Proof of concept: UV-reflectance stacking with MP-E 65

Zeidora

EOS 7D MK II
Feb 15, 2015
668
10
danski0224 said:
Have you considered trying the Coastal Optics UV IR VIS macro lens?
The Coastal optics lenses cost >$5K each, while the Nikon EL 80/5.6 was $40. The UV Nikkor is about $10K, if you can find it anywhere. I have not seen any comparison on the optical performance at >>1:1 of the Coastal Optics or the UV Nikkor lenses to enlarger lenses. Given the rather experimental nature of my project, I am not quite ready to drop 5-10K on it.
 

Zeidora

EOS 7D MK II
Feb 15, 2015
668
10
Danki: have you tried/compared the Coastal optics in for that purpose? Lens rental wants $226/week for it; I do live in the US. If I feel the need, I will buy it, but not quite there yet. Some 20 years ago I used the UV Nikkor at University, but for a UV fluorescence class assignment, which is a waste of that glass. Oh, the follies of youth.

Also, solved the Zerene file reading issue. Had to uncheck the "use external tif reader" in the Zerene Preferences. Rik L. kindly suggested that on the Photomicrography forum.
 

danski0224

EOS 6D MK II
Apr 24, 2011
1,089
4
Zeidora said:
Danki: have you tried/compared the Coastal optics in for that purpose? Lens rental wants $226/week for it; I do live in the US. If I feel the need, I will buy it, but not quite there yet. Some 20 years ago I used the UV Nikkor at University, but for a UV fluorescence class assignment, which is a waste of that glass. Oh, the follies of youth.

Also, solved the Zerene file reading issue. Had to uncheck the "use external tif reader" in the Zerene Preferences. Rik L. kindly suggested that on the Photomicrography forum.
Nope, I haven't tried that lens. I have a Sigma camera that will do full spectrum (no mods other than to remove the dust protector), but it's challenging to get Nikon lenses to work with it, much more difficult than the easy adapter for Canon. Right now there is only one Sigma DSLR camera with live view (and EVF), and that is the new sd series.

Having a Canon converted to full spectrum has crossed my mind, but it would have to have live view. An EVF would also be nice, which points at the latest M. I am near a point where I would consider renting one that has been converted to see if I like it.

$226 seems like a bargain rental to see what the Coastal lens can do.

I saw a UV Nikkor on eBay not too long ago for just under $10k. Haven't used one of those, either.
 

Zeidora

EOS 7D MK II
Feb 15, 2015
668
10
danski0224 said:
I have a Sigma camera that will do full spectrum (no mods other than to remove the dust protector), but it's challenging to get Nikon lenses to work with it, much more difficult than the easy adapter for Canon. Right now there is only one Sigma DSLR camera with live view (and EVF), and that is the new sd series.

Having a Canon converted to full spectrum has crossed my mind, but it would have to have live view. An EVF would also be nice, which points at the latest M. I am near a point where I would consider renting one that has been converted to see if I like it.

$226 seems like a bargain rental to see what the Coastal lens can do.

I saw a UV Nikkor on eBay not too long ago for just under $10k. Haven't used one of those, either.
There are M42 to Sigma adapters, eg. http://www.adorama.com/kasgpxs.html No problem putting a Nikkor EL on a Sigma. Live view/EVF, though, is almost a must for z-stacking. Allegedly, the EL80 has relatively little focus shift in UV, but screwing on a filter after you adjusted a >>1:1 is asking for trouble.

Re renting a Coastal lens, what I do is ongoing. If I get one, I buy it. Rental would just be paying $226 more in final purchase price.
 

Zeidora

EOS 7D MK II
Feb 15, 2015
668
10
Maximilian said:
Zeidora said:
Here's now one that I am reasonably happy with. Oberonia costeriana. Flowers are about 2 mm long ...
Really fascinating, Zeidora.

If I can recall it right, those UV patterns sometimes look even more geometrical.
I wonder if it's just because these flowers are just so small?
Or are geometrical patterns not so common?

Do you know more about that?
Sorry, missed that. The "why" question is the whole point here, and goes towards my botanical research. Photography is just a tool to investigate the biology of those flowers. Not sure whether there is a size correlation to UV patterns. Yellow sun flowers are famous for their concentric rings in UV. I've seen some lines on a yellow pleurothallid orchid (also relatively small flowers, also many flower on inflorescence). Current working hypothesis is that pollinator is not visually oriented, but rather attracted by odor. That could fit the limited observations on fungus gnats, and one species at least smelling like fungus (Oberonia fungum-olens).

To get a reasonably founded answer will take years. Research is painfully slow.
 

Maximilian

The dark side - I've been there
Nov 7, 2013
2,328
186
Germany
Zeidora said:
Maximilian said:
Zeidora said:
Here's now one that I am reasonably happy with. Oberonia costeriana. Flowers are about 2 mm long ...
Really fascinating, Zeidora.

If I can recall it right, those UV patterns sometimes look even more geometrical.
I wonder if it's just because these flowers are just so small?
Or are geometrical patterns not so common?

Do you know more about that?
Sorry, missed that. The "why" question is the whole point here, and goes towards my botanical research. Photography is just a tool to investigate the biology of those flowers. Not sure whether there is a size correlation to UV patterns. Yellow sun flowers are famous for their concentric rings in UV. I've seen some lines on a yellow pleurothallid orchid (also relatively small flowers, also many flower on inflorescence). Current working hypothesis is that pollinator is not visually oriented, but rather attracted by odor. That could fit the limited observations on fungus gnats, and one species at least smelling like fungus (Oberonia fungum-olens).

To get a reasonably founded answer will take years. Research is painfully slow.
Thanks for your reply.
 

scyrene

EOR R
Dec 4, 2013
2,356
195
UK
www.flickr.com
I'm currently sort-of experimenting with this myself. I don't have a modified camera or proper filter, but I've found the results with an astro light pollution filter quite good even if I'm crossing streams, so to speak :) Here's my first direct comparison (few flowers here at this time of year!), it's the rear underwing of a swallow-tailed moth, Ourapteryx sambucaria, in visible vs filtered UV light. Focus stacks using the MP-E and Affinity Photo. Surprisingly few moth specimens I've tried show much in UV but this one was interesting enough to persist with.
 

Attachments

Reactions: Berowne

Zeidora

EOS 7D MK II
Feb 15, 2015
668
10
scyrene said:
I'm currently sort-of experimenting with this myself. I don't have a modified camera or proper filter, but I've found the results with an astro light pollution filter quite good even if I'm crossing streams, so to speak :) Here's my first direct comparison (few flowers here at this time of year!), it's the rear underwing of a swallow-tailed moth, Ourapteryx sambucaria, in visible vs filtered UV light. Focus stacks using the MP-E and Affinity Photo. Surprisingly few moth specimens I've tried show much in UV but this one was interesting enough to persist with.
Neat. I think what you are doing is essentially a blue channel image. There are some faint yellow lines visible in the first image, and those are black in the second. Yellow = no blue, so those yellow areas will be dark. Compare the blue channel information from the first image with the second, and I would think they will be very similar.

To image UV light, you cannot get around a modified camera body.
 

scyrene

EOR R
Dec 4, 2013
2,356
195
UK
www.flickr.com
Zeidora said:
scyrene said:
I'm currently sort-of experimenting with this myself. I don't have a modified camera or proper filter, but I've found the results with an astro light pollution filter quite good even if I'm crossing streams, so to speak :) Here's my first direct comparison (few flowers here at this time of year!), it's the rear underwing of a swallow-tailed moth, Ourapteryx sambucaria, in visible vs filtered UV light. Focus stacks using the MP-E and Affinity Photo. Surprisingly few moth specimens I've tried show much in UV but this one was interesting enough to persist with.
Neat. I think what you are doing is essentially a blue channel image. There are some faint yellow lines visible in the first image, and those are black in the second. Yellow = no blue, so those yellow areas will be dark. Compare the blue channel information from the first image with the second, and I would think they will be very similar.

To image UV light, you cannot get around a modified camera body.
If I follow you right, you're saying that desturating the green and red channels in the visible light image should give the same result, is that correct? I don't think that can be quite right. As a comparison, using this technique made green leaves bright red. That would not be the case if I did what I think you are suggesting.

I agree a modified camera must be better, but UV filters are sold for a reason - and not just to protect the lens's front element! The sensor must be sensitive to some UV, or otherwise to fluorescence in UV? I'm no scientist, so apologies if I'm way off :)
 

Zeidora

EOS 7D MK II
Feb 15, 2015
668
10
You're right, something else is going on. So most likely fluorescence, i.e. UV light is shifted into visible light wavelength.

All sensors are UV sensitive to around 320 nm, but there is filter in front of the sensor that cuts off UV at around 400-420 nm. That filter is taken off when a camera is converted, and replaced with a UV transparent filter. In order to take UV images, you need to exclude all visible light, and also the IR. Sensors are less sensitive to near UV than to visible light, so you need to exclude ALL visible light. If you look through your filter against a bright light source and you see anything, then it is not suitable for UV reflectance imaging. Hope that helps a bit.
 

scyrene

EOR R
Dec 4, 2013
2,356
195
UK
www.flickr.com
Zeidora said:
You're right, something else is going on. So most likely fluorescence, i.e. UV light is shifted into visible light wavelength.

All sensors are UV sensitive to around 320 nm, but there is filter in front of the sensor that cuts off UV at around 400-420 nm. That filter is taken off when a camera is converted, and replaced with a UV transparent filter. In order to take UV images, you need to exclude all visible light, and also the IR. Sensors are less sensitive to near UV than to visible light, so you need to exclude ALL visible light. If you look through your filter against a bright light source and you see anything, then it is not suitable for UV reflectance imaging. Hope that helps a bit.
Thanks for clarifying :) I wonder what the precise cutoff for the in-camera UV filter is? My light source is rated at 395nm so very close, but just under 400. I guess I'll have to get an appropriate filter and see if anything is visible in just that light. I rather like some of the images I've taken with the current setup, however, so maybe it's worth pursuing for its creative merits in any case. Thanks!
 

Zeidora

EOS 7D MK II
Feb 15, 2015
668
10
395 nm is very close to visible. It is essentially the "Disco" light for glowing cloths. There are 365 nm flashlight, and I am currently working on modifying a flash for UV photography: remove yellow plastic UV filter, and replace with UG11. If you are interested in this, look at the http://www.ultravioletphotography.com/content/ forum. UV is hazardous, so take appropriate precautions.
 

Zeidora

EOS 7D MK II
Feb 15, 2015
668
10
Here's am update on my experiments, with which I am quite happy.

Set-up: Canon 5D2, Lens Nikon EL 80 mm f/5.6, Baader U-Venus filter on lens, cheap Russian M42 bellows in between; various adapters all throughout. Illumination with Metz 45 CT1, removed plastic front shield and replaced with cheap Chinese UG11 equivalent filter. Aluminum foil on opposite side brightens right side of image. Metz is triggered via FlashZebra slave, with Canon MT26 EXRT connected to 5D2, firing single head at 1/512 power. The whole rig on Cognisys Stackshot. Have to put 10s firing delay in program, so that the flash can re-charge between firings. Flash head gets quite warm after 60-100 shots.

Processing through DxO with Premium noise reduction removing all color artifacts. No more messing with gaussian blur masks in AffinityPhoto. Easier, but takes a lot of time on computer. Then in AffinityPhoto, steepened a and b curves in Lab space quite bit (lower corners moved halfway to middle, so doubled slope). Plus a bit of fine tuning so that black background is neutral in color.

Biggest problem is low power of flash. Have to go to ISO 400-1600 at full power of flash to get enough UV through the system. May get two Metz 60 flashes, so that should get me in ISO 100-400 range.

Still no pattern on flowers, but there is some differentiation between flower and rachis. About 2.5:1, some 70 frames at f/5.6 and 100 µm steps.
 

Attachments

privatebydesign

Would you take advice from a cartoons stuffed toy?
Jan 29, 2011
7,331
316
118
So you are using a $1,000 flash as an optical triggering device to trigger an underpowered $20 flash? I don't see that as making any sense, why not just plug the Metz into the PC port on the camera?

But a simple 'studio' light is going to give you way more bang for the buck anyway and get you to the power output you need rather than mess around with AA batteries.
 

NancyP

EOS 6D MK II
Dec 17, 2013
1,296
14
Very cool. What orchid? The local terrestrial orchids (Spiranthes) tend to be of this size or a little larger (2 to 3 mm at most). https://en.wikipedia.org/wiki/Spiranthes

Also, do you have a recommendation on a 365 nm UV flashlight?

This right up front, this is not a great picture, it is a proof of concept, to be refined.

I am interested in UV reflectance photography and to see whether there are insect patterns not visible in "normal" light on some minute orchid flowers. The flower is 1.76 mm (no, not a typo) wide. Shot with full-spectrum modified 5D2, MP-E 65 at f/2.8, manually stacked 10 frames, Baader UV-pass visible opaque filter, ISO 400, about 4 second exposure, illuminated with single UV 365 nm LED flashlight.

First off, the MP-E 65 DOES transmit some UV. That's already great. I also have a Nikon EL 80/5.6 but still waiting for some parts to arrive to put it all together. Should have much better UV transmission, so will be interested to see the differences.

Further refinements are additional UV flashlights (can't diffuse them because of absorption and fluorescence issues), and stacking with StackShot (have that).

Some single shots on some other orchid flowers do show some different patterns, so that works as well. The dotted highlights, those are the individual cells, with their rounded surfaces (verified by scanning electron microscopy). In visible light, the flower is uniformly green.
 

Zeidora

EOS 7D MK II
Feb 15, 2015
668
10
Have you considered trying the Coastal Optics UV IR VIS macro lens?
Bought the 105 Coastal recently and am currently working out some protocols. I also upgraded to a Paul C Buff white light 1600 with uncoated, UV-emitting tube. Now I can keep ISO at 100 with Baader U. The problem is that the heat blast wilts the flowers during the stacking sequence. Have to see whether greater distance but higher power will be better than close distance low power shooting. Still a work in progress ...