How to illuminate at WD=0.1mm ?

[Adalbert]

Hello everybody,
Does anybody have any idea how to illuminate the subject if the following lens is used?


I have already tried with 3 COB-LEDs:

BTW, at the moment I’m constructing an adapter for the toslink cables to the flash.
BR, Adi

[enricosavazzi]

Hello everybody,
Does anybody have any idea how to illuminate the subject if the following lens is used?


I have already tried with 3 COB-LEDs:

BTW, at the moment I’m constructing an adapter for the toslink cables to the flash.
BR, Adi

I would strongly consider a form of (modified) axial illumination through the objective, with a beam splitter at the rear of the objective. Since this is an infinity corrected objective, it should be possible to mount the beam splitter between objective and tube lens. This may be impossible if you are using the objective on a microscope (which is otherwise the logical choice at this magnification) and the microscope does not support axial illumination, but with a typical photomacrography setup or a microscope with axial illuminator it should work out.

What I am suggesting is not the typical axial illuminator, which most likely is going to give an unacceptably "flat" illumination of 3D specimens. Instead, you can get an oblique directional illumination by masking out the collimated light beam from the illuminator before it strikes the beam splitter, except for a portion of the beam that passes through a peripheral region of the objective. A 0.95 NA objective collects a strongly divergent cone of light coming from the subject, so using only a peripheral region of this cone for illuminating the subject should allow a sufficiently high angle from the normal to give you a good directional shading of 3D subjects.

[Pau]

Adi, what's in your picture? maybe a hair? The center seems nicely rendered.

Enrico, what you propose is some kind of reflected COL, am I right? It seems a good idea to test.

[enricosavazzi]

Enrico, what you propose is some kind of reflected COL, am I right? It seems a good idea to test.

Not really COL, but what one would obtain by masking off most of the hollow cone of light and using only a narrow sector of the cone.

There may be a show-stopper though, at least for highly reflecting subjects. The type of illumination I propose strongly reduces the NA of the objective with respect to the illumination light beam (because it uses only a portion of the front element of the objective). If the subject acts as a good diffuser/reflector, the light returned by any given point of the subject hits the whole surface of the front element of the objective, and in practice utilizes the whole NA of the objective. If, on the other hand, the subject acts mainly as a specular reflector (even though not a flat one), the light returned by any given point of the subject to the objective still underutilizes the NA of the objective. The consequence is that the objective behaves like one with a significantly lower NA, with all the negative consequences on diffraction and resolution that this entails.

[Adalbert]

Hello Pau, hello Enrico,

Yes, it is a human hair, which I have already shown in the following thread:
http://www.photomacrography.net/forum/v ... 6&start=15

Enrico, do you mean something like the method 3 in:
http://www.photomacrography.net/forum/v ... ht&start=0

BR, Adi

[enricosavazzi]

Enrico, do you mean something like the method 3 in:
http://www.photomacrography.net/forum/v ... ht&start=0

More like method 2, but with a narrower beam of light that passes through one peripheral area of the objective, rather than through the whole objective. Because of the high NA of the lens, this beam would strike the subject obliquely, not along the lens axis, and therefore it would provide a directional shading of 3D features.

Hence it qualifies as a variant of COL (Circular Oblique Lighting), except it would not be circular (more properly ring-shaped), but would come from only a sector of the ring.

Another way to visualize it is: it is the same principle as transmitted darkfield illumination with a sector stop added at the top of the condenser, but incident rather than transmitted.

[Adalbert]

Hello Enrico,
I could combine the method 2 with the toslink-cables.
So, the light-source would be the flash or LED and the light would be transferred by the toslink to the mirror.

BR, Adi

[enricosavazzi]

Hello Enrico,
I could combine the method 2 with the toslink-cables.
So, the light-source would be the flash or LED and the light would be transferred by the toslink to the mirror.
BR, Adi

I don't know how collimated light that enter the optical fibers remains collimated after exiting the fibers. You might need collimating optics between the end of the fibers and the beam splitter.

[Pau]

It exits not collimated. IIRC the typical entrance (and exit) angle is 30deg. so the light cone is 60deg. but may vary with models

[enricosavazzi]

It exits not collimated. IIRC the typical entrance (and exit) angle is 30deg. so the light cone is 60deg. but may vary with models

That seems too wide a cone to me. Ideally, a small planoconvex or biconvex lens with a focal length of not more than 15-20 mm, placed at the same distance from the exit of the fiber, should be enough as a collimator. Then a baffle with a 1-2 mm hole between lens and beam splitter should suffice to isolate a narrow beam of roughly parallel light to send to the beam splitter. Make the baffle mobile sideways, so that you can adjust its position until it gives the desired result.