Rikenon lens matching

[Greg Erker <erker@xxxxxxxxxxxxxx>]

  As well as attaching my light panel to my viewer,
I also managed to test my 5 Rikenon lenses on the
weekend. For those that don't recall, I'm looking for
the best match (in Focal Length) for my Stereo Ricohmatic
225 project.

  I numbered my lenses from 1 to 5, in order of desirability.
Numbers 1 and 2 are perfect and come from cameras that have
the f3.2 viewer lens. Number 3 is perfect but comes from
a camera with f3.5 viewer lens. Number 4 has some rear element
separation. Number 5 has a very badly scratched up front
element coating.

  The results are that the Erker lens magic appears to
be still holding (ask Sam and Eric what this means).
Don Lopp tells me it took him 7 or 8 diacords to get
a good enough match (though he didn't elaborate on
if he was matching more than just focal length).

  My test slide had nine cross hair points. The center,
the four corners and midway along the four edges. The
square they define is 50x50mm (to match the normal mounting
size of the slides).

---
Let's call Lens #1 the reference. So the relative FL's are:
#2: 0.53% bigger image based on diags, 0.34% based on H&V.
#3: 0.36% bigger image based on diags, 0.23% based on H&V.
#4: 1.13% bigger image based on diags, 0.92% based on H&V.
#5: 0.24% bigger image based on diags, 0.58% based on Vert only.
                                 (didn't measure H on this one.)

  So, not only do I have two matched to better than 1%, I have
three which means I can use my third 225 for taking hyper
stereos (by loading film into only half of the stereo 225).

  The best part is that they are my three best condition
lenses. Man, I'm happy!

  And I just won a Diacord auction on the weekend (bought
in case I needed more to get a match). Anyone need a camera?
:)

---

  The serial numbers of the lenses are 35915, 29432, 21316,
14469, and 26372 (in order). So only the very oldest lens
#4 is more than 1% different than the reference lens.

  So maybe Ricoh had better tolerances than the 5% we've
heard about.

---

  Before installing them I plan to look through them with
my Cabin Light Panel to make sure the colour casts are the
same.

---

  I'll end with a question (probably for John B.) If the
lenses had no distortion then the diagonal measurement should
be 1.414 times the horizontal (or vertical) assuming a square.

  They aren't. My spreadsheet shows the diagonals to be too
long by between 0.56% and 0.78%.

  Can anyone tell me how to figure out the barrel/pincushion
distortion amounts from my data?

Thanks - Greg (gloat, gloat :)