I have started weaving the crackle threading as overshot. This is the first time I have ever tried this and I am not totally unhappy with it. I still prefer other treadlings, but I can see some possibilities. Here are three different interpretations:
Each of the three uses 10/2 pearl cotton thrown twice with each shot for the pattern weft and 20/2 pearl cotton for the binder thread. Each of the three uses different colors. And each of the three uses a different technique for the binders.
Reading from the bottom up (as if you were at the loom), the first one (blue and red) uses all the remaining treadles for the binder, one at a time. For example, if treadle 3 is the pattern, as it is in the first block, then I treadled 4 for the first binder, 5 (after the next pattern shot on 3) for the next binder, 6 (after the next pattern shot), and so on. I thought this treadling would be the easiest to keep track of, but actually it proved to be the most difficult.
With the middle blocks (light khaki and red), for the binders I used treadles which raised opposite shafts, but never the pattern treadle. In other words, for the first binder treadle I selected one which raised shafts which were not opposite to the pattern shafts. But for the second binder treadle I selected one which raised shafts which were opposite to the first binder treadle. This treadling turned out to be the easiest to keep track of.
For the last I used three different treadles for binders. The first time I threw the binder weft (after having thrown the pattern weft), I treadled the pedal which raised the shafts opposite to those raised by the pattern weft. The second time I threw the binder weft (again, after having again thrown the pattern weft), I treadled a different pedal. The third time I threw the binder weft, I again treadled the pedal which raised the shafts opposite to those raised by the pattern weft. The fourth time, I treadled the pedal which raised the shafts opposite to the shafts raised which I treadled for the second binder weft.
This last sampling intrigues me. The warp is red, but I used two different colors for the weft: blue for the pattern weft and gold for the tabby weft. I rather like the effect this has on the areas where there are no pattern blocks. For me, this creates a bit more interest.
These three treadlings are those recommended by Zielinski. And, he says, the weaver needs to try each one and see what works best for what he is trying to accomplish.
The question is: why didn't I just take the simple way and treadle pattern and then binder, using always for the binder the pedal which raised the shafts opposite to the pattern? This is what I would do with regular four-shaft crackle. But with this kind of threading, doing that would create a curved kind of M's and O's effect.
By the way, looking not too closely reveals a treadling error where I got the binder wefts wrong. It occurs in the middle sample, the fifth treadling group from the bottom. In case that is kind of hard to see, here is a blow up with arrows where the error is most clearly visible:
At the arrows in particular, you can see where the fabric just kind of collapses because the binding thread did not interact with the warp threads there.