Shaprness reduction due to beamsplitter glass??

[ray_parkhurst]

In my previous attempts at "axial lighting" for coins, I used non-beamsplitter glass. It provided a reasonable split, but there was still some "residual distortion" that I was unhappy with. I figured that using high quality beamsplitter glass would eliminate the problem, so I invested in the 50/50 5x7" glass from Edmund. Unfortunately, the optical problems are still there, with the result that my beamsplitter-lighted images still show a much lower sharpness than my direct-lighted ones. Here are my examples images:

Direct-Lighted, two diffuse Jansjos full coin, f5.6:


Crop from the above:


Beamsplitter-Lighted, flood diffused, full coin, f5.6:


Crop from the above:


Sorry that the images are at a bit different magnification, and not fully white balanced.

The beamsplitter image looks like it is suffering from diffraction. Both images were taken using a 75mm Apo-Rodagon D M1:1 lens at f/5.6, so the lens is probably not at fault. Is there an optical property of the beamsplitter itself that might be causing this effect? Or maybe there is some aspect of my setup, such as not having precise 45-degree alignment (I am less than 1 degree off), causing the problem?

[rjlittlefield]

Is there an optical property of the beamsplitter itself that might be causing this effect?

Yes, at least two of them. First is that there is a systematic aberration (wavefront error) due to the fact that light going through different parts of the lens aperture went through the beamsplitter glass at different angles and thus was delayed by different amounts. Second is that there may be a random aberration introduced by slight variations in thickness of the glass.

The systematic problem is worse with thick glass, better with thin. See for example the images at http://www.photomacrography.net/forum/v ... 680#100680.

Or maybe there is some aspect of my setup, such as not having precise 45-degree alignment (I am less than 1 degree off), causing the problem?

I doubt that alignment is an issue here. I have seen illustrations in an old Kodak publication that certain diffraction effects can be strongly affected by angle. But what you have here doesn't look at all like those.

--Rik

[Pau]

To minimize this effect you must place the beamsplitter at the side of the lens that makes/uses the image forming light rays more parallel (ideally fully parallel like in infinite corrected microscopes). So for magnification under 1X it will be placed at the subjet side and for higher than 1X at the sensor side, assuming a simple lens, not a combo.

[ray_parkhurst]

Is there an optical property of the beamsplitter itself that might be causing this effect?

Yes, at least two of them...
...
The systematic problem is worse with thick glass, better with thin. See for example the images at http://www.photomacrography.net/forum/v ... 680#100680.
...
--Rik

Thanks Rik, my thoughts were in that direction since the previous beamsplitters I used were thinner and had less of this problem. The Edmund 50/50 I used for this test is quite thick!

To minimize this effect you must place the beamsplitter at the side of the lens that makes/uses the image forming light rays more parallel (ideally fully parallel like in infinite corrected microscopes). So for magnification under 1X it will be placed at the subjet side and for higher than 1X at the sensor side, assuming a simple lens, not a combo.

Thanks Pau, makes sense. In this case M=0.7 and I put the beamsplitter between subject and lens, so it is in best position.

[g4lab]

Their used to be pellicle beam splitters available that were extremely thin and would probably reduce some of those aberrations. I think they are moderately expensive but Edmund used to have them.

Canon once had a film SLR called the Pellix which had a fixed beam splitter for the reflex function. It was a bit lighter in weight and as quiet as a Leica.

[MaxRockbin]

I haven't tried this (!), but I read about it and it seems like a neat idea and possibly good for your application.

It's "near axial" lighting and much easier to describe with a picture:



Instead of a single 45 deg. angled piece of glass, you split the reflector in two and create a gap wide enough for your lens to see through.

Your reflectors no longer have to be transparent, so you no longer have to worry about aberrations introduced by the glass and you get more light (shooting through a partial reflector, not only does less light hit the subject, but only half of that light gets back to the lens as well). You can also use white cards to get more diffuse light on the subject. The downside is that it's not perfectly axial. You won't be able to light the inside of a narrow tube, but I think you could light a coin if you have a tolerable working distance.

I can't seem to find the exact reference right now, but it was either from Lester Lefkowitz "Manual of Close up Photography" or Alfred Blaker "Handbook for Scientific Photography."
(Both terrific old books you can get for a few dollars). I can't recall seeing an example in the forums.

[rjlittlefield]

Their used to be pellicle beam splitters available that were extremely thin and would probably reduce some of those aberrations. I think they are moderately expensive but Edmund used to have them.

Edmund still does have them, HERE.

But to call those things "fragile" would be a gross understatement. Their thickness is specified as 0.002 mm.

you split the reflector in two and create a gap wide enough for your lens to see through.

Or cut a hole in it, if you're using white card.

I don't recall seeing that in Lefkowitz. Must be in Blaker, but my copy of that hasn't come yet (and I ordered it a whole 3 minutes ago ).

--Rik

[Choronzon]

Actually, those pellicle mirrors are a lot stronger than glass of an approaching thickness that would eliminate double refraction

[rjlittlefield]

you split the reflector in two and create a gap wide enough for your lens to see through.

Or cut a hole in it, if you're using white card.

I don't recall seeing that in Lefkowitz. Must be in Blaker, but my copy of that hasn't come yet (and I ordered it a whole 3 minutes ago ).

Yes, it's in Blaker, in the section on "Axial and Near-Axial Lighting", pages 49-54. He has both the split mirror (Fig 4-12) and the card with a hole in it (Figure 4-13D, "matte surface mirror with rectangular hole").

--Rik