Sewing that's Retro, Geek, and Chic
Tag Archives: experiment
July 10, 2013Posted by on
So I left all you readers with math swirling around in your heads. I hope Part 1 wasn’t too technical or too much of a cliffhanger!
Now I bet you’re eager to see how my experiment worked when I applied the math. =)
First, I copied my pattern in the bust 34 size at 91% on my printer and taped the pieces together. (As always, photos enlarge when clicked.)
I used some of my favorite pattern tracing paper to trace the other two copies of my pattern at the same size they came in – one bust 34 and one bust 31 size.
Because my traced patterns were on such sheer paper, I could do a good comparison of the shrunken 34 under the actual 31 to see the minor differences.
The differences were subtle but interesting. Here are a few closeups:
To me, the most fascinating part of the pattern comparison was the keyhole detail at the neck. When patterns are graded by the manufacturer, many of the details are not proportionally scaled, meaning that some details remain the same size no matter what size a pattern is drafted to be.
This means that on a smaller version of the dress, the keyholes are relatively larger. Likewise they are smaller on the larger versions. This does not always make for the most aesthetically pleasing details on the far ends of the grade.
After all, a pattern designer designs a pattern to have the correct detail proportions for that prototype size. I think it is somewhat lazy for the pattern companies to then grade without changing all the proportions accordingly.
My shrinking/enlarging method using a ratio solves this grading proportion problem.
Okay, so are you ready to see how the mockup experiment works on a real person?
Here they are on me:
Obviously, the actual bust 34 size is too big for me:
When you zoom in to look at dart placement and where the armhole hits at the shoulder joint, you can really see which fits best.
All three had necklines that were a little too high for comfort. And both the 31 and the shrunken 34 had armholes that were a bit small (typical on vintage patterns). In the above photo, I have cut directly on the stitchline for all the armholes, but I have reshaped and cut on the corrected underarm curve for the shrunken 34.
Obviously, you can see that I have chosen to fit the shrunken 34 because of all the safety pins and drawn lines on the mockup. It had the least amount of alterations needed.
I took in the side seams slightly, and pinned a test hem. I decided it needed a little more flare for the A-line skirt, which I added to the paper pattern when I transferred my pin markings to the pattern.
(Here’s an example of how I transferred fitting marks from another mockup to the paper pattern.)
So what are your thoughts and ideas about my pattern scaling method? I would love to know if this works for others too! Please comment or send me photos or links if you try your own version of my experiment!
And in case you’re curious, here’s the geeky fabric I plan to use to make the real version of my dress.
July 9, 2013Posted by on
This post was a long time coming. I knew I wanted to try this sizing experiment for more than a year; and over the last 6 months or so, I’ve been trying to work on it in between all my other work and personal projects.
(Warning: this post involves lots of words and math.)
First let’s talk a little about body measurements and standard sewing pattern sizing. Sewing patterns tend to be labeled with larger size numbers than mass produced clothing.
I talked about sewing pattern sizes vs. ready-to-wear size numbers and vanity sizing a couple years back in this post (but that was probably before many of my current readers found my blog).
I’m sure you’ve heard people say that the standard measurements used in the clothing industry have shifted over the decades. But they really haven’t changed as much as people think. The only thing that has really changed is the size number on the label – known commonly as “vanity sizing”. The standard body measurements have barely changed from year to year.
Here’s a section of a spreadsheet of my vintage patterns showing sizing for a 34-inch bust. Notice that the biggest change from year to year is the size number not the measurements used:
You’ll see that in the 1940s and ‘50s, a 34-inch bust was considered a size 16 in sewing patterns. In the late 1950s and through the 1960s, that same bust size became a size 14 with a slightly smaller waist and hips. Then in the 1970s, the 34-inch bust became a size 12 while the waist and hips went up a little in proportion – proof that there is a bit of psychological vanity sizing in sewing patterns, but not as drastic as it is in the ready-to-wear industry.
Today’s sewing pattern size 12 is the same as it was in the ‘70s with measurements of 34-26.5-36.
Most patterns are drafted in a mid-range size and then graded down for the smaller sizes and graded up for the larger sizes. This method involves some math and a whole bunch of measuring. (Threads Magazine provides a good explanation of grading here and a quick-guide chart here.)
Now, on to my experiment.
I admit I’m lazy when it comes to grading patterns. The math isn’t complicated, but all that measuring is a lot of work.
I started using my cheater’s method years ago when I needed to increase the size of flower-girl dresses for little girls who were larger than a standard pattern but proportionally still the shape of a standard girls’ pattern. (Part of the reason I started doing this stemmed from last minute clients who didn’t give me the time to do much patterning.)
I compared the differences on the pattern’s standard measurement charts and used ratio computations to find the percentage of the needed size increase. Then I just copied the pattern at the larger percentage on my home printer.
This enlarging method also helped me make a quick bodice pattern for a cousin in my sister’s wedding, who was really tall for her age but still not shaped like a woman yet. So far, this method for enlarging a kid’s pattern has worked for me every time with very few (if any) alterations needed.
When I wanted to sew a vintage pattern for myself but only had it in a size that was 2 or 3 sizes too big, I used my ratio method to shrink it to my size on the copier.
I did this for myself for a few years, getting a nearly perfect fit every time I shrank a pattern. For a long time, I thought vintage patterns were just drafted better, or differently.
But then I discovered that a vintage pattern in my actual size had as many fit issues as most modern sewing patterns. And I began to reexamine why shrinking by a percentage seemed to work better for me.
I think it’s because I am proportionally smaller than the average size used to draft most patterns. (I am small but not petite by clothing standards.) I don’t need merely a few areas changed, or some areas changed more than others (like when grading). I need everything equally shrunken down.
I had decided to test my theory by doing a real comparison of a vintage pattern that I owned in both a bust size 34 and a bust size 31 (my actual size).
I chose the Simplicity 7136 pattern from 1967 for a couple of reasons – I had it in two sizes and it had very basic style-lines, making it easier to compare small differences.
I made three mockups – one straight from the B31, one straight from the B34, and one B34 copied at 91%.
NOTE: Most vintage patterns have actual stitch lines printed on the pattern pieces. If you are using a pattern without marked seam allowance, you will need to either cut off the seams allowance before copying or make a line to measure what the new copied seam allowance ends up being. You will want to sew on the actual copied stitch lines, so adjustments to seam allowance may be required.
Here’s the math I used with the bust measurements printed on the pattern envelope to determine the scale factor:
Bust 1 31
Bust 2 34 = 91%
My theory was that I could also apply this percentage to both of the larger waist and hip measurements to obtain the corresponding chart values:
Waist 2 x 91% = Waist 1: 26 x 91% ≈ 24 Hip 2 x 91% = Hip 1: 36 x 91% ≈ 33
I’m obviously rounding to the nearest whole number because that’s the precision on the pattern envelope. And look! It’s pretty darn close. Here is how my exact ratio computations compared to Simplicity’s size chart:
Scaled 91% Simplicity Waist 23.66 24 Hips 32.76 33
My experiment worked!
Stay tuned for Part 2…