Bob Broder wrote and asked a question about the merits of single vs
multi-coated filters.
Brian Huber replied:
Bob,
As I've always understood it:
1. Using no filter gives the best optical result.
2. 2nd choice (1st for many of us when we think of damage to lens and
consequences) would be to use a multicoated filter.
3. 3rd choice would be a non-multicoated filter. Quality of filter
(manufacturer) would also be important. Glass isn't just necessarily
glass. How flat it is, small defects, color consistency, mounting (is
it
plane?) can be contributing factors.
The reason to use a multicoated filter would be that internal
reflections
are kept to minimum. Each time a new optical surface is introduced, a
new
reflection (or two) is created. There have been volumes written about
this. I'm sure others on this list can give much more detail in better
form than I.
-------------------------------------------------------------------
Now I have to add my piece:
I'm in general agreement with Brian's reply except that I don't know why
he thinks a non-multicoated filter might have something to do with
damaging a lens (see point #2 above). Perhaps he meant to say something
else and it just didn't come across the way he meant it.
Having agreed that multicoating is better than single coating I have to
add that there seems to be (amongst some list members) an over emphasis
on optical quality. (Doris Fang being the major exception) Better is,
of course, better but sometimes the difference is mighty hard to detect
and might not be worth the price to some folks.
When it comes to multicoating the first thing we need to keep in mind is
that single coating is already about 97 0.000000e+00ffective in eliminating
reflections from a single surface. The first glass surface of the lens
is going to pass 970f the light and 3 0s going to reflect off. Off
the first element it's only light loss. Off the interior elements that
3 0s trying to bounce around and go other than where you want it to.
Multicoating gets you into the 98-99% range. Doesn't sound like much of
a difference until you consider a multi-element lens. Some of the
elements don't count because they're cemented and there is no air/glass
interface to cause a reflection. But, assume there are 4 surfaces we
have to contend with (the front of the lens and 3 air gaps between
elements). At 97 0.000000e+00fficiency the total light transmission will be
.97 x
.97 x .97 x .97 = 88.5%. If the multicoating is 98.5 0.000000e+00fficient on a
single surface then the total transmission will be .985 x .985 x .985 x
.985 = 94.1%. So we get 6.6% more of the light through the lens and we
reduce the pesky reflections.
Pretty good but you have to care a LOT to call it dramatic.
Now consider adding one more piece of glass and another air gap... the
filter! Whether the lens is single or multicoated we're only going to
add one other multiplier. If the lens is multicoated and the filter is
single coated we'll end up with .941 x .97 = 91.3% transmission. If the
filter is multi-coated we'll get .941 x .985 = 92.7%.
As you can see, the effect is VERY small. Far less than the difference,
say, between a single and multicoated lens... and I haven't heard of
anybody throwing away their single coated OLY lenses.
Brian's third set of points (flatness, color, small defects, mounting)
are also considerations. However, the manufacturing of a thin, small
diameter flat such as a filter would have to be positively atrocious to
cause any appreciable effect back at the film plane. Since we are not
refracting the light the glass almost doesn't matter. It just needs to
be clear and free of any blemish. Also, the filter would probably have
to be out of plane by several degrees before there would be noticeable
quality effects. With the consistency given by automated manufacture my
guess is that it would be difficult to find an unacceptable filter...
even from the lowest cost manufacturer.
ps: Treat my 97% and 98.5% as representative for the illustration of my
points. 25 years ago I knew better what the numbers were. These aren't
gospel but they're in the ballpark. This also has some import for the
discussion that was going on here awhile back about whose multicoating
technology was best, who owned what patents, etc. Twenty five years ago
multicoating was already at the 98 0.000000e+00vel or even better. That didn't
leave a whole lot of room for improvement. I don't know about Japan but
in the US patents are only good for 17 years. The point is that any
lens manufacturer today probably has access to coating technology that
is almost as good as it gets and without paying any royalties.
Maybe Gary has a cheapo filter and can prove me wrong on the next lens
test. OK. Flame suit on... fire away!
Chuck Norcutt
< 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 >
|