Motivations
I like garments to be kind of effortless. Easy to think about, easy to construct, easy to put on and wear. This is not just to give an idea of my preferences, it’s a synthesis of my objectives when designing clothes:
Ease of thinking comes to me from a concept of pattern making that allows, when looking at the base pattern, to visualize in a single step how the garment is going to fall on the body.
Ease of construction comes from minimizing the number of steps during construction, removing unnecessary seams and features. It’s partly a consequence of ease of thinking.
Ease of wearing is perhaps the most complex to achieve, it depends most importantly on having the correct choices in fabric, functionality and style for the environment the garment is meant to be used in, but is also necessarily a function of how the garment fits.
This article is about a method of pattern making that tries to satisfy the first two conditions, while keeping the third in consideration.
Why traditional patternmaking doesn’t fit this approach
Traditional pattern making techniques split the garment into a minimum of three tubes of cloth, with two pieces folded and sewn into cilinders for the arms, and two body halves forming another tube that the arms then attach to.
It can get much more complicated when talking about something like a tailored jacket, where everything is split into more parts to provide more seams that can be used to build a more complex 3D shape.
There are good reasons to build clothes this way: the human upper body can be approximated to a set of tubes, and splitting a pattern in more parts allows more seams for shaping and more versatility in cloth usage when cutting out the pieces.
To me, the most simple t-shirt pattern doesn't feel simple enough though; while an understanding of the way the set of shapes is going to transform into something that goes on the body is kind of intuitive, it still requires multiple shapes taking multiple paths, and sewing them together in steps that are sequential.
Compare that with a poncho for example. When looking at the pattern for a poncho, understanding of how it's going to fit the body is immediate. This is what I call true ease of thinking.
Conic Projection Patterns
To get myself closer to the objective, I wanted to start from the most basic possible way to wear cloth, something even simpler than a poncho, which is simply draping a rectangular blanket over the shoulder.
Observing how the blanket behaves around the body gives a lot of important information, in fact it gives all the information that I needed for the next step.
The first observation is the fact that there is an excess of cloth around the neck, this comes from the fact that the neck wants to sit at the centre of the rectangle and so is pushing against the side of the blanket and bunching up the cloth.
The second observation is that because human shoulders are not square but sit at an angle, the blanket curves in a sort of conical shape around them, so that when we look a where the "cross" formed by arms, center front of the body and center back sits under the cloth, it's not really shaped like a cross, instead the arms and the two halves of the center front are angled backwards.
Going straight from this to an actual pattern is very direct, one simply needs to add a neckhole in the center, remove the excess cloth between arms and body (in the process creating two side seams that allow the garment to have actual sleeves), and to straighten the connection between the two center front halves, so that they can be pulled together with either a seam or some closure system.
And there we have a conic projection pattern. It's called conic because the only real shaping is in the cone it forms around the shoulders when closed, also it looks a lot like a conic map projection.
Sewing it together is easy, and the way the pattern will go, from a flat piece of cloth, to wrapping around the body, is very intuitive to me.
Parametrized model
The fundamental measurements for this type of pattern start from the neck.
Theoretically, they should start from the center of the neck, but sticking the measuring tape inside the neck is not possible so a practical compromise is used.
Sleeve length is measured directly from center back neck to wrist, since there is no shoulder seam.
Back length is taken from the same point.
Front length is either taken from around the adam's apple, going down to the hem, or just chosen in relation to back length.
Shoulder slope angle can be chosen to perfectly approximate the target's shoulders, but it also influences grain alignment and can be modified to adapt to that, this can be really important as grain alignment is challenging with this method of construction.
As for other measurements like the chest and waist widths, and sleeve openings, they work exactly as in traditional patternmaking.
Additionally, the seam/opening between the center fronts can be moved to the back to achieve a seamless front when making a garment that doesn't have a front opening, like a t-shirt. The flow of the cloth is inverted, with the cloth wrapping around the body front-to-back, but all the measurements work exactly as before.
Also, if the available cloth isn’t big enough for a single piece, or incompatible grain alignment is wanted between front and back, the pattern can be split at the back, merging the back halves in a single piece that is attached to the front half before sewing the side seams.
Obviously, the basic pattern can be split in many more ways, but using too many pieces kind of defeats the original purpose of this method.
Characteristics
My conditions for ease of thinking and construction are satisfied.
The run of cloth over the shoulder, where the whole garment hangs from the body, is seamless. This is comfortable and gives a very relaxed feel to the piece.
A bit of shaping can be done by playing with the side seams. I prefer "explicit" shapers though, for example a half belt to tighten the cloth around the small of the back.
Grain alignment is fundamentally a compromise, since various parts of the single-piece cut are at an angle to each other, some will necessarily be at an angle to the grain. The basic block can be split in various ways to compensate, but another interesting option is to use a 22.5° shoulder slope angle, which results in a 90°(4x22.5) opening between the center fronts. This can be really useful because, when using fabrics with similar behaviour between straight and cross grain, the pattern can be angled so that a single piece cut has bot the center front halves on the two grains with the back piece on a perfect bias, providing a clean, straight look on the front, with very comfortable forward arm movement thanks to bias elasticity on the back.
This method is not that suitable for making tight garments, it may work when using fabrics with good elasticity, but the limited shaping available through seams doesn’t allow to follow the body very well.
The usage of striped/checked/patterned fabrics poses the same challenges that are already present with grain alignment and more.
Overall, this not a replacement for traditional methods, but an interesting alternative for a specific look.
In the future, I think most builds on this blog will be based on variations on this concept. The first one to come will be a short sleeve henley, using the paper pattern at the top of the post.