In your message dated: 23 Aug 2003 00:59:15 -0000 you write:
>
>
>Lee,
>
>The implication is that color is useless for telling SC and MC apart. I guess
No. The implication is that MC lenses have a lot more colors than SC
lenses, especially early SC lenses.
> the only way to really tell is to measure the strength of the reflections, ra
>ther than just eyeballing it, although at least for planar surfaces one can vi
>sually compare reflection strengths with that from an uncoated filter.
Yes, reflection strength is probably the only way to be 100 ertain.
But seldom are SC lenses "trickily coated" to show lots of different
colors from SC. It's just not usually done that way. Why? Because if
you are gonna coat different lenses to different thicknesses, it becomes
a lot more expensive process for not very much gain. You might as well
bite the bullet and multicoat.
>
>>My guess is that your 50/1.8 is multicoated, and to a much higher
>>standard than the 55/1.2.
>
>Based on selling price alone, I would have expected the contrary. The 55/1.2
You are forgetting economies of scale (are you in marketing, btw?).
Thousands, probably tens of thousands more 50/1.8's were made than
55/1.2's. Also, the 50/1.8 was made over a lot more years, which
represented more changes in technologies.
Also, once you develop a coating plant that can do multicoating, it is
really not a lot more expensive to multicoat everything you make, as
opposed to just a few lenses. I realize that different elements got
different coating prescriptions, but this is a minor tweak to the above
reasoning, from a coating plant perspective.
>And the colors of the reflections from the 1.8 were all over the place, includ
>ing straw and white.
Which means not all surfaces were multicoated. Or coated at all. Which
is not unusual. A lens doesn't have to have all surfaces of each
element multicoated to qualify as a "multicoated" lens.
>
>
>>As for process control in coating plants, it would be basically useless
>>to build a coating plant that performs as you imply. If you couldn't
>>control the coating thickness to the wavelength required, it would be
>>hard to know what results you would get. The coatings that resulted
>>would likely (very likely) end up doing more harm than good.
>
>I believe that too, but what I see is perfect consistency in the 55/1.2 and co
>mplete inconsistency in the 50/1.8, and from economics alone one would expect
>more care and tighter tolerances in the more expensive lens.
Yes, but the flaw in this argument is that multicoating was a practice
that was used at lot when the 55/1.2 was designed. It wasn't. I'm
sure the designers knew they would be replacing the lens with a 50/1.2,
and with the few made (compared to other lenses) doing the redesign
wasn't worth it.
Bill wrote:
>One must factor:
> - time period whence manufactured.
> - where manufactured and the vendor for MC
> - evolutionary period for the 1.8 versus the 1.2 which is more
>than an
>order of magnitude when adjusted to include the number of units
>produced for each design X options for low cost bid for large
>production runs versus small / custom lots.
>FACT: "Inconsistency" in coating does NOT equate to inconsistency in
>optical performance.
I agree with everything Bill says, except the last item. What do you
base this "fact" on Bill? Are you actually suggesting I could take a
set of elements for a lens, throw them in the coating chamber, put
random thickness coatings on them, and end up with a lens that performed
well optically? Coatings are governed by physics. I don't claim to
know everything about coatings... although I have designed them and
taught my students how to design basic coatings. It is a very complicated
area. I still have volumes to learn. There is a reason why only the most
expensive optical design programs can deal with multicoatings. And
there is a reason why when you change your coating design, you often have
to tweak the optical presciption of the optical system. I don't want to
start a war here, but one of the reasons I quit reading this list in the
first place was statements that people swore to that were just plain wrong.
It appears nothing has changed. I'm not even gonna comment on the
recent "get a Mac based on Unix* thread. That type of thread has no
business on this type of list, period. *sigh*
--Lee
< This message was delivered via the Olympus Mailing List >
< For questions, mailto:owner-olympus@xxxxxxxxxxxxxxx >
< Web Page: http://Zuiko.sls.bc.ca/swright/olympuslist.html >
|