Tiramisu- a dress cut (mostly) on the bias
When this question comes up in classes, I used to reply “The short answer is, no. Knit fabrics behave differently to wovens, only wovens have a true bias.” That’s not inaccurate, as far as it goes. The long answer is a little more nuanced, and it’s taken me the better part of a year to satisfy myself about whether knit fabrics have a “bias” grain or not. I think they do, even though the nature of the fabric is different than a regular woven.
I run a few regular searches on google, to find any new blog or forum posts on the topic. I never found much information online or off about whether knits have a bias grain. Opinion varies widely on the topic of knit bias, so I knew I would need to experiment on my own to find an answer that satisfied me.
Earlier this year, I made the Bow Tie Tee, a hack of the Blank Canvas Tee and also available as a pdf pattern. The jersey I chose to use is a relatively beefy cotton t-shirting with moderate stretch. It’s pretty stable, a good place to start playing with knit grainlines. When I worked with the pattern, I didn’t approach it as a “woven” or a “knit” pattern, but as a kind of hybrid. For most woven patterns, you want at least an inch or two of wearing ease to allow you to move and laugh and jump around. Most knit patterns tend to have “negative” ease, which means the fabric is cut smaller than the body’s measurements so the fabric will stretch and cling to the body. For the Bow Tie Tee, I worked with 0 ease- neither extra fabric like a woven, nor “negative” ease like a knit.
I did this because I wanted to play with stripes. Why else? The front yoke is cut with horizontal stripes, and the stretch runs the direction it should in a typical knit pattern- that is, the stretch lies around the body. For the lower front piece, I turned the fabric on its side. I wasn’t sure this would work, because it’s not usually the way knit fabric is used. Turns out, it works just fine- probably because I only did this on one section of the garment, not the entire shirt. This is one of my favorite tops, I’ve been wearing it for months now and I find no problem with the cut or the grain. It doesn’t bind or ride up, in fact when I put on this shirt I cease to remember I’m wearing clothes. That’s always my aim.
The back uses another kind of grain altogether- a “bias” grain. I did this because I have an undying love of back interest and chevron stripes. Again, I had no idea if this would work because it’s not usually done, but it turned out well. I noticed that the back seems to have a different kind of “give” and “hug” than a regular knit top. When I wear the shirt backwards for a plain v-neck front (this works well!), I notice the fabric molds and skims over my bust and waist in a pleasing way.
front, “Lamington,” a work in progress
I tried a similar back bias treatment when working up samples for another upcoming pattern. (I make and wear samples extensively before working them into a full multi-sized pattern.) This time, the front is quite plain with the stretch running around the body as it would on a normal knit pattern.
I left this one unhemmed to see the effect of the CB downward “growth” more clearly. You can see that the back has indeed grown longer than the front, and as it stretches downward it gently pulls the side seam to the back. This is quite clear when I lay the garment out flat, but when I put it on my body fills the shirt and the seams lie where they should. The back “bias” pieces behave almost exactly the way I would expect a woven bias cut to behave, but with the knit fabric the bias effect is somewhat more pronounced. Interesting!
The stripes show you clearly that the side seam is cut on the “straight” of grain, and the CF is “bias.”
I took it to the next level with my samples of the Tiramisu dress. The skirt is cut with the straight of grain at the side and a “bias” seam at the center front and back. When I first started playing with this concept, I had no idea if it would work in a knit. I know that “straight” sides and “bias” CF and CB seams works to make a very flattering silhouette in a woven, and I was keen to try on a knit.
thats some good rippling… The jersey is one of those “modern” jerseys that stays wrinkled… Not sure I love the effect… But I do love this dress.
I’m happy to say it does work quite well on knit fabrics. I’ve tested it on rather light weight (striped version), medium weight (several muslins), and even a double-layered jersey (red version) to see how the fabric behaves when treated in this manner. I couldn’t be more pleased with the gentle rippling effect, though I have discovered that a knit bias dress should be hung unhemmed overnight so it can “settle” the same as a woven.
Tiramisu front bodice piece
The bodice for Tiramisu was also interesting. I liked the stripe placement on a similar woven dress (1950’s pattern I made for Mother’s Day) and hoped it would work for a knit. It does. However, I discovered that due to the way the stretch lies on the body (nearly vertical) and the weight of the skirt, the top bodice section tends to stretch.
Made of a heavy weight double layered jersey, works quite well.
Further, I discovered that it’s more or less a standard amount of stretching regardless of the weight of the fabric used. Lighter fabrics usually stretch more than heavy ones, but the lighter dresses don’t weigh as much as the heavier ones, so it seems to come out even. In this case, we’re working with the nature of the fabric (stretch), the weight of the fabric, and gravity.
This means that for the underbust seam to lie under the bust, the pattern itself must be quite a bit shorter that the shoulder-to-underbust raw measurement. If I hold the upper bodice pattern piece up to my own body, it doesn’t seem like it will work. However, comprehensive testing and understanding how the fabric will behave has shown me how to produce a consistent result.
I also use a carefully-calibrated piece of neck binding to help “snug up” the neck opening to prevent tumbling out of the dress. I hate having to think about my clothes after I put them on, especially having to worry about that particular issue. The short binding works well for keeping the neckline in shape without using another kind of stabilizer. The binding itself is cut on the cross-grain (without much stretch) and about 3/4″ shorter than my neckline opening. I eased it in, and the result is solid and light.
I have satisfied myself that knit does have a “bias.” While the nature and structure of a knit fabric is quite different than a woven, I can not deny that knit fabrics behave differently on the body (and with a pattern) when cut on the bias than when cut “straight.” Furthermore, I think that when a knit is cut on the “bias,” it behaves in much the same way as a woven bias, but more so.
What do you think? Have you ever experimented with knit bias? Do you know of a great blog post, article, or book on the subject? I haven’t really found much that was helpful, just a handful of forum postings and a few about.com pages that rambled about knit fabrics. I’d be quite happy to hear any and all thoughts on this!
(Also… I have some fun sewing for Stephen coming up… I made him some long-sleeved linen work shirts years ago, he gets compliments from the other ecologists whenever he wears them and has requested a few more! Cool. I’m planning to use Negroni, I’ve been dying to make it up! And Lila could use a few new pieces, I’m designing some girlie versions of my ladies’ stuff to try on her. Should be fun!)