Musings on Digital Patterns

There has been a discussion on one of the sewing boards about a company that produces patterns to be downloaded and printed at home.  When someone commented that a digital pattern should fit better, my mind started whirling and I started pondering.

I have always thought that the idea of digitally
transmitted, downloaded and printed patterns is wonderful. But having worked with computers for many
years, and having written CAD programs and programs that do 3 dimensional
modeling and problem-solving, I am well aware that this is not as easy as one
might think.

First of all there is the issue of getting the measurements. Four, eight, or even twelve measurements are not enough to
adequately reproduce a human body. Then,
everyone takes measurements differently. I suspect that if you have 4 people measure you, the results will be different. When we fitted out our dress forms last winter I had three sets of measurements of myself, taken by my mom, my step-daughter, and my DH. They were all different and I do not believe this is unusual, even if you use people who sew regularly to take the measurements.

So if people aren’t accurate at measuring, who is? Well people can be trained, but different
body types are easier or harder to measure.

What about digital scanners or whole body measurement
systems? I used to think these would be
the wave of the future but they have been less than successful. First of all, for manufactured clothes, you
lose all of the advantages of mass-production if you have to cut garments
individually. Things automatically
become much more expensive. Secondly
even digital body scanners may or not be all that accurate. Each person’s measurements may appear
different depending on the way they are standing when they are scanned. Also, different undergarments or lack thereof
will affect they way the body is measured. And last but far from least, the amount of body fat will definitely
affect the fit of the garment. This is
probably why these systems have been most ineffective with larger clients. In order to measure accurately measurements
must be taken in reference to some structural point on the body (shoulder
point) etc and on a human body, these usually relate to the bones which make up
our sub-structure. But body fat can make
it very difficult to find the underlying structure, even when you are measuring
by hand and are able to palpate the body for its underlying structure. A
scanner, unless it is also an MRI (not likely) will not be able to determine
how the person is padded referable to the underlying structure. Human bone structure comes in a very small
set of variations compared to the variations in human bodies because human
bodies have a widely varying degree of padding. If you have a total measurement of body mass, but you don’t know how
that mass is distributed relative to the underlying bones and joints, you are
still going to have trouble making a pattern to produce a garment that fits
that person and is able to move with that person.

Once you have a set of numbers representing a person’s measurements, you get to the entire issue of producing patterns
from the measurements. This takes a
highly skilled programmer with experience in three-dimensional modeling and
converting complex mathematical systems into working programs. It also takes a highly technical
understanding of the human body and how the measurements you have taken
actually relate to the body and the way it moves. Thirdly you need outstanding pattern making
skills to understand how to make a pattern to fit that body using all the other
mathematical data you have incorporated.

Sounds impossible doesn’t it?  It is not, but hopefully you can appreciate
why we don’t have anything that does this yet. A program that would work, say even 75 % of the time, would take thousands of
thousands of man-hours of work by highly skilled and therefore highly expensive
people. It would simply cost too much,
and it would probably cost too much for most manufacturers to implement the
program and provide custom–fitted clothes because most customers would not want
to pay the price. And I am not convinced that you can really model all the
extremes of the human-body spectrum. But
with a population in the billions, even the relatively small percent of people
at the extremes can number in the hundreds of thousands or millions. You can’t get custom fitted jeans for
Wal-Mart prices. And I think even $100
for custom fitted jeans is too cheap once you figure in the overhead and time
spent cutting, sewing, shipping, selling, etc. It is just significantly cheaper to cut large quantities of the same thing. And most people want good salaries and cheap
clothes. In this case it is a lose-lose
situation. This would be even more true
for sewing patterns, which are a smaller market.

Years ago, when I was writing CAD programs my DH thought I
should write a program that would model my body measurements and create
patterns. I thought about it
briefly. But I realized that I don’t
have the pattern making skill to write a program that could translate any
desired shape of garment into a pattern that would fit me. Most importantly it would take me a LONG TIME
to write such a program because you would have to write all the possible
variations and allow for them to adapt to future design ideas and body
shift/weight change. OK CAD programs
are more advanced now and I wouldn’t have to do it all myself, but even so it
would require more programming that I want to do. I figured I can buy a pattern
for a shape I am not familiar with, alter it to fit me, even with muslins, and
make hundreds of garments before I would ever get that program perfected, and I
would probably be tinkering with it the rest of my life. Leave these things to the people who live to
program….me, I’d rather be sewing.