Returning to the list of topics from the reader’s poll, stretch knit pattern grading is next so that’s what we’ll do today. To refresh your memory, I’d said, knit pattern grading the way that Stuart does it which is very different from how we do it in North America. To discuss the dichotomy of Stuart’s approach requires that you already understand grading to see why his method is a departure from what we typically do and we can discuss why it may be better and easier. My role is best described as a facilitator of this discussion because my experience with his method is very limited, albeit successful. And by the way, I refer to Stuart Anderson whose website is better known as Pattern School. Pattern School is offline for now but you can easily access it via the WayBack Machine and all links I’ll use will direct there. Ready? Okey dokey.
Stuart says there are two types of grading, incremental and proportional. Traditionally, we use incremental which means we attach a value to grow or shrink a given point according to X or Y coordinates. [At right is an illustration of rules applied to a pattern the traditional way.] Stuart uses another way, namely proportional. This requires calculating the degree of stretch in the fabric and using a feature common to CAD programs, stretch the pattern pieces in accordance with the stretch properties. Doing it this way can be easier, faster and more accurate than grading via XY coordinates.
For a simple example, let’s say the fabric stretches by 6% width wise and 3% lengthwise (this is actually less than the full extension, knit recovery must be included in the calculation of stretch). Let us say you also know that you want the pattern to change a total of 2″ per size (XS-XL). So you would do the math to figure out how that 2″ translates with respect to the stretch percentage. The formula can be found on Stuart’s page that I already linked to twice. It is better to do it this way because the patterns aren’t getting bigger than they should be once you consider the stretch. And I know, we traditionalists guesstimate that a 2″ grade actually means we should use 3/4″ or 7/8″ instead of a full inch per back and front but that’s what it is -a guesstimate. Using this proportional method means no guessing! The added benefit is that newer graders can catch up more quickly with experienced graders who have developed (over many years experience) arcane rules of thumb.
The solution is to use the stretch/shrink tool in your CAD system to grow or shrink each given piece by the percentage you’ve established. This is pretty simple for a tee shirt -but what if you had a very complex style with side panels, yokes, pieced sleeves and all that? If you grade the traditional way, you have to spend some time figuring out how much of the grade goes here or there so the garment is proportional to the mid range size. With Stuart’s proportional method, you can forget doing any and all of that (we’re free! we’re free!). All you need to do is apply the same exact percentage to every piece in the pattern. Too bad this won’t work for wovens…
Of course, I have to try these things out. It is not so much that I don’t trust Stuart but I have to do it to understand it. My concern was that seams might not measure –exactly– to line up again with the pieces they sew to —but they did.
After I graded several pattern proportionately, I mapped the pieces by stacking them and “graded” the piece as we would traditionally to see what the grade rules were of the proportionately graded pieces. I have a screen cap below but there is also an excel file if you want to look at it more closely. The conclusion? These grade rules don’t look anything like what we typically see. Very complex stuff, too complex to actually attempt to do it traditionally.
See what I mean? Those are some hairy grade rules. Traditionally, every point (1-6) would be the same for all sizes or if not, there would be a readily identifiable pattern.
Questions? Comments?
I have some pending issues with this as it relates to my CAD program knowing I have a graded set -which I can post about another time. I also have a help ticket (in a manner of speaking) out with Stuart so following up is not a problem. I wouldn’t be surprised if Stuart drops in to impart some of his imitable wisdom.
Hello,
I tried Stuart’s method when I was grading a colour-blocked bikini pattern. (You can see the pattern here http://strawberrymilkrun.blogspot.com.au/2011/10/colour-block-bikini-pattern.html).
I do my pattern-making in corel, not a dedicated pattern making software.
Initially, I split my pattern into its colour blocked pieces, then applied the vertical and horizontal stretch amounts (In corel, you apply separate vertical and horizontal stretch amounts, not an overall amount). I found that my seam lines did not match up on the graded pieces…I think because my horizontal and vertical stretch amounts were different (which is appropriate for swimwear), but my seam lines were not completely horizontal or vertical.
Next go, I drew the colour blocked lines on my bikini, but kept the piece as a whole. Then I graded the pattern up, again using different horizontal and vertical stretch. After the pattern was graded, I split it into its colour blocked (which was easy to do because I drew the split lines on before I graded). This time around, the seam lines matched perfectly.
I just mention this to save somebody the same pitfall I had :)
I am not a math person and my brain fuzzes over when confronted with numbers. Also I only sew for myself so even though I have graded by hand I don’t have to. Inspite of my limitations this makes sense to me and corresponds to a knitting technique in which one starts with a given measurement, ie chest and then you calculate the rest of the garment/sweater using percentages rather than actual measurements. With this method of knitting one does not need a pattern. You calculate neck, sleeve, etc. by increasing/decreasing a certain percentage. As best as I can determine the concept is the same/similar.
Interesting Katherine; that could be good to know for those who don’t have CAD. I wonder if Illustrator is the same?
Just an fyi for those unfamiliar with apparel cad programs: I would be surprised if anyone had a problem such that Katherine describes. We routinely have to adjust patterns for shrinkage. Grading in this way uses the same utility. I’ve adjusted a lot of patterns for shrinkage and have been very pleased with the outcome.
Donna: now I know why that when I was searching for knit grading online, only hand knitting came up!
Kathleen, is there a reason that one couldn’t grade patterns for wovens this way?
I know that you said the standard is to grade up or down a fixed amount per size. But, given the advantage in not having to divide the grading amount between different pieces, it seems this would make the grading much easier. In theory, sequential percentages could make the differences between sizes progressively larger as you get farther from your middle, but this could be adjusted for. Or maybe it would actually result in a better fit.
down to 64ths of an inch…I don’t think I have a ruler with anything past 16ths.
Why not this for woven fabrics? I’m guessing it’s that usually wovens don’t have a difference in lenthwise/widthwise dimensions like knits do? I wonder what the results are if you did apply those formulas to a 0% stretch knit??? I’ll leave that for others to ponder for a bit, I haven’t had enough coffee for that yet.
I am the biggest fan of Pattern School! Stuart is amazing! I do bikinis and I could never gotten as far with my pattern development with out him.
I draft my patterns on Illustrator and print them out on a 11×17 inch printer. (small pieces) and use proportional grading. It seems to work great. With that said, I have found one small problem, that I will try to explain.
I had mentioned my “field trip” to see my mechanic’s place in another post. When I was there I asked his partner if she knew of anyone who would be interested in spending a couple of hours and show me how to sew the damn things right, she said she would be interested. Well, 4 hours and 200 dollars later (worth every penny) she asked me if I would like to see how she graded. She showed me how she graded a halter top. Basically she moved one point out and down, the strap was kept the same width for all sizes. It took me a minute and I realized that because the elastic from the inside and outside butted together that if you went smaller you would have to use thiner elastic, bigger it would just get wider but because they did not have any fancy binder, turner what ever, they used the elastic as the guide when sewing and the tolerances being so small that your eye can really see 1/16 of a inch difference. Then when you are doing something like a triangle top that has for lack of a better word Tab that you attach the 1/4 inch spaghetti strap and the tab gets bigger or smaller it will not fit right.
Sorry if this was hard to follow……I hope you understand what I am trying to say.
Hi, I’ve been trawling the internet for some info and am hoping I’ve finally found help!
I have used CAD before but don’t have access to it now, so am manually grading and percentages would take me a little while to calculate… I’m also used to grading in metric so some figures in your example take a bit for me to translate!
That aside, here’s my real problem-
I am trying to figure out how best to grade some shirts I’m working on, they are supplex/lycra t-shirt more or less, so everything above would apply for stretch etc, BUT they have woven collar stands and plackets (they’re for equestrian sports so quite high and close around the neck).
Trying to avoid having a huge size run ie AUS 6-14 as it’s not needed for the main shirt is stretch, but don’t know what sort of increments I should use to cater for both the stretch and the woven aspect (especially the woven being around the neck) and have fewer sizes – XS, S, M, L, or just S, M, L
And I have sampled it with the collar stand/placket in the stretch but it doesn’t sit how it needs to, so had to go back to the wovens for that, which incidently causes this problem!
Have had three people- an AUS 6, 8 and 10 in the one size shirt and it fit all three, but where to go from there I just can’t decide what’s best!
Any insights? Think my poor little brain has probably way over-complicated it too!
Any help would be greatly appeciated :)
I also wonder why it can’t be done for wovens that way?
Somehow I do not understand why it shouldn’t be possible?
You wouldn’t know it from the looks of things but I’ve been trying to figure out how to do video screen capturing so I could create some videos for demo purposes and upload them to youtube. I’ve got a simple test project working now so if it loads successfully, I’ll have more to report tomorrow. Thanks for your patience.
Clothing is one of the lines we make custom. We have this same issue, however, because we show samples in softshells that have a small variance in stretch from style to style, and can guess how much more we need to add to compensate. Really, its just experience and fly-by-the-seat-of-your pants. But occasionally we use an outerwear fabric like Powerstretch, and making a jacket in this turns out to be 1-2 sizes larger than the same jacket in the regular stretch softshells, such as Powershield. It is very frustrating because we don’t want to keep samples of every fabric, in every size jacket we make.
Don’t I know it.. I’ve noticed even different colours from the same brand/weight can have a different % of stretch.. I’m pulling my hair out here trying to figure out how to adjust my patterns–usually when there isn’t enough vertical stretch in the fabric.. (some of the same fabrics purchased from you).
Wayback Machine is no longer working for pattern school. Is there any way to get the proportional formula to grade knits this way?
Hi,
I was unable to find Stuart’s page, after some research, it seems it has been taken down?
I’m looking for a swimwear grading resource that I can buy such as a book. If anyone has any suggestions please let me know!
I am a swimwear pattern maker by trade and unfortunately we don’t use a CAD system at my work- it’s all manual!! I’d love a guide I can consult as I won’t be able to use Stuart’s method described above manually! I have long sleeved swimsuits with collars, multi-panelled, all very complex.
Otherwise, if anyone has ideas for basic swimwear grading the old way (eg grade for body rise / length for a full piece suit)?
I could upload a diagram that we have been using here to see if people agree with the grade.
Looking fwd to responses!
I found a link, I’m not certain it’s the same but it’s a good access point. http://web.archive.org/web/20130113081552/http://www.patternschool.com/?page_id=22
This link looks like the home page where you start: http://web.archive.org/web/20130114064706/http://www.patternschool.com/?page_id=54