View Single Post
 
Old 07-20-2007, 02:36 PM
rlwhitt's Avatar
rlwhitt rlwhitt is offline
Member
 
Join Date: Jul 2007
Location: Clemmons, NC, USA
Posts: 274
Total Downloaded: 12.31 MB
Build: Up-Scale 1:20, Republic P-47D-25, Halinski 3/06

NOTE:This thread is an extract of my posts from the thread of the same name at the Zealot/CardModels.net site. I have grouped
them together more than they were originally where it makes sense, and edited a bit of extraneous text, in addition to using thumbnails
instead of full size pics. Some posts will still read a bit disjointed because of the time compression and lack of intervening posts
from other members of the other forum.



For my next project, I've opted to Up-scale a printed kit! I know that the vogue seems to be to build smaller, but I want to go the
other way and see what happens. So here I'm going to make a stab at building the Halinski P-47D-25 Thunderbolt upsized 65% to
approximately 1:20 scale.

I've wanted to tackle this kit as I really like the Jug, but it weighs in at around 1200 parts and I just could not see doing it at 1:33
scale. The Zero was bad enough! I think the level of detail of this kit will work out well at the larger scale, plus I'm tentatively
planning to add some extra detail where it makes sense (and is visible!), for example landing gear and wheel wells, etc. And I’ll
probably add some extra cockpit detail too as long as I can see it (No more totally gratuitous detail for me!).

As for why I picked 1:20 scale; if I'm going to do this, I wanted a big enough increase to make it worth doing. I toyed with the idea of
a double to 1:16 or so, but quickly realized that quite a few parts (skins mostly) would span more than 1 letter size sheet. One of my
criteria is to print at home to normal size paper, so the thought of joining a bunch of skin parts in the middle nixed that idea. Plus –
where in the world would I PUT such a monster! As it turns out, a 65% increase will cause me to only have to split one part - the main
wing skins - which have a pretty natural seam printed on them about midway out where I can make the joint.

For a rescale from a printed kit, there is a lot of preparation - particularly for an upscale. First there is scanning. I have been
looking at some old threads and in particular posts by Leif and Gil that have some good info for scanning and rescaling. So a big hat
tip to those guys because I'm stealing a lot of their info here!
The first thing I had to do was cut each page of the book roughly in half, weaving the cut among the parts, so that I end up with chunks
that fit on the scanner bed.

Scanning:

The important thing to know about scanning is to NOT allow the scanning software to assume you are scanning photos, or otherwise be
allowed to make "exposure" decisions. If you allow this, scans from one page to the next will not come out the same - will not match
color and darkness. My scanner (Epson) has a Document setting, and in addition I had to force it each time to turn OFF Auto Exposure,
even though I had picked Document.

One reality about scanning seems to be that the color and darkness you get from a scan and subsequent prints WILL NOT exactly match the
kit. Perhaps with really high-end equipment this may be possible, but with the type of stuff we have at home, my attempts seem to
indicate that you will frustrate yourself to insanity if you try. The important thing is to get consistency between the scans so things
match and to try to just get reasonably close on color. For this kit I really only had one noticeably bad result and that was with Reds.
This model has a red cowl and rudder, and these were coming out more dark, almost rust-colored. So for these parts only I rescanned that
page and allowed the software to do auto-exposure and they came out a good bit brighter. So I will just substitute those parts only.

About DPI (dots per inch). I was tempted to use higher DPI settings (like 720) to help bring out more of that fantasic Halinski printed
detail. I could, in fact, see differences in greatly magnified sections at higher DPI. But printed, not so much. And the file size
increase is HUGH! (doubling DPI quadrouples the number of pixels). Load/save/editing times dictate that you not go overboard on DPI!

To summarize, my settings for my Epson scanner are:

Type: Document
No Auto-Exposure
Arrange the input size to just fit each page fragment (using the crop box)
300 dpi
165% Size
Save to TIFF files (do NOT use JPEG!)

Editing:

As long as you get a good scan using a document setting, you might have the resulting files’ color pretty like you need them, but one
tip I got from reading the experienced guys' posts at this point was to ensure that your files have their “white point” adjusted – again
to help make sure that there is color consistency from one file to the next. The process here would vary in details from one editing
program to another. In Photoshop, you open up a file, select Image/Adjustments/Levels, click the Set White Point dropper, and click
somewhere in the “white” area between parts. If the page background is a little off-color, everything will be adjusted until that point
you clicked is white. In my experience scanning this model, there was rarely any difference made when I did this, as they were quite
close to white already.

One thing I did after setting the white point was to save the files as Photoshop PSD files. This made the files about half the size and
thus allows faster load/save times. You just want to make sure that at no time in the process you allow the files to be saved in any
lossy compression format (JPEG, etc)

The MAJOR task now is segregating parts. A lot of printed kits are printed on paper a good bit larger than our “letter” size (or even
European A4 size), plus the 65% increase obviously means that the resultant files are much larger than you can print with a typical home
printer! So, now you’ve got to split things up to fit on whatever you are printing to. The easy way would be to just get into your
editor and select letter size groups of parts and then cut and paste those into new files. Or even just print sections of each file
letting the parts get cropped at the margins. I decided that while I was going to all this trouble anyway, I’d segregate them into
assembly order – a big job! But I always get tired of hunting for parts so I thought I’d do the work on the front end. I’m not really
convinced it’s worth all the trouble. I’m not yet finished with it yet, but I’ve got enough pages made to get a long way into the
construction.

You need to at least segregate parts into categories by laminating thickness – all the “1mm” parts together, “0.5mm” parts together,
regular card stock parts on the rest, etc. This kit has 6 thickness categories to worry about (more on that later).

Printing:

On the printer side, you also want to make sure the printer driver is not making “exposure” decisions for you, so you’ll want to select
something for output that is NOT Photo. I’m printing on a Canon i860, and it has custom settings to print as type: Graphic. I also make
the Brightness setting = Light to more closely match the original book’s darkness.

Materials:

Upsizing the parts themselves is not the end of the battle. You also need to make sure your stock and laminating thicknesses are
multiplied as well. Some basic facts on current Halinski books:

Card pages are 0.21 mm thick (same as 65 lb Wausau Bright White)
Paper pages are 0.10 mm thick (about the same as 22 lb inkjet paper)

I have a pad of Strathmore Bristol, which measures 0.32 mm thick, a 52% increase over the Halinski card stock. Close enough I think for
upsizing the card stock parts. This stock is pretty stiff, and does not fold nearly as cleanly and easily as our regular weight card, so
I think for many small parts that will be rolled and folded, I’ll still print a 65lb stock version. The thicker Bristol will be used for
skins and parts that are basically flat, and any other things where dimensions might make a difference in fit.

As for paper parts, anything that’s just a pattern I’ll print to regular paper, but for those parts that get rolled around wire, etc, I
will probably try to upscale it to something like 28 lb paper.

As for laminations, 0.5 mm and 1 mm come out to 0.825 and 1.65 mm respectively. I’ve got a stock of ~.80 mm cardboard from the back of
X-ray films (hat tip to Ted!), so I think I’ve got that one nailed pretty close.

That leaves the parts to be laminated to “Bristol” or double that (marked + and ++). I will probably just use the Bristol I’m making the
skin parts from for these, but some experimentation may be in order to see for sure.
I’ve got the feeling that this one is going to take a while! And I might need to buy a bigger cutting mat
Reply With Quote