How to check the accuracy of graded patterns

Posted by Kathleen Fasanella on Jun 22, 2012 at 4:50 pm / Grading, Patterns / Trackback

grade_post_seam_measure2Edited 6/23/12
So you had a pattern graded, what do you do next? By way of introduction to what can go amiss if you don’t know what to do after you get a pattern graded comes this comment (emphasis is mine, edited for clarity):

In the midst of my first production, I visited the factory today and found out that the sleeves of my dresses were much larger than the armholes they are supposed to fit in to. In some cases, the sleeves were 1″ larger on each side, front and back armholes (total 2″). This was across the board, 5 styles in 5 graded sizes. Result: waste of fabric and sewing time trying to correct the problem. The sample pattern was checked against the sample size in the marker. Don’t graders have the ability to check seam lengths on graded sizes? Has this happened to anyone?

The short answer is yes, graders have a tool to check grades which I will show you and then I will describe the process to follow after having a pattern graded to prevent this sort of thing. Before proceeding, one off topic mention: Sleeve and armhole seams should match evenly; see Sleeve cap ease is bogus. Additionally, it bears reminding that a grader might not be at fault if the pattern was wonky before it got to them and no, they’re not responsible for checking it unless you pay for it. Which is why a lot of pattern people don’t want to grade patterns they didn’t make because patterns must be checked for accuracy (see how to do it yourself) before grading to prevent cumulative error.

Using my CAD program (StyleCAD), I can display seam measures in a variety of ways. This first one (top right) shows seam lengths (the key is F7 if you have this software) from notch to notch. In the above example I’m showing two adjoining pieces. From the left corner to the notch, the seam is 3 63/64ths on both sides. From the notch to the right corner, the seam length is 2″. In this way I’m checking not only total seam length but that the notches are exact on the two pieces that will be sewn together.

Once the pattern is graded, I can compare all of the lengths easily, even per size.  The illustration below is another seam (not the same as above).  The piece names are SFRT and FRT, this is found under “pattern” on the left side of the chart with the graded sizes (10-18) listed beneath. [The base size is 14 which is why the chart shows zeros for that size.]


The perimeter column lists seam lengths of each size. The increment column shows how much the seam (red dashed line) changes per size. In this case, the seam lengths are changing plus or minus 3/16ths per size.  In all cases but size 10, we have an exact match in seam length. The size 10 SFRT is 1/64th shorter. By the way, if you are wondering why the pieces have such strange names, this will help.

The grader should check their work (the uniformity of the grade) before sending you the nest. This means that the column labeled “increment” should be uniform (usually logarithmic). However, it doesn’t necessarily hold that the perimeter column will be a match like mine is unless you hired the grader to walk the pattern to verify seam lengths. A lot of designers are certain their patterns are perfect because it was tested with a sewn sample but this isn’t the same thing because people make adjustments during sewing of one-offs and think nothing of it.

But I digress. It isn’t possible for a grader to export charts such that I’ve shown you so you can check more easily. This means you have to do something else.

There are two ways for you to check the grade. One is to request that the grader export the grading chart to Excel, any industrial program should do this. I don’t know what the output looks like for other CAD systems but I really like how StyleCAD sets theirs up. Below is an image sample of one piece.


Analyzing one of these Excel spreadsheets is rather mind numbing (just my opinion) because you have to compare the grade of points 1-2 of FRT to points 4-6 of SFRT. Those point numbers are just examples but you have to jump around in the spreadsheet a lot. I would much prefer to do this visually but if you don’t have a CAD system, it is a bit difficult to do it. So what can you do?

When you have a pattern graded, it needs to be approved and the person doing the approving is usually you. In anticipation of that, the grader will send you a nested pattern (all of this is described in much more detail in my book so if you are confused about any of this, reading it is the first step). The nest will have all of the pieces laid inside each other. The usual process of checking the grade is with a ruler. You measure seam lengths of various adjoining seams to make certain they will sew together. It is critical this be done well because once you approve it, if anything goes wrong with it after that, whomever proceeded you is technically off the hook. With approval, you “had it last”. In real life though, unless you’re working with a total jerk, an error discovered after the fact should be repaired at no cost to you provided you or somebody else didn’t fiddle with it. I think it is a good practice for you to save the nest and for the pattern grader to save that version with a unique file name.

However, these days since so many people don’t feel qualified to approve the grade, I suppose you could ask the grader to sign off on it. You need to make it clear so they save a version of this file independent of any later changes. Depending on the complexity, it may take time for the grader to highlight anomalies. For example, in the image I showed you measuring perimeters with the chart, the size 10 was 1/64th off. Sometimes -depending on the grade you requested- a small variation like this may be unavoidable. In this particular case, the designer wanted a companion piece (not shown, a sleeve) to grow a non-standard length so this affected the grow of this piece. [1/64th is not much of a problem but it does bring to bear my constant preaching that your base pattern must be perfect. If it were not, the differences magnify, errors are cumulative when it comes to grading apparel.] Personally I don’t think it is out of line for a grader to charge for the time it takes to explain these things to you because this is the responsibility of design and it would take you time to do it yourself. Once you understood (learning as you went along), you wouldn’t need so much hand holding so it would cost less.

To be certain there are still other ways to check the grade but they are much more laborious and costly in terms of materials. By way of example comes this query:

Once I get the nest, do I test it by sewing one sample? Or should I sew every size?

Perish the thought! I will answer this in my next post. Til soon~

13 Responses to “How to check the accuracy of graded patterns”

Comments RSS feed

June 22nd, 2012
6:35 PM

Kathleen, I just had this problem with a client this week and am buying them your book as a gift. This is such an obvious time waster and happens around here in shops more than you would expect.

Paul V
June 23rd, 2012
3:48 AM

Based upon the output table the 10FRT is 1/64″ inch longer. Could it actually be a problem in cell formatting (as in Excel) and there is actually no error in the the graded pattern? Regardless, having a print or a table of the measurements would allow anyone to physically check the grade where an error appears in the table.
SFRT – side front – seam appears to be on a ‘princess’ line offset to the underarm

Kathleen Fasanella
June 23rd, 2012
9:15 AM

Fyi, the post has been edited a bit. Perhaps worth re-reading.

Dara: the book is very popular with service providers. There are quite a few who won’t take someone unless they’ve read it. Even experienced companies use it internally it because it makes training new people so much faster and less expensive.

Paul: the chart that shows the shortfall is within the CAD program itself so the error can’t be attributed to [an export to] Excel.

Paul V
June 23rd, 2012
10:51 AM

Yes, I see the editing and now I understand the growth
Kathleen, I wasn’t thinking it was exported to Excel but I was wondering if the cell formatting in the CAD program could be reporting it as a “rounding error” that you might see in an Excel spread sheet. In other words, is the pattern really off by 1/64th of an inch? Not sure how a CAD program could make an error when the mathematical calculation should be the same, getting the same result as the adjoining seam. Three-sixteenths of an inch subtracted from 5 and 51/64 inches is not 5.625 inches. That is what made me think there is a rounding error within the cell definition or in the programming of the CAD system. Strange thing, don’t you think, even if the pattern is supposed to grow? Doesn’t it usually grow as it gets larger rather than smaller? Whatever – just an example I suppose.
Yes, it does show that you need to get this information before signing off on the graded patterns, though too, I would think the person/people hired to do the grading should check their work. I would also wonder why the CAD system wasn’t programmed to highlight or underline this. In some cases a difference is desired, but it should still be highlighted or underlined, in my humble opinion.
In CAD programs I use, a line that doesn’t match will be highlighted or will change color until you either change the dimension or over ride the “error”.

June 23rd, 2012
11:13 AM

I was pretty sure I understood what you meant so thanks for the confirmation. I thought it would be good to address it anyway in case someone else didn’t.

The difference is very difficult to explain and amounts to grading on a logarithmic scale. If you look at fig 5.66 on page 172 of my book, this is how the grade was done. It was not an even 2″ per size; the smallest size took a smaller grade (1.75″). If it had been graded for still other small sizes, those sizes (2 or so) would also graded at 1.75″, after which still smaller sizes would grade at 1.5″ and so on. (see this related entry if you’re new to these parts)

Point is, it would have been easier to even it all out (eliminating the 1/64) by creating a size break but it can’t be justified for only one smaller size.

June 23rd, 2012
11:20 AM

Forgot to speak to this:

I would also wonder why the CAD system wasn’t programmed to highlight or underline this. In some cases a difference is desired, but it should still be highlighted or underlined, in my humble opinion.

For all I know you can do something comparative. It would make sense. It would be very complex to program it because of the related dependencies. You’d have to specify that X seam pairs with Y etc and then whether they could change with respect to direct measure or proportion. As a user, I would find this feature unwieldy but I’m not an advanced user.

June 25th, 2012
1:17 PM

Thanks for this Kathleen. Grading matters cause my head to swim. Not sure if this is exactly on topic but I thought I’d throw my question out there for comments….

My patterns created/tested/sampled in a size small. However, when they are graded I’ve had several styles where the fit got horribly funky (sorry for the technical term). I know the grader followed the grade rules so it’s not that component. It’s something with my intial sample, but I’ll be darned if I can figure out what it is. Does anyone ever have different grade rules for different types of styles? To refresh you, I do women’s sleepwear. And to be more specific, on occassion in the larger sizes, the bust comes out way too small.
Sorry for the typos, having PC issues….

Paul V
June 25th, 2012
9:58 PM

I think I remember this being addressed in an earlier post. It may be in the archive. What I do remember is that for larger sizes and there is a break point depending upon the intended customer, the grading rule has to change. Maybe in your case the grading rule has to change in the chest or bust more than it changes in the body length.
If you are noticing this at a certain size, then it seems that is the break point size where the grading rule needs to change. The unfortunate part is that grading alone may not take care of all the potential problems in fit at the break point size and larger.
I have had similar problems trying to develop off-the-rack corsets. In the old days, the majority of corsets were off-the-rack but the shops that sold them could and would custom fit them to the client or customer. When I was a kid my mother would shop for girdles and long-line bras in a store that specifically sold these items. There they would reduce the girdle to properly fit the waist. Girdles were sized to fit the hip properly and the waist could be reduced, if necessary to better fit the customer, much like custom tailoring. There are few places that do this now.

[…] up where we left off in part one, someone I’ll call Mary asks: At this point I’m not 100% sure what the procedure is, […]

June 26th, 2012
8:18 AM

My patterns created/tested/sampled in a size small. However, when they are graded I’ve had several styles where the fit got horribly funky (sorry for the technical term). I know the grader followed the grade rules so it’s not that component. It’s something with my intial sample, but I’ll be darned if I can figure out what it is. Does anyone ever have different grade rules for different types of styles? To refresh you, I do women’s sleepwear. And to be more specific, on occassion in the larger sizes, the bust comes out way too small.

A couple of things.
One of the main reasons you use a medium (as explained in my book) is to prevent grade distortion.

Second, it is more common than not (or should be more common) for grades to change based on styling. One can’t really say a given grade is right because a grade should changed based on given styles and even silhouettes. A useful analogy may be food portion sizes. A serving size depends on calories and satiety. A serving of butter is one tablespoon but a serving of tossed salad may be a full cup or more. Each of these are foods (styles) but serving sizes vary greatly -as should your designs based on fabrication weight/type and styling.

To whit, using the analogy of a nutritionist using portion sizes interchangeably between food items, one’s confidence in a given grade should be reexamined.

[…] you need to catch up, see parts one and two; today I’ll post three and four. Part three includes reasons why I don’t think […]

[…] colleague who wishes his company did sew samples in all sizes (if you need to catch up, see parts one, two and three). This email is very telling -it says much more than on the face of it. It is a very […]

Ngoni Comfort Akarakiri
December 28th, 2013
4:36 AM

I went through your site and I really appreciated what I read about needles and entrepreneurship in garment industry, I must confess I really learnt a lot. Your BOOK IS A MUST BUY. QUESTION – Is there
any sewing/fashion school where I can obtain Masters Degree in Fashion or where can I learn practical at an affordable rate. I have a Degree in Home Economics from a reputable University in Nigeria and have interest in sewing/fashion and entrepreneurship. Pls help me, I really have interest and I need help. Regards. Ngozi Comfort Akarakiri

Leave a Reply



The Entrepreneur's Guide to Sewn Product Manufacturing

Often described as the garment industry “blue book”, the most highly rated book in the business is guaranteed to get you off to a solid start or your money back. Many service providers require you read this before they’ll work with you. Learn more »

Subscription Options

RSS Feed My Yahoo

Subscribe by email: