More reposts. Material originally appearing on 17 June 2004.
More reports on the great filet experiment.
First, thank you to Gayle Roehm, Judy Gibson and BJ Knitslikecrazy. Gail and Judy sent me info on a method shown in a Burda Magazine. Burda’s instructions were for a set of curtains in a cyclamen pattern. They offered up the pattern graphed like any standard filet crochet or darned net design, and in prose gave directions on forming both filled and open squares. Gayle sent a photo of her project knit from those directions:
Burda’s method takes four rows to complete a tier from the graphed chart. On the first row, spaces are formed by a YO, K2tog unit. On the second row (the purl side) the K2tog is purled, YO is slipped, and another YO is made. On the third row (knit side again), both YOs are slipped and another is made then the purl is knit. On the fourth row each space starts with a P1, and then the three YOs are purled together. That makes a mesh with heavier verticals than horizontals, but the mesh is more or less square. Not as delicate as filet crochet (which in turn is not as delicate as the darned net or withdrawn thread family of embroidery techniques). Here is Gayle’s note – reprinted with her permission:
A couple of years, I too had a bee in my bonnet about knitting that
looks like filet crochet, and I tried three methods:
a) Burda method. I actually knit the cyclamen panel — a partial scan is
attached. Comments follow.
b) Mary Thomas method.
c) Sandy Terp method.
None were quite satisfactory, though all work but only sort-of. The
problem is that you don’t get the nice squares and open holes that you
get with 3 dc, 3 chain. The solids are okay, but the holes and bars are
The Burda method is stockinette-based. From the scan, you see right away
that (a) the holes are still round, and smaller than the closed squares;
and (b) the vertical “lines” are thicker than the horizontal “lines,”
making the squares even harder to see. I think each square of the chart
accounted for two rows.
The Terp method, which is garter-based, was only a little better. The
holes were marginally squarer, but the vertical lines are still thicker
— perhaps inevitable because you have to do a k2tog or k3tog. And one
square was (as best I recall, but I’m probably wrong) three or four rows
— slow. If you don’t have Sandy Terp’s method at hand, I think she has
it on her website somewhere, or written up in a leaflet. <snip>
The Mary Thomas method was also unsatisfactory for reasons similar to
Terp — the details escape me.
BJ Knitslikecrazy sent in mention of a style of filet knitting in a needlework technique omnibus called Stitch Wise. The description she sent sounds a lot like the Thomas method – with 3 stitch by 4 row units, with two half-height open spaces stacked to balance the solid knitted squares.
Here’s my swatch. The bottom several rows of garter stitch and bars is the Thomas method. The top checkerboard is my method:
As you can see, the Thomas method first few rows is interesting, and can be done following a graphed chart, but it just doesn’t have the filet look.
The upper part is my stab at doing it differently. Some of the vertical bars in the lower part are sloppy because I was experimenting with several methods of making them. I’m not entirely pleased with the later methods, but progress is being made. In the Kim-method, two rows make one tier of the graph.
Solid squares are formed this way on the first pass (knit side):
On the returning purl row, all of these stitches are purled.
Note that groups of solid blocks can be ganged together. If for example, if the chart shows three solid squares in a row, the knitter would do this: K9, turn; p9, turn; K9.
Spaces are formed this way on the knit side pass:
(YO)3x. Retain 3 loops on needle.
Slip one stitch as if to knit. Slip the next stitch as if to knit. Pass the first slipped stitch over the second and off the end of the needle (sort of a no-knit bind-off).
Slip one stitch as if to knit. Pass the previously slipped stitch over this one and off the end of the needle.
(Reinsert the left hand needle tip into the stitch at the end of the right hand needle. K1.)3x – This makes a 3-stitch vertical “chain”
On the returning purl row do this on each space:
P2tog, K1, P1
Problems with my method:
- I don’t like the way the bind offs in the open spaces pull away from the previous solid space.
- I don’t like the relative thickness of the horizontals and verticals. My verticals are thinner than ones made by decreases, but they’re still thicker than I’d like.
- I don’t like way there’s a little vertical slit left when a solid square follows a space.
But I’m getting there… Constructive criticism and idle thoughts graciously accepted!
i love your blog! thanks for writing to the laceknitters yahoo group, so i could find you. your blog is worth studying! you’re an inspiration. do you have a degree in fiber arts?
btw, years ago i tried the Mary Thomas filet knitting and was so disappointed i never tried it again. your results here in the pink cyclamen example are stunning – from the perspective of my own experience. my result looked more like your lower white swatch. i also tried my own method, which at least produced some open spaces, irregular though they were!
i hope to check back to your blog regularly, and learn a thing or two! (*smile*)
Thanks! The cyclamen panel belongs to Gail Roehm who kindly shared a picture of her work with me, then gave permission for me to put it on the blog. That panel is from Burda, and she did it in the Burda method, not the Mary Thomas method. I haven’t tried that one yet, but it did give me some leads on the method I’m trying to develop on my own. In the mean time, I’ll keep experimenting! -K.