Weft: Candy silk
Brocade: Spun silver (devere) and gilt passing thread (Hedgehog handworks)
Soumak: Red silk (devere)
Cards: 23 pattern + 2x7 border
I think this pattern is going to be my nemesis. As you may recall, I have already woven it twice, after New Zealand Post managed to lose it the first time. At the time, pretty much all my information on the band came from Hansen. However, since then I came across Lise Ræder Knudsen's article Brocaded Tablet-Woven Bands: Same Appearance, Different Weaving Technique in NESAT VII, which talks about this band and also some others in a similar style. She has a quite different idea about how the band is constructed:
Hansen says that all cards are threaded with silk and the band is executed with "card idling"- ie turn the odd-numbered cards in one pass and the even-numbered ones in the next. Knudsen says that all 4 holes were threaded, but that excepting the 3 tablets on each border that never show brocade, alternating holes were threaded with a vegetable fibre (probably flax) which has since completely deteriorated (leaving the remains indistinguishable from a 2-hole band- a theory put forth by Collingwood).
The cunningness of this plan is that your band ends up looking like it's made entirely expensive silk, when really a lot of it is just linen! Super cunning. There are some further steps you have to take to make this effect work:
- The tiedown should always be one thread only (and always the silk thread of course!).
- Brocade floats should always cover an odd number of cards. This means your tiedown will always be the right type of thread. My original Mammen pattern was mistranscribed from Hansen and had some floats over 4 cards, so I had to adjust it.
- In areas where there are more than one adjacent tiedown (ie the border between the gold and the silver in the pattern) the tiedown that is of the wrong fibre can be hidden by covering the tiedowns with soumak. I've never done soumak before. There are some descriptions in Hansen and Collingwood but no pictures, and in Gertrud Grenander-Nyberg's artivle in NESAT IV, Soumak Techniquie in Swedish Medieval Textiles, there is good information including diagrams on soumak in the context of (non tablet-) weaving. Hopefully I did it right. I can describe in more detail of anyone's interested (Conversely, if there's already a good description online I'd like to know)
I'm pretty sure the problem comes down to my choice in gold thread. I tried hard for this band to use high quality materials which is why I use the gilt passing thread. But it's just not nearly flexible enough. If you zoom in on the picture you can see that the gold thread always comes in pairs, two close together, then a gap. That gap is the inside of the loop where the gold thread turns around. It won't corner any sharper (without resorting to special tools or something, anyway). Passing thread appears to be completely the wrong choice.
I've bought myself some tambour thread, which is meant to be more flexible, and am going to have a test to see if that's any better. But it's quite thick, and I don't like my chances. It seems that if I want a dense weft I'm going to need to stick with naff synthetic spun gold.
So I'm going to need to weave this band one more time, once I've sorted out my issues with the brocade thread. I will leave it for a while though, in case some other article comes along with more insight into its construction :-)