Everything about it seems OK; it has firm hairs with soft tips, has been cleaned thoroughly, and 'feels' OK to the touch. But it just doesn't work - pigment gets bunched up here and there on the block, and just won't get spread out properly.

Last time I visited Kubota-san, the very experienced older printer who is doing some of our Ukiyoe Heroes printing, I took it with me for his inspection, and he suggested that it perhaps had too much hair. The maker had perhaps jammed a few too many hairs into each of the holes when making it. And when I compared it with some of his (excellent) brushes, it did seem too 'tight'.

So when I got home that day I tried a little experiment. I took the cap off the brush, to expose the wire that ties all the bristles in place, and marked around one third of the holes with a black dot.

I then squeezed a dab of epoxy glue down into the neighbouring holes, so that the hair wouldn't all fall out after what I was about to do. (The entire brush has all its hairs tied in place with a single length of wire, that threads form hole to hole - and if this gets broken in even one place, the thing will eventually all unravel ...)

I then took a drill and reamed out each of the black-dotted holes, severing the wire and causing all the hair tied into that hole to fall out the bottom. (The photo above was taken just after this was done ...)

After I was done, I put the cap back on, and ended up with this:

And the result? A huge improvement! It has a much lighter and 'open' feel, and the pigment is spread evenly.

So ... having this experience under my belt, I made a trip yesterday to one of the last surviving brush makers in Tokyo, to see what they thought about perhaps making some brushes for us with less hair.

They weren't very excited, and when I asked them if they could make a brush for me with fewer clumps than the normal pattern, they said that it wouldn't be possible. All their 'kata' - the pattern jigs they had inherited from previous generations - were based on the traditional layouts, and there were no craftsmen left who could make new ones.

I asked to see what these looked like, and they brought one out to show me:

This is nothing! It's just a piece of very thin brass, with a bunch of holes punched in it. "No craftsmen left who could make one?" Are they kidding me?

It's an interesting demonstration of the way that these traditional craftsmen think. I looked at that and thought to myself, "Give me a half-hour with a few rudimentary household tools, and I can replicate this with no problem ..." but they just can't make that leap. This was prepared for them by a 'specialist', so they wouldn't dream of trying to make one themselves. It's the "this is the way it has always been done ..." syndrome, and they just go blank when faced with any kind of dead end.

So anyway, I asked straight out, "If I can produce a kata with a pattern that I would like to investigate, could you make a brush based on it," and they replied that it would be no problem. They would put the pattern on a piece of wood, and drill and sew to match.

You know what comes next, of course - once I got home and got started, it took about an hour in all:

I designed my hole pattern in Photoshop, printed it out on paper, lightly glued this to a piece of thin brass sheet, then drilled it and cut out the outline. My new 'kata' is identical in structure to the old model, except that the old one shows caliper marks where the guy must have laid out his dot pattern manually. (And of course, now that I see how their system works, it's obvious that this brass kata isn't an essential requirement at all - it's only used to transfer a required dot pattern to the blank wood. You could print the pattern on a sheet of thin paper directly from Photoshop, paste that onto the un-drilled wood surface, and proceed from there. Each and every brush could be 'customized', if that were needed ...)

Anyway, I sent this brass one off to the brush maker last night, and we'll see what comes next ...