Newbie question about macro with DSLR

tony cooper wrote:


Microscopes can be very expensive, but it is unlikely
that you need a fancy one. It should be fairly easy to
find an older used model that will do fine. They often
are available on eBay at reasonable prices, as are the
parts and pieces such as objectives and "relay lenses"
for photography. Keep in mind that objective lenses are
matched to the eye pieces (and to the relay lenses used
for projecting into a camera), and hence should not
necessarily be mixed and matched between manufacturers.

I can't figure out what you're saying here, and I'm pretty familiar
with microscopes. I have no idea what "relay lenses" are.


They are what the words say. I own a late 1990's inverted Zeiss Axiovert
microscope, a top of the line gizmo. The objective sends the light,
COLLIMATED, through an optional beamsplitter used for epi-fluorescence.
This beamsplitter lets the fluorescence through, and reflects the light that
comes IN to excite the fluorescence. It also contains transmission
filters for the incoming and outgoing light. It is optional, and
is not used for brightfield illumination. This is not a metallographic
microscope and does not have darkfield.

The light then continues downward through a "tube lens". This is a lens
of 163 mm focal length that focus the collimated light. The light then
goes out the bottom of the scope onto a TV camera. There is a mirror that
slides into this path and reflects the light up toward the eyepiece.
But it focuses too soon, so there is a field lens near the focus,
and, farther on, a set of relay lenses that image the first virtual
image, at the field lens, up into the eyepiece assembly. That's what
a relay lens is used for.

Because Zeiss now uses "tube lenses", an objective from my scope would
not, by itself, work properly as a camera lens. You would need the specific
Zeiss "tube lens".

Different companies (i.e. at least the big three, Nikon, Zeiss, and
Olympus) make different tube lenses. They are part of the aberration
correction system and using the wrong one will result in lateral
chromatic aberration. Olympus used to sell ... maybe still does,
and you certainly can get used, objectives that do NOT need a tube lens,
instead doing all the focusing themselves. Some of these, unfortunately,
don't correct the later chromatic, instead leaving that to the eyepiece.
However, many of them (or rather did). Any one that was intended
for photography without a tube lens did.

Doug McDonald

I wrote:
I've looked at the Olympus site, and am somewhat lost. If it's not too
much trouble, can you give me any more specific idea as to what I'm looking
for?

Floyd L. Davidson wrote:
[lots of good advice and references, including...]
The Olympus webpage has much good info, and these are a
couple to start with:

http://www.olympusmicro.com/primer/faq.html

This one is long with lots of general information:

http://www.olympusmicro.com/primer/microscopy.pdf
[etc.]

Thanks! That was exactly the sort of information I needed!

Eric

Doug McDonald wrote:
What are you going to photograph?

PMOS integrated circuits fabricated between 1975 and 1978.
Minimum feature size (as drawn) is at least 3 microns.

Remember that lenses are available at 20x magnification that will
resolve .5 micron ... that is, 1000 line pairs per mm.

Really? Can you point me to any specific models?

Thanks!
Eric

Archibald wrote:
You should also be able to achieve this by reversing an ordinary Canon
lens.
[followed by a detailed explanation]

Thanks for writing that up!

Eric

Floyd L. Davidson wrote:
You need a microscope.

I suppose you're probably right. I was hoping some other, less
expensive possibilities, like a Raynox MSN-505 or CM-3500 conversion
lens used with a telephoto lens might do the trick. I can probably
find other uses for the conversion lenses anyhow, so I don't mind
buying them even if they don't work out for my current project.

Do you want _each_ image to be even higher magnification
than stated above, and for the resulting final product
to be as described above???? Or do you mean each
exposure will capture 0.5 to 1.0 mm of a larger object?

The overall size of the subject is under 10mm on a side. I wanted
to get shots of 0.5 to 1.0 mm square portions of that, then stitch
them together.

If you mean that each exposure will capture 1mm of an
object, then a typical dissecting microscope would be
appropriate (and would be very convenient to mate with
your X-Y state). Another possibility is a toolmakers
metallurgical microscope.

OK, I'll look into those.

The feature you'll want to look for is the ability to
mount it in a way that you can use your X-Y stage (which
is not true of all metallurgical scopes).

Some of the relatively inexpensive microscopes I've looked
at have a single-boom or double-boom mount that should do
that. I'm not sure which would be better for stability
(vibration).

Another feature, which is almost universal though there is a
chance you don't need it, is episcopic (or "epi" for
short) illumination. That uses a beam splitter to
illuminate the object through the same optical path that
you view it. (That is probably _very_ useful to you.)

Definitely. I hadn't even heard of it. The microscopes
I've looked at have ring illuminators, which obviously
wouldn't be useful if the objective has to be very close
to the subject.

Thanks for all the advice!
Eric

Eric Smith wrote:
Doug McDonald wrote:
What are you going to photograph?

PMOS integrated circuits fabricated between 1975 and 1978.
Minimum feature size (as drawn) is at least 3 microns.

Remember that lenses are available at 20x magnification that will
resolve .5 micron ... that is, 1000 line pairs per mm.

Really? Can you point me to any specific models?

Thanks!
Eric

Unfortunately, not usefully. I can point to plenty of ones
that need a tube lens, but not ones that don't. They do exist.

Remember this: resolving power is determined by "numerical aperture",
N. A.

f/number = 1 / (2 * N. A.) (for lenses working in air)

In other words an N. A. 1.0 lens is f/0.5
0.5 lens is f/1.0
0.3 lens is f/1.6
etc.

resolving power, defined as the half-width of the blur function at the
subject, is roughly f/number, in microns.

In other words, the blur of an N. A. 0.5 objective is about 1 micron.
This is fairly rough, formally it is actually 0.83 microns.

You don't need a big-name objective to get the job done.

Google on "microscope objectives". You need a "plan" one (i.e. flat field) for "no tube lens"
or NOT "infinite focus" and also if it matters for that objective, NO COVER GLASS.

Also, you need to worry about working distance if you intend to send light in from the side.

I found many objectives that will work under $150 ... but I didn't check to see
if they will work well with no cover glass.

Doug McDonald

lid wrote:
Eric Smith wrote:
Doug McDonald wrote:
What are you going to photograph?
PMOS integrated circuits fabricated between 1975 and
1978.
Minimum feature size (as drawn) is at least 3 microns.

Remember that lenses are available at 20x magnification that will
resolve .5 micron ... that is, 1000 line pairs per mm.
Really? Can you point me to any specific models?
Thanks!
Eric

Unfortunately, not usefully. I can point to plenty of ones
that need a tube lens, but not ones that don't. They do exist.

Remember this: resolving power is determined by "numerical aperture",
N. A.

f/number = 1 / (2 * N. A.) (for lenses working in air)

In other words an N. A. 1.0 lens is f/0.5
0.5 lens is f/1.0
0.3 lens is f/1.6
etc.

resolving power, defined as the half-width of the blur function at the
subject, is roughly f/number, in microns.

In other words, the blur of an N. A. 0.5 objective is about 1 micron.
This is fairly rough, formally it is actually 0.83 microns.

You don't need a big-name objective to get the job done.

Google on "microscope objectives". You need a "plan" one (i.e. flat field) for "no tube lens"
or NOT "infinite focus" and also if it matters for that objective, NO COVER GLASS.

Also, you need to worry about working distance if you intend to send light in from the side.

Objectives for metallurgical work are the ones which fit
the description. They have longer working distances and
are intended for use with no cover glass. Typically
there will be an "M", "N", or "E" in the model number.
Zeiss might label something as "EpiPlan".

A typical description would look like this:

40x/0.70 160/-

The 40x indicates the magnification when used as a
microscope, the 0.70 is the N.A. value as discussed
above (indicating the aperture), the 160 is the tube
length it was designed for (avoid any listed as "inf"),
and the "-" indicates that is it designed to be used
without a slip cover (otherwise a number, such as ".17"
would be listed, and indicates the thickness of the slip
cover).

The problem that I'd wonder about is that typically the
objective and the ocular or projection lenses are
matched pairs with aberrrations compensated partially in
each. Apparently Nikon may not do that, and therefore
might be the best choice of manufacturer, but I have
never been able to positively confirm that. I've never
seen any numbers for Unitron either, though the
nomenclature they use for oculars would indicate
compensation too.

Typically available objectives that might meet Eric's
needs a

10X:
Leitz 10x/0.25 170/-
Nikon E Plan 10x/0.25 160/-
Olympus Neo 10x/0.25 200/
Unitron MF (Plan Acro) 10x/0.25
Unitron MPL 10x/0.25
Zeiss NPL Fluotar 10x/0.30 160/-

20X:
Olympus NeoPlan 20x/0.40
Unitron MF (Plan Acro) 20x/0.40
Unitron MPL 20x/0.40
Unitron M 20x/0.45

40X:
Unitron Epi MF (Plan Acro) 40x/0.65 170
Unitron MPl 40x/0.65 170/-
Unitron M Plan 40x/0.65 170/-
Unitron M Acro 40x/0.65 170/-
Zeiss EpiPlan-HD 40x/0.85 160/0

Olympus and Unitron objectives are very common and
usually relatively inexpensive. Nikon and Zeiss seem to
run hot and cold on prices. Regardless, it should be
possible to come up with a representative example for
less than $100 that is at least useful for testing. If
it is worth the difference, there are certainly a number
of objectives that cost several hundreds of dollars.

--
Floyd L. Davidson http://www.apaflo.com/floyd_davidson
Ukpeagvik (Barrow, Alaska)

Joseph Meehan wrote:
"Jürgen Exner" wrote in message
...
...

Yeah, for whatever reason Nikon calls "Micro" what everyone else calls
"Macro".


Yea. Frankly everyone else is wrong, :-0 Well Nikon is wrong in
their use from time to time as well. I don't think it is going to get
straightened out, but it is good to keep in mind that there are two
terms and that we should all expect them to be used properly and
improperly so we don't assume that they will be used in the manner we
expect.

Nikon calls them micro because they already used macro for an obscure
old line of extreme closeup lenses:
http://www.naturfotograf.com/lens_spec.html