Sweater 2: Electronically generated boogaloo

So what’s the first step in making this sweater? Well, I’ve already measured myself (oops), so that leaves one input parameter (technically either 2 or 4 but whatever): Gauge. So let’s make a gauge block! This sweater will be made out of the last color of cheap acrylic yarn I bought to test with, which is this pinkish-brown which I personally find quite dubious. But, this is a sweater I’m testing for fit, I guess, so let’s go for it! Here’s 40 stitches and 50 rows between some waste yarn: 

I think this color’s starting to grow on me.

For context, this yarn is at the very end of what my machine can handle, in terms of yarn diameter / size. That means I’ll be using tension 10 (loosest possible) throughout this process. With other yarns, I’d need to take better note of tension, because that’s a significant player in yarn gauge.

While I work on this, I’m listening to Sixpence None the Richer’s self-titled album, only because I had “There She Goes” stuck in my head when I was thinking about what to play. It’s pretty good so far, I like the bass.

The yarn I used to make this swatch weighs about 14g (measured change of mass of cone), which means I’m getting about 0.007g / stitch, or 142.857 stitches / gram, which might not actually mean anything for me right now, especially since I don’t have an awesome way to determine the number of stitches in a sweater right now, but it’s neat to know, I guess. So what about gauge? In theory, you’re supposed to let this rest for two days, and/or wash it like the sweater would be washed, and/or block it, depending on who you ask. I really should do that, but break is only so long, and I’m really just in the mood to make a sweater right now. This is acrylic yarn, and I’m not really expecting it to change super drastically. I’ll probably run a bunch of tests down the line to see how various processes (waiting, washing, blocking, etc.) have on gauge for different fabric types, and possibly get a couple of cool graphs or even a paper out of it, but for right now, that’s a little more effort than my resources allow. I’ve tugged on it a bit to get it reasonably settled, and I’m measuring this sample to be 5 5/8 inches long, and 6 3/8 inches across. That gives 6.27 st/in and 8.89 r/in. Out of curiosity, I’ll also use my Susan Bates Knit-Check, which only measures over 2” square and in theory a less rigorous measure. Let’s see how far off they are! 12.5st/2in, 17r/2in, which gives 6.25st and 8.5 r per inch. That’s not too terribly far off, but I think I’m going to go with the 7s for this exercise.

Also, if you were wondering, this whole shebang is built in inches because that’s what my tools are in, and it’s the measurement system that I have a feel for in the context of the human body. Everything should be just as easily done in cm, or even just converting the gauges if you needed to.

Here we go! We’re ready to rumble, baybeee! That album is over, so I’m switching to Wheatus’ self-titled album, which I have heard before and absolutely has no right to go as hard as it does. My spreadsheet kicks out stitch and row information down to three decimal points, which I recognize as excessive, but I wanted to be able to do my own rounding for whatever reason. I guess I think I know something. For the back, I’m casting on from 70L to 71R to cover 141st total.

After finishing the turned hem and knitting roughly 70 rows, Wheatus is over, which again goes incredibly hard, so we’re going to go with Electric Dirt by Levon Helm for a nice little tone shift. I’ve heard “I Wish I Knew How It Would Feel To Be Free” off this album, and I’m really grasping at straws here. Cool album art, though!

…

Whew. It’s been a hot minute. I started working on this sweater on the 7th, it’s now the 15th. I finished the sweater last night, and there were definitely a couple of notes I have, both about the process and my spreadsheet. The actual labor of knitting the sweater definitely did not take that long, and in theory could still be done over one, two, or three days, especially once I’m done troubleshooting.

The spreadsheet itself works great. It presents the information that I need really clearly, it’s pretty easy to work with, and honestly I’m not sure I really need to translate it into Python, at least until I motorize Reka and start using the outputs of this program as something other than text I read. There were some logical errors that I made in the calculations, though. Some pretty rough ones. For one, the front and back panels made different allocations for the set-in sleeve, because I forgot to account for the top-of-the-shoulder shaping. I had designed it with the idea that it was better to imply measurements than to bother to actually calculate them, which in theory results in a more flexible design with less counting, etc. but among other things, these two sleeves were different lengths. Also, I had to figure out how to decrease stitches for the sleeves, which is a set of problem-solving that will get its own post soon, because it was A. a big pain, B. a great lesson, and C. something I’m pretty proud of.

As a sweater, this sweater is terrible. It’s pretty wonky, there are things wrong with both sleeves, and it makes me self-conscious to wear because of how entirely wack it is, which is exactly the opposite of the point of this entire project. But this sweater was not the culmination of this entire project. The goal of me knitting this sweater was not at all to have a good sweater. The point of this sweater is to test and refine my generation of sweater patterns. It really helped me do that, so it’s definitely something I qualify as a success. Also, I ran out of yarn on the last sleeve, which is why the rest of that sleeve is yellow. I think it looks nice, but I didn’t account for (or measure for) any difference in gauge, so we’ll see how it measures up. Here’s some pictures of the sweater, so you know what I’m working with:

Here’s what testing this out is going to look like: 1. I’m going to measure the dimensions of the sweater as it is, and compare those to the intended values from the pattern. Those values are actually my measurements plus some added room, which is based on numbers that I kind of guessed at, based on what I thought would make sense and make the sweater fit and function (and have a neck hole big enough for my head to go through…). Then I’m going to wear the sweater and figure out what parts of it are too tight or loose, relative to what I think is good in a sweater. Then I’ll adjust that parameter.

Step 1: Actual vs. goal:

• Waist (intended: 43”): 21” per side, total 42”. Pretty close, error more likely due to seam allowance.

• Neck (21”): 21”, right on the dot. Nice!

• Sleeve length (25”): Armpit to end of cuff; Right: Just under 24”. Left (yellow…): 27.5”. Horrid.

• Total length (28”): Top of shoulder to bottom of hem; 31”. It feels super long, too.

• Sleeve at wrist (8”): Right: 11”; Left: 14”. Egregious.

• Sleeve at shoulder (19”): R: 26”; L: 29”. Godawful.

So there are a couple of things about this that I knew were going to be poorly matched, because of some of the issues I’ve already spotted (and hopefully fixed) in the spreadsheet. I’m not entirely sure why the two sleeves are quite so different, because every part of the pattern is directly symmetrical, but I think it had something to do with how I hung up the stitches at the shoulder. They were quite so far off from their intended values because, again, the spreadsheet wasn’t accounting for the top-of-the-shoulder shaping when considering how long the arm holes needed to be, and that should be handled now. I’ve also added in a section that says, when you’re hanging those stitches up, “there should be about this many.” which I think will help in the future. I’m tempted to blame the much greater total length on that as well, as well as the sleeve circumferences. In fact, everything that wasn’t touched by that issue is pretty much spot-on, the waist, the neck (which is the least straightforward to calculate for, I think.). The differences in sleeve lengths is something absolutely crazy that at this current time I cannot explain.

Step 2: Actual vs. me:

• Waist: (as measured)
  The important thing to note about these measurements is that they were taken when the fabric was not under any tension- so when it’s on my body, there’s actually some more room that it can stretch. I think I could possibly take this dimension down a couple of inches and get a sweater that is more form-fitting, but that still fits. I’ll eventually do some tests as to evaluating elasticity and expansion in this knit fabric, but for now I think I’d be willing to actually knit to 3” less than a waist measurement, and go from there.

• Neck: (5” larger than neck)
  I’m not sure how exactly to make this measurement inference more generalizable, but the reason I added 5” was because on my first sweater (not something I blogged about, but actually a better sweater than this was), I had knit it with the size of my neck in mind, and it was hard for me to get my head through when putting it on. Figures that your head is usually bigger than your neck. Funny how things work out like that. Anyways, the reason I added 5” was because the circumference of my head is 5” greater than that of my neck. I’m not sure if adding 5” will be enough for everyone that will be getting one of these sweaters, but that’s something I guess I can find out later.

• Sleeve Length: (One too short, one too long)
  Even the one that’s technically too short feels too long. Maybe not when I fully extend my arm, but since this fabric ended up kind of stiff, it just always feels like it’s up past my wrist, which is just uncomfortable when I’m trying to do stuff. The big sleeve is absolutely cavernous. I can make my left arm into the freaking sand dragon from the Mandalorian. It can go take a nap in the depths of this sleeve. Much too large.

• Total Length: (Like 6” longer than needed oops)
  Yeah, holy cow is this too long. I’m really not a fan. I feel like I could wear this sweater without pants on and it would be only as terrible as even just wearing this terrible sweater at all. I think it really would have been the right length if things had gone to plan.

• Sleeve at wrist: (Much larger than measured)
  The smaller one is pretty good, I like it the way it is. It’s not too tight, which would have been horrible, and I can still do the thing where you grab onto the cuff and contract into a smaller being, which is something I’m usually unable to do as being very nearly a giant, so this is promising. The big sleeve could be used as a sweater for a python. I doubt even The Rock could fill this sleeve.

• Sleeve at shoulder: (Much larger than measured.)
  I waved my arms around a bunch and then realized I looked a fool. Then I lifted up a chair a couple of times? It’s not bad to get in and out of the sweater, which is where this tends to cause trouble, and they’re not so tight that you can tell, which my first sweater was. Overall, pretty good. Pretty bulky, I don’t think they need to be quite this big however.

Maybe I’ll pin things and see how much smaller I could go in terms of making a sweater look at least kind of like my human body. At least nothing was too tight, I guess. Then I couldn’t wear the sweater for reasons other than “pride“ or “general public decency.” I’m kidding. It’s not that bad.

This has been a post which is 75% devoted to picking apart every way in which the sweater that is currently on my body is absolutely terrible. It does not read as a person who is happy or proud of their work. In reality, this sweater was made for the express purpose of being analyzed and improved upon. It has certainly exceeded expectations in this regard. Not only that, but I have actually managed to generate sweater patterns from human measurements. From scratch, mostly and kind of. Hot damn. That’s very impressive, and both you and I should be impressed. That was most of the project, wasn’t it?

I’m pretty confident that I’ve worked out the more glaring issues with the calculations. It’s going to take a bit of tuning, but in general, I think I’m pretty much at the point where I’d like to start testing it on other people who are not me. Maybe these “add” parameters shouldn’t be hard-coded measurements, and instead should be percentage tolerances, or something else. How else can I be sure but to add extra data points? That’s going to be a whole other slew of blog posts. Also, I have pretty much run out of cheap yarn at this point, and am having one hell of a time finding more. I guess I’ll have to set up an Etsy and start selling headbands. Weather-wise, it’s not really the year for it, I don’t think, but we’ll see!