the full repeat and the half repeat version shown in the diagram. That would allow a full complement of sizes, but the necklines would be different. Writing a pattern for both versions would involve two separate sets of charts and written instructions, meaning that I’d basically have to write two separate patterns.

The third option would be to add some kind of spacer column at the edge of the piece that would give it the extra inch needed to make the size adjustment.

I don’t really like any of these options. The first is by far the easiest, and if you account for changes due to blocking it covers the whole range of sizes. But I don’t really like the idea of having to change the base fabric of the pattern to get the size that you want (also, the grading may not work out perfectly because blocking does stretch the knitting in all directions simultaneously, and proper grading shouldn’t).

The second option is a lot more work, but I have to say I’m currently leaning in that direction. It would keep the basic gauge/fabric the same, which to me is a more attractive option. It’s not a perfect solution, though, because the necklines will be different. As with option 1, I don’t really like the idea of the design being controlled by the size you want to knit.

I’m reluctant to say that the third option probably makes the most sense; adding a non-lace panel would give me the extra flexibility that I want in the sizing without requiring as much fussing with the lace itself. It would still change all of the underarm increases/decreases, though, so it would probably still require two versions of the pattern. I’m also concerned that adding an extra panel would affect the flow of the lace shaping, which is really fun to knit.

Well, then. On to option four, which is really two in one: change the basic lace repeat, or change the design. Altering the lace to be a smaller repeat would help to make smaller steps in the sizing. It would also change the overall look of the fabric, but it could be worthwhile. I could also experiment with placing the increases and decreases differently, working them into the pattern a little bit more so that they can be adjusted more subtly. In the lace ribs sweater, I altered the individual lace repeats in order to get the waist and bust shaping I wanted, and so didn’t have to add much extra shaping.

Spreading the shaping throughout the sweater also gives me more power to control increases and decreases without changing the overall look of the piece. It doesn’t utilize the lace pattern as nicely as my original design, but I’m wondering if this might be a case of needing to “kill your darlings,” which is common advice given to writers, and basically means that sometimes you have to cut the best stuff in favor of producing a better flow in the whole. I don’t like this advice (few writers do either, I presume), but it’s often a lot easier to work with the design rather than fighting against it in pursuit of some ideal vision.

I’m mostly thinking out loud at the moment, so I haven’t yet come to any real conclusion about what needs to be done. I think I’m likely to end up with some version of option 4. I don’t really want to go all the way back to the drawing board, but it would be worth it if it gets me a better design. I have a sinking feeling that I’ll be reknitting that front panel in any case, but first I think some swatching is in order.