Test result in last post - Beam splitter cube vs plate

Have questions about the equipment used for macro- or micro- photography? Post those questions in this forum.

Moderators: Pau, rjlittlefield, ChrisR, Chris S.

zzffnn
Posts: 1839
Joined: Thu May 22, 2014 1:25 pm
Location: Boston, Massachusetts, USA
Contact:

Test result in last post - Beam splitter cube vs plate

Post by zzffnn »

Thank you for checking this thread.

I am looking at this Edmund Optics 50R/50T Plate Beamsplitter:
http://www.edmundoptics.com/optics/beam ... ers/43359/

Compared to a plain 1mm thick glass plate placed at 45 degrees, how much better it would be? Would that save a lot more light?

I would like to put the beam splitter at 45 degrees under my microscope condenser, to direct both my microscope light and (side-mounted) camera flash up the condenser.

I tried a 0.8mm thick acrylic plate placed at 45 degrees and it worked (please see photo #3 of my thread http://www.photomacrography.net/forum/v ... ht=#184790 for a sample photo that the acrylic plate produced - you can see some halos, but it is most likely objective/darkfield's fault). The problem was that, I had to use flash at 1/4 or even 1/2 (though I might have used a flash focal length that was too wide).

Would a dedicated beam splitter plate save a lot light output, compared to 0.8mm acrylic plate? How about compared to 1mm plain glass plate? I understand that dedicated beam splitter should produce less aberration, when palced at exactly 45 degrees.

Edit: I have also tried a thick arylic right angle prism. But light loss from transmitted light was too much for it to be useful. So prism cubes may not work.

Also, some eBay beam splitter plate has nm numbers on their description, such as 830 nm. What does that wavelength value mean?
http://www.ebay.com/itm/Melles-Griot-Pl ... SwT~9WlNHW

I am also considering the following:
http://www.ebay.com/itm/MICROSCOPE-MOUN ... 3641.l6368

Not sure if it is worth it though, compared to a plain glass plate. I would like more light reflection efficiency, so that my flash can run faster for freezing protists movements. At 1/4 flash output, flash may be too slow for freezing motion. I am only looking for decent image quality, and not willing to pay a lot more to get that final 5%.

Thank you very much for your kind advice.
Last edited by zzffnn on Thu Nov 03, 2016 10:35 am, edited 2 times in total.
Selling my Canon FD 200mm F/2.8 lens

Pau
Site Admin
Posts: 5317
Joined: Wed Jan 20, 2010 8:57 am
Location: Valencia, Spain

Post by Pau »

So prism cubes may not work
I use a prism cube and it works very nicely, is easy to mount and its surfaces are much less delicate than a plate metal coated half mirror. For illumination is excellent, to image through it the plate BS is much more convenient. Both the Edmund plate and the prisms provide 50/50.

Take a look at: http://www.photomacrography.net/forum/v ... hp?t=26185

Mine is from SurplusSed and is inexpensive: http://www.surplusshed.com/pages/item/l2047d.html
They have it in 10, 15, 20, 30 and 40mm versions, mine is 30mm, I formerly used it just sitting over the illuminator as you intend.

Edmund also sell cube BSs at much higher prices but they offer them also in different T/R ratios
Pau

zzffnn
Posts: 1839
Joined: Thu May 22, 2014 1:25 pm
Location: Boston, Massachusetts, USA
Contact:

Post by zzffnn »

Thanks Pau. I have read those previous threads / posts or yours.

I wonder why my acrylic right angle prism block that much transmitted light. I was using a big 40w (very bright) LED. But once my prism was placed over field lens, light from scope was no longer bright (most was blocked by prism, though a small amount did pass through - it was quite dim though). A acrylic plate of 0.8 mm did let enough transmitted light pass through, in comparison, though reflected light seemed less.

Ideally, I don't mind giving more priority to reflect (flash) light due to my flash arrangement, but my acrylic prism blocked way too much for visual and chasing protists.

Acrylic or polycarbonate should have RI of 1.49-1.6, pretty close to glass. So my acrylic prism should not be much worse than yours, I assume? And I am guessing transmitted light loss would be less, if light only needs to pass 1mm of beam splitter plate, verses a 20 mm of glass prism?

Do you think a plain 1mm thick glass plate would work as well (have you compared them)?

The following microscope beam splitter looks like a (45 degrees) mounted plate, though it is not cheap (is it worth it, compared to plain [non beam splitter] glass plate?): http://m.ebay.com/itm/252263553456?_trk ... mwBanner=1
Selling my Canon FD 200mm F/2.8 lens

Pau
Site Admin
Posts: 5317
Joined: Wed Jan 20, 2010 8:57 am
Location: Valencia, Spain

Post by Pau »

So my acrylic prism should not be much worse than yours, I assume? And I am guessing transmitted light loss would be less, if light only needs to pass 1mm of beam splitter plate, verses a 20 mm of glass prism?
Is yours a cube prism or a simple prism?. If the light is so dimmed I'm sure it's working very differently than my cube prism. With a 10w LED I have enough light for DIC.
Of course a plate BS is fully adequate, the ebay one seems nicely mounted for your purpose although the T/R ratio is unknown.
Pau

zzffnn
Posts: 1839
Joined: Thu May 22, 2014 1:25 pm
Location: Boston, Massachusetts, USA
Contact:

Post by zzffnn »

Thank you, Pau.

My bad. That maybe where the problem is. This is my simple prism: http://www.amazon.com/Large-Right-Angle ... B00WL3OGEU

Maybe mine was made to reflect a lot more than transmit? It does not say T/R ratio. I would guess it is NOT anti-reflection coated. I bought it a long time ago for a different purpose.

By the way, I hope you understand - I was only trying to learn more by asking why (not arguing with you).

I read previously that beamspliter has to be mounted at exactly 45 degrees, otherwise there would be aberration, correct? I did make a DIY mount as close to 45 degrees as possible, but that may not be good enough? And you are right, a plate beam splitter may be too fragile.
Selling my Canon FD 200mm F/2.8 lens

zzffnn
Posts: 1839
Joined: Thu May 22, 2014 1:25 pm
Location: Boston, Massachusetts, USA
Contact:

Post by zzffnn »

Any other kind comments? Thank you!
Selling my Canon FD 200mm F/2.8 lens

ray_parkhurst
Posts: 2819
Joined: Sat Nov 20, 2010 10:40 am
Location: Santa Clara, CA, USA
Contact:

Post by ray_parkhurst »

I have an Edmund 50/50 splitter and built a nice frame for it to use in shooting coins. It works reasonably well for lighting, but I find it introduces an unacceptable amount of distortion/unsharpness to the image. I rarely if ever use it because of this.

I'd also like to comment that if I purchased another splitter, I'd lean toward 60/40 or even more reflective (30/70? not sure what the next steps are) and point the reflective side down. You get a bit less overall light at the sensor, but having more light hit the subject and less transmit to the sensor has 2 advantages:

1) Less light passes through from the source, and what does pass through is reflected less.

2) Ambient lighting hitting the sensor is reduced.

The biggest problem (after the unsharpness issue) I've had with beamsplitter axial lighting is maintaining contrast due to stray light.

Ray

zzffnn
Posts: 1839
Joined: Thu May 22, 2014 1:25 pm
Location: Boston, Massachusetts, USA
Contact:

Post by zzffnn »

Thank you very much, Ray.

You reminded me of my own stray light issue: I think I set my flash light too wide at 50mm, so flash light beam did not go out as ideal parallel beams, thus I had light waste and more stray light

In your case, you were not using beam splitter under a microscope condenser, correct?
Selling my Canon FD 200mm F/2.8 lens

ray_parkhurst
Posts: 2819
Joined: Sat Nov 20, 2010 10:40 am
Location: Santa Clara, CA, USA
Contact:

Post by ray_parkhurst »

No, the beamsplitter was between subject and lens. Worst place to be!

zzffnn
Posts: 1839
Joined: Thu May 22, 2014 1:25 pm
Location: Boston, Massachusetts, USA
Contact:

Post by zzffnn »

Ray,

I saw some filter or mirror for "imaging through" at Edmund, but cannot remember where. Have you talked to an Edmund tech support, they offer online chat and phone consultation. Just in case you did not notice........

I often get lost in their web site and have to use their selection criteria (on page's upper left hand) a lot to find stuff (I missed lots of items, if I just go in a category and pick myself).
Selling my Canon FD 200mm F/2.8 lens

ray_parkhurst
Posts: 2819
Joined: Sat Nov 20, 2010 10:40 am
Location: Santa Clara, CA, USA
Contact:

Post by ray_parkhurst »

There are pellicle beam splitters available, with very thin membranes rather than glass, which should reduce the problem, but I've never gone far enough down the path to invest in one. I also worry about their durability and susceptibility to degradation in dusty environments.

I do have a couple of beamsplitters with very thin glass mirrors, one by B&L, and one by Nikon. The B&L was designed for their StereoZoom line of microscopes, and attaches to the bottom of the microscope pod. It does not introduce much distortion, but the illumination it produces is nearly pure axial, which brings up a interesting point...in coin photography, pure axial is not particularly useful. All of the "axial" systems folks use for coins are actually used to create a sort of hybrid lighting. Pure axial reflects strongly off the flat surfaces of coins (or any object) but is scattered by any angled/non-flat surfaces. This creates a strong shadowing effect around the devices or features of the coin, and thus a high contrast situation.

I made a diagram a while back to explain how an "axial" system can produce non-axial light, see below. By having the light be much larger than the subject, more angles of light can be created than just the 90-deg of pure axial. This allows a more diffuse illumination to be created, minimizing shadows and excess contrast, while still producing a pure axial component.

Image

zzffnn
Posts: 1839
Joined: Thu May 22, 2014 1:25 pm
Location: Boston, Massachusetts, USA
Contact:

Post by zzffnn »

Thank you very much, Ray. I found your post very educational and easy to understand.

In my case (beam splitter under microscope condenser to converge flash and scope light), I actually want pure axial light. I am guessing any non axial lght will not get utilized well by my condenser.

And durability aside, I think a beam splitter plate will serve me better than a cube prism, correct? Assuming BS plate and prism both have highly similar optical specifications.
Selling my Canon FD 200mm F/2.8 lens

Pau
Site Admin
Posts: 5317
Joined: Wed Jan 20, 2010 8:57 am
Location: Valencia, Spain

Post by Pau »

And durability aside, I think a beam splitter plate will serve me better than a cube prism, correct?
Both will do exactly the same for your application. Because you actually prefer the plate BS, just go for it.
Pau

zzffnn
Posts: 1839
Joined: Thu May 22, 2014 1:25 pm
Location: Boston, Massachusetts, USA
Contact:

Post by zzffnn »

Thank you, Pau.

I should have mentioned that I may have future applications similar to Ray's, though not as demanding. In that regard, I think since I am buying one anyway (and prices are similar), I may be better off buying a multipurpose plate BS. I have made a 45 degree platform for it.
Selling my Canon FD 200mm F/2.8 lens

zzffnn
Posts: 1839
Joined: Thu May 22, 2014 1:25 pm
Location: Boston, Massachusetts, USA
Contact:

Post by zzffnn »

Thanks to Pau's kind help, now I have a beam splitter plate and a cube prism to compare.

http://www.ebay.com/itm/50R-50T-Plate-B ... 7675.l2557

http://www.surplusshed.com/pages/item/l2048d.html

Both are coated. Surface quality rating of the cube is slightly better than the plate (though neither is of very high quality). The cube is 40 mm by 40mm. The plate is about 1.6mm thick and carefully installed on degree 45 degree DIY Lego platform.

My test reaults showed that the plate passed on more light visibly, in the direction of transmission.

In terms of reflection efficiency, the plate also does slightly better:
Comparing diatom dot brightness in 97x NA 1.1 darkfield, plate did best at 1/4 +0.3 or +0.7 flash power. While the cube did best at 1/2 flash or slightly higher.

I am keeping both. The cube is more durable and a lot quicker and easier to use. The plate is fragile and needs a precise (and not easy) mount. When I have more than enough light (for exmaple, when using high magnification oblique), I would use the cube. When I am doing high magnification darkfield (and close to running out of light), I would use the plate.

97.5x NA 1.1 darkfield test was done using a DIY 40w LED running at near full power, when beam splitters are used.

Post Reply Previous topicNext topic