2014-08-15

Copying a T-shirt

No doubt there's all sorts of advice on copying RTW out there. You can do a much more precise job of it if you want, and this can be well worth it for a very detailed garment. I'm copying a simple T-shirt, though, and I think a quick method can be good enough. (This post probably doesn't give enough detail for someone with little or no patternmaking experience; in that case I think it might be better to start with a less quick-and-dirty method anyway, so you can be more confident you've got an accurate copy. YMMV.)

First, you need some sort of paper. I'm using a roll of newsprint, but if you don't have a lightbox, you might want tracing paper instead. This is helpful for making the pattern symmetric later on.

Second, you want a hard, flat surface (ideally a cutting mat) that can stand being poked.

I rolled out my paper, then put my T-shirt on top, trying to get the torso as smooth and flat as possible, without being stretched.

This top turned out to be even simpler than I thought: the front and back were probably meant to be identical, except at the neckline.  So I tried to get the folds to be along the shoulder and side seams, and lined up the hems as best I could.


Next to the seam gauge, there's a needlepoint tracing wheel on there — very useful for tracing seams through to the paper underneath. You can use pinpricks instead, although it takes a lot longer. I don't like regular tracing wheels and tracing paper for this; they move the fabric.

The top was fairly skewed, so I couldn't line up the armscye seams all that well, and the back hemline was a bit shorter than the front (this can be done for a full bust, but I'm not sure it was intentional here). This happens in knitwear. Even woven RTW isn't perfect, but I think in knits it's worse. It could be a combination of inaccurate cutting, skewed fabric (if you knit, you can even get this in hand knits if the yarn twist is wonky), and any sort of asymmetric stretching that happened as the garment was worn. Just do the best you can.

If it won't stay put for whatever reason, you can put weights on, but I find cotton knits on matte paper don't cause much trouble. I free-handed around the side and shoulder seams, 3/8" from the folded edge, and around the hem, 3/4" away. Then I used the needlepoint tracing wheel to do the same with the armscye seams, feeling through the fabric to get the back armscye seams as well. The same thing was done with the front neckline.

(This photo was seriously washed out; I've tried to adjust it to make something, anything, visible. Sleeves are up top.)

The sleeves were done separately, but in a similar manner. These ones happened to be symmetric, fitting the symmetric armholes. I ordinarily consider this a no-no, but this top does seem to fit pretty well, so I'll give it a pass. Makes it easier to copy, anyway. There, I made sure the hems were lined up and the underarm seams on the fold, then did the same thing — freehand around the folded edges and hems, and traced the armscye seam with the tracing wheel. You may notice that with the sleeve flattened out, the armscye is a little longer than it was when the torso was flattened out. This is fine, as long as nothing's seriously stretched; it means there's a little bit of ease in the sleeve cap (which is a whole other debate I won't get into now). It's a good idea to trace both, so that you can average out any discrepancies.

By folding the torso piece in half, you can find out how asymmetric your pattern is. This one will line up OK along the side seam, and close along the shoulder seams, but one shoulder is noticeably wider than the other. The armholes were way off compared to everything else. The same comparison can be made by trying to match the right and left sleeve tracings.

After that, you basically want to trace this off, interpolating as best you can, to get a symmetric pattern. A lightbox or tracing paper will help here.

So here's the pattern.


I will have to copy off a front or back, and I'm going to recheck the shape of the armhole — the sleeve cap looks like it might be just a bit too long for the armscye, which might not be concave enough. But overall, the pattern looks more or less as I would have expected.

For knits, it's also a very good idea to measure the stretch percentage of the original fabric (how far a crosswise fold will stretch — say, from 4" to 5" is 125%, or 25% stretch. The pattern will fit similarly if you use a similar fabric, but could end up seeming quite a bit smaller or larger otherwise. Noting down the fiber content helps too, since different fibers have different recovery.

No comments:

Post a Comment