Unbranded Camera Adapter on Olympus BH2

siliconGary · Post by **siliconGary** » Tue Dec 27, 2016 3:03 pm

Having recently bought one I thought I would share my findings, as prior to purchasing I couldn't find much concrete information on them and saw several mixed (Mostly negative) comments about these adapters.

The background is I recently resuscitated an Olympus BH2 microscope with an LED retrofit. I wanted to add a camera, and got a Canon EOS1300D (Rebel T6) in a black Friday sale. I knew after reading the excellent resources on Charles Krebs and Alan Woods sites that the right way to do this was to buy a 1.67X (or 2.5X) NFK projection lens and a 150mm adapter. However the 1.67X appear only marginally more common than unicorns, and even for 2.5X in 6 weeks have only seen one on Ebay that would ship to Canada, and previous ones sold recently for ~$150-200 which together with the adapter tube and EF mount would mean ~$500 (CAD) to fit the camera properly outside of my budget presently.

.....so with hesitation I bought a no-name camera adapter.
This is the adapter I purchased
http://www.microscopenet.com/microscope ... p-223.html
Its a 2x lens adapter (Likely identical to OMAX adapter but I am not certain). At $108 it was within my budget for this.

First the unexpected. It fits inside the BH2 trinocular adapter and is perfectly parfocal with the eyepieces. This was a pleasant surprise as I was expecting to have to gerryrig the trinoc head.

Overall initial impressions are good, with the Canon T6 it gives a image that is a good representation of what I see in the eyepiece. FOV in the camera is smaller than the eyepieces but not frustratingly small.

Based on comments I had read from Charles K and others I was expecting to see a lot of chromatic aberration as the NFK ocular is correcting where these adapters are not. For most samples it is not obvious, but I found a micrometer slide on white background showed it up nicely.

20X objective 0.01mm line/space

Zoomed area at right hand edge of the image showing CA

For me I think it is worth the price, it gets me a photographic capability at a price I can manage and I don't think the CA will affect me too much - time will tell.

Lastly I will show a couple of regular images taken. I do this with some trepidation as I am in awe at the quality of images that are posted to this forum.

This is an image of a Gallium Arsenide integrated circuit with gold metal routing
20X objective

And this is the same image in dark field

ChrisR · Post by **ChrisR** » Tue Dec 27, 2016 3:20 pm

The first (completlely monochrome) image - did you correct the CA in post?

I'm confused about the size too - is it a crop?

siliconGary · Post by **siliconGary** » Tue Dec 27, 2016 3:34 pm

Yes the second image is a crop from the very right edge of the first. Without zooming in at the edge of the image you (I) cannot see the CA.

I don't have a stronger objective than the 20x currently, so don't know if it will be worse at higher magnification.

Pau · Post by **Pau** » Tue Dec 27, 2016 4:08 pm

Nice!
ND maybe means new design, I've seen pretty bad results with this kind of adapters, yours working so well is good news.

Because chromatic aberration due to the need of compensation of the objectives is symmetrical you can correct it pretty easily shooting RAW and correcting it during RAW conversion. It can be done with generic software like Adobe CR or with the free Canon DPP, in the second case you need to use a chipped adapter that tells the camera that a (fake) Canon EF lens is mounted, if not lens corrections are disabled.
A bit of sharpening and contrast adjustment will also improve the images

Aussie Phil · Post by **Aussie Phil** » Tue Dec 27, 2016 5:06 pm

Nice find, thanks for posting this. I'm looking for an adaptor for Nikon cameras and this looks good enough to get me going, for a reasonable price.

ChrisR · Post by **ChrisR** » Tue Dec 27, 2016 8:29 pm

Yes the second image is a crop from the very right edge of the first. Without zooming in at the edge of the image you (I) cannot see the CA.

It doesn't fit that. The first image is true monochrome - all pixels have equal values for R and G and B. They're all greys.
None of the pixels except those which are burned out, are grey in the "crop". Something else must have happened!

Here's your upper image (rotated, it's the LEFT edge) cropped, with your "cropped" image on top just translated lower. The outlined rectangle, I desaturated.

I don't have a stronger objective than the 20x currently, so don't know if it will be worse at higher magnification.

That's about 0.45mm, though, so the field you can use is about 9mm x 6mm on your sensor?

siliconGary · Post by **siliconGary** » Tue Dec 27, 2016 9:40 pm

ChrisR wrote:
It doesn't fit that. The first image is true monochrome - all pixels have equal values for R and G and B. They're all greys.
None of the pixels except those which are burned out, are grey in the "crop". Something else must have happened!
I don't know what happened, all I did was shoot in JPEG then opened the image in Preview (Mac) enlarged it 6x moved over to far right and cropped.

I have just repeated the process with another image - same process. The subject is a black printed on a glass slide that I put over a sheet of white paper, so it is pretty monochrome. For this image I opened it in Lightzone and checked that the pixels are not all monochrome before uploading it

cropped area

That's about 0.45mm, though, so the field you can use is about 9mm x 6mm on your sensor?
Thats a good observation so I am only using about 2.5M pixels of the 18M pixels. It is a shame they do not make a 1-1.25X adapter.

mjkzz · Post by **mjkzz** » Tue Dec 27, 2016 10:53 pm

SiliconGary, a side question, what chip is it?

siliconGary · Post by **siliconGary** » Wed Dec 28, 2016 8:10 am

Mike I don't know what the actual chip is or does. I know it is a GaAs chip made by Nortel in the mid 90's but don't now much more than that. I actually have the wafer attached to a commemorative plaque

I looked at and took a picture because the design, layout and look of a GaAs chip are very different from any silicon chip and not many people have seen them

siliconGary · Post by **siliconGary** » Wed Dec 28, 2016 8:23 am

ChrisR wrote:That's about 0.45mm, though, so the field you can use is about 9mm x 6mm on your sensor?

Chris I was thinking about that this morning - as I am using an RLA-vertical illuminator which has a tube lens I think the field of view on the sensor is greater.
If the adapter lens is 2x (As stated) and has correct focal length then the diameter it is imaging on the sensor I think should be 28mm/2 =14mm. That would mean that the total magnification is 14mm/0.46 ie ~30x. Is it reasonable to assume the RLA tube lens is magnifying 1.5X ? - I cannot find anything that gives the value of this lens.

mjkzz · Post by **mjkzz** » Wed Dec 28, 2016 9:26 am

Thanks Gary.

ChrisR · Post by **ChrisR** » Wed Dec 28, 2016 1:57 pm

Gary - I don't know. I pass to those with greater knowledge! I read that the adapter was 2x, then immediately omitted it from the sums, so I was a long way out.
Something seems to be not quite right.

Pau · Post by **Pau** » Wed Dec 28, 2016 4:21 pm

Quote:
That's about 0.45mm, though, so the field you can use is about 9mm x 6mm on your sensor?
Thats a good observation so I am only using about 2.5M pixels of the 18M pixels. It is a shame they do not make a 1-1.25X adapter.

Assuming that the whole micrometer image is not cropped and with each micrometer fine division measuring 0.01mm, your magnification on sensor will be 22.2/0.47 = 47.23X

If your adapter really delivers its nominal 2X (likely not exact) your epiiluminator is providing 47.25/40=1.18X , roughly 1.25X, a pretty standard specification for that kind of devices.

BTW, if the image is not cropped you're actually using all your camera pixels, what you're losing is a respectable portion of the objective primary image (and of the subject field, of course)

siliconGary · Post by **siliconGary** » Wed Dec 28, 2016 9:16 pm

Thank you Pau. Your maths make sense (I realized I mixed diameter with horizontal dimensions in my last post). It makes sense that I am losing a "respectable portion of the objective primary image" given the field of view is smaller than what I can see in the eyepieces.
It seems to me that the 1.67 NFK compensating ocular is considered an optimum for an APS-C sensor. Thus there is a missed opportunity for a 1.5-1.67x camera adapter that is not currently available. (Don't know if technically making a 1.5-1.67x lens is more expensive than a 2x lens)

Charles Krebs · Post by **Charles Krebs** » Wed Dec 28, 2016 10:12 pm

Gary,

What 20X objective are you using?