I did some Photoshop simulations of cross-dyed double katazome – my shorthand for “katazome on the warp and on the finished fabric”. This simulates what happens when you do katazome on a cellulose warp, weave the fabric with a protein-fiber weft, and then do katazome on the finished fabric using acid dyes (which will dye the protein fiber weft but not the cellulose warp).
These simulations are, obviously, just simulations. It’s hard to know exactly what the finished fabric will look like, since crepe weaves crinkle up when washed and real fabric doesn’t look like pixels on a screen. Nonetheless, these are encouraging! Here are some examples using a split complementary color scheme with different structures.

In this one, I used a 2-2 twill structure, made warp and weft slightly different colors in the background, but kept the two colors identical to produce a “solid” color in the feathers/leaves. (Click in for a close-up; it’s hard to really appreciate the pattern except at full size.)

In the next one, I changed the structure to a goose-eye twill, cranked up the contrast in the background slightly, and added a trace of contrast in the feathers/leaves (click to open the larger version as the pattern is more obvious there):

Next I went a little bolder on the motif, using draft #31525 from Handweaving.net (Page 262 Figure 5: Donat, Franz Large Book of Textile Patterns, Germany, ca 1895)

I thought this was actually a bit distracting, so I tried a crepe pattern from Alice Schlein’s A Crepe Is Not Just a Pancake monograph – also available as draft 63090 from Handweaving.net:

This is one of my favorites!
I then decided to try some secondary triadic color schemes:


While the actual fabric will likely look fairly different from the simulation, this looks like a very promising path to explore!
So far I have mostly used subtle patterns, with relatively low contrast, that don’t distract from the images. I think that, with detailed images like these, that is the right way to go. However, those are not the only options. Remember my example from a previous post:

Here the figures are very simple and the interaction between pattern and color dominates.
I plan to try examples of katazome with both simple and complex figures over the weekend, and will report back with results.
Meanwhile, regarding the Celtic Braid Coat buttons:
I hadn’t thought of using no buttons at all! I will investigate that and see if it’s practical and looks good.
I had thought of using polymer clay buttons. The trouble is that polymer clay buttons cannot be dry cleaned and this vest (being leather) is emphatically dry clean only. i know it’s done, but I don’t think it’s a good idea to use findings whose care requirements contradict the garment’s care requirements. Yes, you can remove the button for dry cleaning and then reattach it (and a good dry cleaners will do just that), but if you forget once, you’ve just ruined your button and possibly the entire garment. Better to avoid that route, I think. But I might be able to do something similar with silver clay and an enamel-like pen, and I’m considering that possibility.
I had thought of using self-fabric as a button, but early experiments were not promising. The trouble is that a button is round and the braid runs straight, so the motifs clash. A four-cornered Celtic knot might work, but I can’t weave one that looks good on only 24 shafts (even if I had any of the warp/weft remaining). I’ll continue to mull it over, though.