Posted by Dave Bull at 10:31 AM, December 27, 2008 [Permalink]
Continued from [Seacoast in Winter - 2] | Starting point of the thread is [Seacoast in Winter]
Yes, about the circles ... it's kind of a long story.
When I was doing the printing for the Forest in Spring print (Chapter 11) some months ago, I had a very difficult time with the registration. I just couldn't get the tree trunks at the top of the print and the background to 'line up' properly, and the patterns on the forest floor also just wouldn't fit into place. When I lined up one side of the print, objects at the other side would be out of registration.
At first, I put this down to stretch in the paper due to the large number of impressions. When I printed the second batch of 100+ sheets I tried to be more careful with the paper condition, but still had the trouble. No matter how I shifted the paper in the registration marks, I was unable to get everything properly in place. The forest floor on the right half of the print is sharp, that on the left side is blurred, although I doubt that anybody would notice anything wrong by looking at their print.
A couple of month later, when doing the print of the moon (Chapter 3), registration was also trouble at first. I cut the registration marks very carefully, because I really wanted the smallest craters around the perimeter of the moon to be sharply delineated, but had a lot of trouble nonetheless. I was able to solve the problem that time, because the critical part of the design is in the center of the paper, and the margins at right and left need no registration at all.
I was annoyed at myself for the carelessness again, put it down to over-confidence, and made a promise to be much more careful with the next print.
But when I did the first test printing on the next one - the River in Autumn (Chapter 4) - it was even worse. I had actually carved a keyblock for that design, but the registration was so bad - off by nearly two millimeters from side to side of the design, that I had to discard it. And I was only saved from total disaster by the fact that the circular black over-printing covers all the details at the far left and right edges of the design.
Something was clearly wrong, and I finally realized that it wasn't me. At least I mean it wasn't just me being careless at printing time. The blocks themselves were inherently wrong. Something was spectacularly wrong with my hanshita preparation process.
The first three prints in the series - Chapters 1, 5, and 6 - had all been made in the traditional way, with the cutting of a key block first, then using that keyblock to print 'kyogo', colour separation sheets. I had no trouble with registration at all.
But these last three prints were not made that way. For these, I prepared the colour separations completely in Photoshop, printed them out with my laser printer, and pasted them onto blocks for carving. Photoshop couldn't be the problem, as everything there was controlled right down to the last pixel. The printer just spits out the correct data. But my pasting technique is very well honed after all these years - since some disasters very early on, many years ago, I have been rigourous with this.
Clearly though, one of those three steps must be the problem, so this time, I tested and checked at every step, to try and isolate what was causing the trouble.
The Photoshop data was of course perfect, - each of the colour separations in the master file had (of course) exactly the same dimensions. I printed the first one out, and then grabbed a ruler to measure it.
Bingo! Problem located ...
Not only did the dimensions of the printout not match what was in the original file, the printout wasn't even perfectly rectangular. It was 'twisted' and distorted by around 1.5 millimetres across the length of the image. I printed another one. Also twisted, but in a slightly different way.
I have two laser printers here, a small Canon I use for my invoicing, and a larger A3 Epson I use for printing the pages of the books for this series. I experimented back and forth with both of them, and found the same thing. They both were producing distorted output. It seems that the paper twists and turns ever-so-slightly as it moves through the printer and there are always variations in the finished sheets.
No wonder I had had so much trouble with the registration on the previous three prints; they had been 'doomed' from the start, as the distortions were carved into the blocks!
The solution seemed clear, just use a high-quality printer. I took the file down to my local printing company, and got them to run off a sample sheet on their room-sized giant digital machine. Same problem! It seems that this problem is inherent in copier technology. There is no way that the paper can go through all those rollers and twists and turns without the image becoming ever-so-slightly distorted, even on a top of the line professional machine.
So, what to do? Giving up the Photoshop separations, and returning to the traditional method of keyblock and kyogo would obviously work, but for this type of design, that was not a practical solution. I don't want outlines, and I do want the textured surfaces. So I need a type of printer that will output the Photoshop data without distorting it.
I also have a large (A3) inkjet printer here, but have never used it for hanshita preparation. Everybody 'knows' that output from these printers is susceptible to smearing when it becomes wet, and if I used this printer, of course the details of the design would instantly become smeared and lost when the sheet was being pasted down onto the wood. And could the sprayed output have anywhere near the sharpness and clarity I had been getting with the 2400 dpi laser printer?
But I had never tested any of these things, so I tried a small sample. This printer is quite an expensive model, and uses what Epson calls 'pigmented inks'. When I tried pasting down the sample swatch, it didn't smear at all. As for sharpness, there are dozens of settings available in the printer driver, and I experimented until I found the one that gave the clearest output onto my hanshita paper.
And as an added bonus, this printer also has a slot in the back, so the paper can be fed through it without bending around any rollers at all.
And here - long story, I'm sorry - we get to the point of the circles, which by now, you can presumably guess. To minimize any possible chance for distortion in these sheets, I have broken up any wide areas of solid colour, to avoid having the liquid ink cause any swelling in those areas.
It all seems to work. The first test sheet showed no perceptible difference in dimensions from the original file, so I went ahead and ran off the whole batch, checked each one carefully, and then pasted them down ready for carving.
At this point there is nothing I can do to 'fix' the blocks for the previous three prints, and any future editions taken from them will have to be done with the same struggles, but hopefully, this next one will come out clear and sharp, from edge to edge.
We'll see!
The thread continues in [Seacoast in Winter - 4] ...