Jump to content
UltravioletPhotography

macro z-stacking in UV


Daniel Geiger

Recommended Posts

Daniel Geiger

Been working on getting UV reflectance z-stacking done. This is

 

for ongoing botanical research into a particular genus of orchids, and to figure out who might possibly pollinate it.

Set up is (see image to get some idea):

Canon 5DII full spectrum, cheap Russian M42 bellows, Nikon EL 80 mm f/5.6 silver, Baader U-Venus. Light source, Metz 45 CT1 plastic shield replaced with Chinese UG11 type filter, triggered with slave from FlashZebra (the orange thing dangling from flash; not sure about sync voltage of the Metz), triggered by Canon MT26EX RT, single reflector at 1/512 power. stacking with Cognysis Stackshot, at 3:1 100 µm steps (2:1 with 176 µm steps), with 10s settlement to allow Metz to charge up again. The attached shot has 89 frames. Camera set to ISO 1600. Flash from left, aluminum reflector on right.

 

post-135-0-84329900-1514659007.jpg

 

Processing of RAW/CR2 with DxO using Premium Noise reduction (level 18). This is to get rid of color artifacts, that lead to swirly streaks in z-stack images. Previously tried to do that with a gaussian blur mask in Affinity, but the DxO noise reduction is WAY better. Then stacking in ZereneStacker with Pmax. In AffinityPhoto, steepen a and b channel curves in Lab color space, some color adjustment to make black background neutral, a bit of shadow lifting in L channel. Tried a bit of channel mixing, but don't like anything but the "natural" look.

 

post-135-0-49335700-1514659224.jpg

 

In vis light, everything is yellow-green. In UV, the rachis (central axis of flower stalk) is different than the flowers, but no patterns on the ~2 mm flowers. Note the granules are the cells. The species is Oberonia mucronata. The lack of patterns suggest dipteran pollinator.

 

I would like to get ISO down to 100. There are options with more light, and better lens transmission.

Dual Metz 60s would get me 2 f-stops so down to ISO 400. There are some dedicated UV flashes (Quantum). Worth it? Anything else?

How much better would a Coastal Optics UV lens be? I don't think I will get another f-stop out of that.

 

Any other suggestions?

 

I still don't get notifications for any replies from forum. Will check in periodically. Or send me an e-mail. geiger@vetigastropoda.com Thanks!

Link to comment

This is impressive! Thank you for showing us this result.

And it is nice to see the Stackshot in use for a UV stack.

 

Such tiny tiny flowers!

 

For maximum light I'm thinking that you might want to look into obtaining Blak-Ray lamps. A couple of those should supply sufficient light for a lower ISO setting. They will also tend to fry the specimen being photographed if not used carefully.

http://www.ultraviol...vf-photography/

Link to comment
Also, I looked at our Email Error Logs and found no errors. So perhaps also check your email Spam filter to make sure notifications are getting through?
Link to comment
Daniel Geiger

Dear Andrea,

 

finally got e-mail notifications on your postings. Glad to note that things seem to work.

 

Re blak-Ray, that is continuous light. Any idea on heat output? I'd be a bit concerned about movement of the specimen over time due to elevated temperature. That's why I prefer flash. I guess I could use quartz cold-light guides, but there is a huge hit in intensity, and there is also a price tag. Not sure about warm up either. I remember on epi-fluorescent microscopes, you had to let the Hg-lamps heat up for about 10 minutes.

 

Consider that the flash with filter and slave cost me about $100, I suspect that the blak Ray will set me back by just a tad more. Any idea on the Quantum UV flash? Significantly more UV than going old-school with vintage flash off eBay? I think I could also optimize geometry a bit more and take advantage of inverse square law. An inch closer in macro makes a big difference. I think I could snip off the inflorescence and get more organizing freedom.

 

Here's one more stack, that I even like a bit more. About 2.5:1 and 70 frames (Oberonia cf. padangensis).

post-135-0-11724200-1514697010.jpg

Link to comment

Lots of questions and clearly, areas of potential improvements. Last example seems very nice though.

 

Shooting up this close yet needing an ISO 1600 setting indicates the system output and response to UV is poor. A dedicated UV lens will provide "more light" as you could shoot at f/4-f/4.5 instead of f/5.6 or smaller with your current lens. However, I don't think the lens per se is the item most in need of replacement. More incoming UV and/or a better responding camera (for UV) might be better to concentrate on.

 

More light means using flashes with more UV output. However, a balance has to be struck as you correctly surmise, subject movement due to heat dissipation from the light source is an unpleasant side effect of bringing too strong flashes in too close. Thus a sensible balance has to be struck. The inverse square "law" will be ineffective or void for the close range, once you get the flash so close it no longer becomes a point light source. However, there will be a significant interaction between the reflector of the flash and the emitted light pattern. A bare bulb and a snoot are typically the least efficient for close-ups. In between these alternatives, there is a range of reflectors that can improve flash output dramatically for a given distance to subject. One can readily observe that certain reflectors, up close, have the ability to defocus the light instead of concentrating the light beam. Experimenting with a neutral subject and your flash at various distances facilitates finding its optimum position.

 

I am not familiar with your camera, but if it is capable of providing a daylight exposure of a typical UV flower at say f/8-f/11, 1-2 secs, ISO 100-400 with the EL-Nikkor, it is good enough for the task you put it to.

Link to comment
Daniel Geiger

Thanks for the various pointers. Don't know about daylight performance of my system. hopefully today it will clear up a bit and I can test it. I have a focusing helicoid so can do outdoor infinity shots.

 

Had been thinking about light modifiers on flash as well. May be add an aluminum sleeve that I can extend past the front element. I have done something similar on microscope objectives with paper sleeves for diffusion of external epi-source in visible light.

 

While the inverse square law may not be exact, I'm still in the range where it does affect light output. The UV window is about 1x2", so quite a bit smaller than the full surface of the Metz 45. Rest is taped off with silver tape, so should get as much internal reflection as possible. Reducing distance from say 12 cm to 5 cm should have a some decent effect.

 

One thing I wondered is whether the Chinese ZWB1 filter (allegedly UG11 equivalent) absorbs more UV than a true UG11. Cutting up a Baader U to put in front of the flash seems a bit excessive, so try to find a middle ground.

Link to comment
The Chinese glass apparently does not have tightly controlled tolerances and so the results could be better or worse than whatever spectrum was advertised.
Link to comment
Daniel Geiger
Alaun, I got a 45CT1 so only one (full) manual setting. After some 70 flashes 10 s apart, it sure gets warm in the upper part of the handle. Consider that effective f/stop at 3:1 is f/22. I did remove the plastic cover over the reflector housing.
Link to comment
Daniel, here in Australia we have a man who started photographing our Australian native terrestrial orchids then noticed various insects around them and discovered little had been done to photograph the pollination of orchids, his work is really amazing: Orchid Pollinators of Victoria, Rudie H Kuiter <rudiekuiter@optusnet.com.au> His images blow your mind, all under natural conditions outdoors under the Aussie sun using flash of course but most of the insects only fly on HOT days.
Link to comment

Why are you using a filter on the flash at all? For reflected UV I think you can shoot with full spectrum flash, and it may emit more UV light than when filtered.

The filter on the flash is reducing some of the UV output of the flash.

Your Baader U only lets the camera see the reflected UV, so that should be all you need.

Maybe give that a try and see if there is any difference.

Use the filter on the flash if you are shooting visual or IR fluorescence, but for that you should also stack the filter with S8612 2mm to suppress the Red/IR (~700nm~) range light that leaks through the 'UG11/U-340' type filter.

 

Quite an impressing set up you have there, I need a wiring diagram. ;-)

Link to comment
Daniel Geiger

Thanks for the pointer to Rudie Kuiter. Know of the name. I am working mainly on one genus (with 470 names, some 200-300 species) of orchids, but then globally. Occurs in QLD and N-most NSW, but not further south. Will be back down under in Fall, with Sydney Canberra and Brisbane on the list of herbaria to visit.

 

Re UG11 filter on flash, the idea is to cut out IR pollution by 3x, but have good UV transmission. The S8612 has less UV transmission, but better IR suppression. Maybe I am making things more complicated than need to be, and Baader U is all I need. Also got a 58 mm Baader alternative from kolarivision. Still have to test that as well.

 

May also have to upgrade the bellows. The M42 causes some vignetting on the FF body. I guess I should not have sold my OM or CY bellows. Oh well. Still plenty of data to print a 3" wide picture in a journal.

 

Re wiring diagram (joke well noted):

 

Stackshot X3 controller has incoming power from a Paul Buff Vagabond battery, Sends power to stackshot focusing rail, and has connector to N3 remote release on camera body. To keep weight down on camera body, I connect the Canon MT 26 flash with a 6' TTL cable from FlashZebra. Metz CT45 has a sync connector with mono jack, dual male jack connector, mono jack slave sensor (all from FlashZebra). The jacks are more secure connectors compared to PC sync.

Link to comment
Daniel Geiger
Forgot: went out today in daylight and get 0.5-1 s exposure at f/5.6 ISO 400 with Baader U. So sensor and filter seem fine.
Link to comment

Thanks for the pointer to Rudie Kuiter. Know of the name. I am working mainly on one genus (with 470 names, some 200-300 species) of orchids, but then globally. Occurs in QLD and N-most NSW, but not further south. Will be back down under in Fall, with Sydney Canberra and Brisbane on the list of herbaria to visit.

 

Re UG11 filter on flash, the idea is to cut out IR pollution by 3x, but have good UV transmission. The S8612 has less UV transmission, but better IR suppression. Maybe I am making things more complicated than need to be, and Baader U is all I need. Also got a 58 mm Baader alternative from kolarivision. Still have to test that as well.

 

May also have to upgrade the bellows. The M42 causes some vignetting on the FF body. I guess I should not have sold my OM or CY bellows. Oh well. Still plenty of data to print a 3" wide picture in a journal.

 

Re wiring diagram (joke well noted):

 

Stackshot X3 controller has incoming power from a Paul Buff Vagabond battery, Sends power to stackshot focusing rail, and has connector to N3 remote release on camera body. To keep weight down on camera body, I connect the Canon MT 26 flash with a 6' TTL cable from FlashZebra. Metz CT45 has a sync connector with mono jack, dual male jack connector, mono jack slave sensor (all from FlashZebra). The jacks are more secure connectors compared to PC sync.

 

I didn't say to use S8612 for UV reflected shots, but stack it with UG11/U-340 on the flash if you do any UVIVF.

There is no reason to use anything on the flash for UV reflected.

Give it a try, see what the difference is.

Link to comment

I second that advice. No need for any flash filter if one is using a decent UV bandpass filter with adequate IR-blocking over the lens (or at the rear of lens, or inside camera).

 

I'm shooting with a very powerful 1600 W/s Broncolor studio flash (uncoated xenon tube) and get excellent UV images at f/16-f/22, ISO 100, with the flash almost 1 m away from the subject. Only by directing the flash straight into into the lens is there some signs of IR leakage through the Baader U filter.

 

Such a powerful UV source necessitates the use of efficient UV goggles for eye protection. One will be very quickly reminded by extremely sore eyes that protective measures are required.

Link to comment
Daniel Geiger

So I tested the no-filter option on the Metz 45 CT1

post-135-0-52353200-1515279747.jpg

At closest distance, I can use ISO 400.

 

and also the Einstein 640 with glass dome removed and using a silver bowl reflector.

post-135-0-67799900-1515279748.jpg

At closest distance at full power I can get down to ISO 200. So UV yield is pretty poor. I could not measure flash output at the actual plant distance (out of range for Sekonic 558), at equal distance the Einstein gives me f/128.3 while Metz CT45 gives f/19.1, so 5 stops apart, but only double the UV.

 

 

Z-stacks are not the greatest, but this is just some testing of different light sources. Processing more or less the same (DxO premium noise reduction, steepening of curves in a and b channel of Lab).

The colors are similar. The blue is getting progressively more pronounced from Metz with ZWB1 -> Metz no filter -> Einstein. This could be some IR leakage as the relative proportions of UV to IR decrease with each set-up. Could possibly also adjust in AffinityPhoto to get closer color. At any rate, no radically different patterns. The difference is greater between w vs w/o ZWB1 on flash. That makes sense because of the interaction (multiplication) of filter curves of ZWB1 and Baader U.

 

I am looking in getting an Alien Bee 1600 with uncoated tube, and if my assessment is correct, then that result in slightly less blue images (and lower ISO/power).

 

True UG11 hasn't come in yet ...

Link to comment

So those are a comparison between two different flash units with no filter on the flash.

I am still not clear if the exposure time changed between with/with-out the flash being filtered or not.

 

Your questions about the Quantum UV flash and the Coastal Optics UV lens, you might want to ask Boon at this link below,

he uses both of those for UV shots, and should be able to answer any questions.

 

http://myphotojourne...wer/#more-10395

Link to comment
Daniel Geiger

Yes, comparison between two flashes, no filters. Exposure on camera is 1/200 flash sync time for all shots. Both flashes at full power so guesstimated in the ~1/1000s range.

 

Thanks for the pointer to Boon. Will check it out.

Link to comment

Daniel, Thanks.

You think the shots have more blue?

The Baader U will not leak blue, visual, or red/IR (700nm +/-), it will leak higher IR possibly, around 900nm, which can look white/gray in dark areas,

added blue is mostly likely from upper 400nm range.

I think I prefer the look of the latest shots you show. They don't look like they have any IR or visual in them.

Also, I liked your second flower shot you show on the first page, preceding the non-filtered flash shots.

Link to comment
Daniel Geiger

The images with Baader U in front of lens seem to me to have more blue-false-color contribution. I do understand that the Baader U eliminates all visible light 400-700 nm. What seems to happen is that the interaction of ZWB1 transmission spectrum, Baader U transmission spectrum, and camera CMOS spectral sensitivity interact in such a way that the blue channel of the CMOS gets higher readings. Haven't found a spectral response curve for a "naked" CMOS of the 5D2, the best is here

https://maxmax.com/s...al_response.htm

As I said, I could possibly adjust them closer to one another by curve adjustments in AP. I just note that with similar processing, the shots with non-ZWB1 filtered light source have more blue-false-color contribution. Baader U has a non-zero transmission around 700-900 nm, not much, maybe 1-2%, but it's more than in visible range.

 

Re no visible contribution, agree 100%. Just for comparison, same species (different specimen, but very similar otherwise) in visible light.

post-135-0-65754200-1515393251.jpg

Enjoy!

Link to comment

....the Coastal Optics UV lens

 

Just so that you are aware -- for very close work the Coastal Optics 60/4.0 requires an extension tube to mitigate its hot spot. Their website tells the magnification range where the hot spot occurs. A long lens hood also helps mitigate the hot spot, but this would get in the way for the close range in which you are working.

Link to comment
Daniel Geiger

Andrea: Thanks for the info on the CO 60. I use the dedicated conical lens hood on the Canon MP-E 65 mm, and it is no problem even at 5:1 mag. May also be a good alternative for the CO lens. At 3:1 I have gobs of working distance with the Nikon EL 80. I've been thinking more about the CO 100 mm anyway, as it goes to 1:1 with no extension and has better UV transmission <360 nm. Half an f-stop slower, and a bit more extension required for same magnifications, those are the trade-off. In fact, better transmission may offset the f-stop for overall exposure.

 

For IR, I use my regular Zeisses, don't have a 60, but the Otus 55 is very close and a spectacular lens in its own right.

 

Sorry for slow reply. That last notification ended up on spam filter. Weird.

Link to comment

Please sign in to comment

You will be able to leave a comment after signing in



Sign In Now
×
×
  • Create New...