macro, stereo, compound, SEM natural history guy from So Cal

Lets get to know each other better. Here's a forum to post images and short autobiographies of ourselves as well as any other info you would like to post about yourself.

Moderators: rjlittlefield, ChrisR, Chris S., Pau

dlg
Posts: 26
Joined: Wed Nov 04, 2015 11:47 pm

macro, stereo, compound, SEM natural history guy from So Cal

Post by dlg »

Finally joined here, after a bit of renewed interest in natural history imaging of mainly <1-5 mm shells and orchid flowers. Use a range of equipment from 5DsR with Zeiss 100MP/ZE, MPE65, stackshot, to Zeiss Discovery V20 mot with planapos 0.63 and 1.5x and Axiocam HRc, Axioskop IIplus with set of planapos, and am in charge of our institutional Zeiss EVO40XVP SEM. Also Arca 4x5" from close-up to about 1.5:1.

Been doing SLR for about >30 years, SEM for 25, LF for 10, z-stacking for about 5. I shoot mainly to illustrate my papers and books, some ended up in nat hist museum temp and permanent exhibitions. Have written a few isolated chapters and papers on photography over the years.

Greatest challenge currently is flash illumination of orchid flowers, that should look a bit more "natural". Also still learning the 5DsR and how it behaves in macro. Thus far figured out that above 4:1 on MPE65 there's only empty magnification. May also want to explore ZereneStacker a bit more; thus far just use Pmax, then edit in PS. Good enough, but maybe I can do better.

Pau
Site Admin
Posts: 6052
Joined: Wed Jan 20, 2010 8:57 am
Location: Valencia, Spain

Post by Pau »

Welcome aboard!

Will be nice to see some of your images here.
Pau

Lou Jost
Posts: 5944
Joined: Fri Sep 04, 2015 7:03 am
Location: Ecuador
Contact:

Post by Lou Jost »

I look forward to hearing about your orchid photography experiments and techniques. Tiny orchids are also my specialty (pleurothallids particularly).

dlg
Posts: 26
Joined: Wed Nov 04, 2015 11:47 pm

Post by dlg »

Lou Jost wrote:I look forward to hearing about your orchid photography experiments and techniques. Tiny orchids are also my specialty (pleurothallids particularly).
Hi Lou, pleuros are a lot of fun, too. If you want, you can check out
Geiger, D. L. 2013. Imaging small orchid flowers using visible light. Orchid Digest 77: 112–123.
Pdf can be downloaded from
http://www.vetigastropoda.com/abstracts ... papers.php

Does not cover the 5DsR, but still plenty of info and experiences. Maybe a bit basic if you hang out here, but you may still enjoy it. More images also in
Geiger. D. L. 2014. Oberonia: the microscopist's delight. Orchids 83(9): 558–563.
Geiger, D. L. 2014. Adventures in misidentified orchids. Orchids 83(3): 174–176.

Are you also in the Pleurothallid alliance? Maybe see you at the SF show in February? Found your website: nice work!

Lou Jost
Posts: 5944
Joined: Fri Sep 04, 2015 7:03 am
Location: Ecuador
Contact:

Post by Lou Jost »

Thanks for the links, I will enjoy reading them. You have an intriguing set of publications!

I used to be in the Pleurothallid Alliance but not now. Many of those people are friends though: Carl Luer, Ron Parsons, Mary Gerritsen, and many others.

I am currently working on a paper to split Neooreophilus (formerly Lepanthes) nummularia into many different species, based on population behavior (sympatry without interbreeding), DNA, and morphology. The latter is where microphotography comes into play. These are almost certainly pseudocopulation flowers, and the distinguishing features are the fake female genitalia on the tiny lips. I need to resolve the same kind of detail that insect photographers resolve when photographing genitalia.

So I came here to learn, and it has been very productive. I think I am nearly ready to make some acceptable stacks of these orchids' "genitals" for my publication. And the forum has opened my eyes to really small things. Very exciting.

dlg
Posts: 26
Joined: Wed Nov 04, 2015 11:47 pm

Post by dlg »

Have you thought about SEM for your flowers? You will need critical point dryer. I tried it with hexamethyldisilizane, but does not work even with 2 week incubation. That will be part of a ms I just submitted to AOS-ORCHIDS/Lindleyana section. The cell surface sculpture you get with SEM is simply unattainable with any light imaging. For Oberonia, the variety of surface sculptures in very discrete areas is astounding. Will be a great set of characters for systematics. The complementary nature of color-LM and B&W SEM is really powerful. Currently I have about 1-2K SEMs, hopefully will get some loans in from Naturalis/Leiden and Canberra in not too long. The Kew set was great!

Also getting into the mol. syst. game in collaboration with an old high school friend.

Haven't meet Luer yet (know his name, of course); Ron is a great guy and also have met Mary. I wrote a review of the fabulous Miniature Compendium for Orchid Digest. Visited Selby so met Antonio (Stelis), and will be in DC in April talking about small flower imaging, something that Tom Mirenda got rolling. The orchid world is just a village.

PS: did you see my other post here re 5DsR and z-stacking?
http://www.photomacrography.net/forum/v ... hp?t=29140

you may find that relevant and also shows some sample images.

Lou Jost
Posts: 5944
Joined: Fri Sep 04, 2015 7:03 am
Location: Ecuador
Contact:

Post by Lou Jost »

Yes, it is a small world. Tom Mirenda is a friend too.

For me SEM is out of the question. I live in a remote Andean village with no access to such technology, and have very little money (especially now after buying a bunch of macro stuff). But if I am very careful, and if I use monochromatic blue or green light, I think I can get a photo resolves the biologically relevant surface detail at cellular level, though not smaller details.

Today I was photographing the surfaces of a local Phragmipedium pearcei lip at 10-20x. As you say, the variations in surface texture in different parts of the flower were astonishing. I had no idea! I think very few people have seen this sort of thing. If the stacks look good I will post them here on the forum.

I have to work a lot with translucent specimens in alcohol. Have you found any good way to light them or treat them so that the surface details are more clearly visible? A low-contrast ghostly translucent white blob will not play nicely with stacking software, I fear.

dlg
Posts: 26
Joined: Wed Nov 04, 2015 11:47 pm

Post by dlg »

If you need a few SEMs, and can get the preserved flowers into the US, I can take some SEMs for you. Send me an e-mail directly for details.

Re EtOH specimens, they all go in the SEM, sorry. However, any standard stain such as carmine red or methylene blue should do the trick. This should give you structural contrast. I would then convert the images to B&W so reader are not mislead by the "wrong" blue/red color. If your camera can be set to B&W, then you may get marginally better resolution, because the interpolation from the Bayer pattern falls away. Eventually you hit the hard optical resolution limit, and then B&W won't help anymore.

One caveat though. If you have truly translucent structures, then you have multiple areas of the flower in focus at the same x-y coordinates. This will confuse the stacking algorithm. You may try with top illumination, and also limit the stacking range so that there are no two areas at the same x/y coordinate in different focal planes. Stacking is not the same as confocal microscopy.

Stacking subsets, and then stacking the partial stacked images may help. Or you may have to combine parts of the images in Photoshop with masking layers.

I hope that made sense.

Alternatively, you could dry the flowers. But that requires critical point dryer.

If you are really brave, you could also do osmium tetroxide impregnation. Nasty stuff! You need a lab with a good hood, double gloves, goggles. I'm "old-school" so stick my arm into EtOH no problems, but OsO4 is nasty, and not cheap either. However, it should render the entire flower opaque black.

Lou Jost
Posts: 5944
Joined: Fri Sep 04, 2015 7:03 am
Location: Ecuador
Contact:

Post by Lou Jost »

Thanks very much for the advice and the offer to make a few SEMs. One of my coauthors (Mark Wilson, Colorado State College, a pleurothallid guy) does have the ability to do SEMS now, but getting the flowers to the US is very hard because of permit bureaucracy here. On top of that, some of my specimens are unique and cannot be destroyed, as would happen in a SEM. Though perhaps a metal-coated flower could still be preserved somehow?

The point about the translucent specimens confusing the stacking algorithm is a serious one. So far I have only tried working with flowers freshly immersed in alcohol; though they had lost their colors they were not quite translucent. Stacking worked well on them. The key is to find some stain that makes them opaque. The osmium sounds frightening and I could surely not get it here anyway.

I've discovered that preserving a flower in an iodine solution makes it opaque, but this only works when the flower is fresh. If the flower was originally preserved in alcohol, putting it into iodine does not have the same effect. Most of my specimens are old and were originally preserved in alcohol, so I am stuck.

I think the stains you mentioned would not help-- they would just change the color, but the flower would remain translucent. Am I correct about that?

I wonder if Zerene can be made depth-aware. Maybe Rik can tell us. The photos could be ordered according to z-distance by the user, and then Zerene could pick the topmost layer that is in focus, and ignore the rest of the stack in that particular small x-y region.

Regarding BW, yes, when I do my green or blue monochromatic shots, I convert to BW. They look great. Since they are monochromatic they have no chromatic aberrations, of course, so they are sharper regardless of wavelength. But using blue light also doubles the theoretical resolution (compared to light that contains a lot of red) because of the short wavelength. Nevertheless the reduction in pixel efficiency wrt green light (because of the 2:1 ratio of green-sensitive pixels to blue-sensitive pixels) may make green light a better option in spite of its slightly longer wavelength. I have to experiment more with this. I do love my green and blue LED monochromatic flashlights though! Very cheap Chinese things but very powerful.

rjlittlefield
Site Admin
Posts: 23561
Joined: Tue Aug 01, 2006 8:34 am
Location: Richland, Washington State, USA
Contact:

Post by rjlittlefield »

Lou Jost wrote:I wonder if Zerene can be made depth-aware. Maybe Rik can tell us. The photos could be ordered according to z-distance by the user, and then Zerene could pick the topmost layer that is in focus, and ignore the rest of the stack in that particular small x-y region.
Doing this fully automatically is a hard problem that nobody has ever solved as far as I know. The main difficulty is with areas of the surface where there are not enough features for software to figure out that there's actually a surface there. Of course your subjects might not have that problem, and in that case perhaps something could be done. If you can upload a representative stack of source images, I'd be happy to take a look at that possibility.

Some people have successfully attacked the problem using a combination of slabbing and manual retouching. "Slabbing" refers to first processing the stack in groups of a few source frames, then combining the outputs of those groups to give a final result. The main value of slabbing is that it significantly reduces the number of frames that have to be accessed during the manual retouching. For an example that has different physics but is perhaps similar from the standpoint of the image processing, see Surface of dragonfly eye.
But using blue light also doubles the theoretical resolution (compared to light that contains a lot of red) because of the short wavelength. Nevertheless the reduction in pixel efficiency wrt green light (because of the 2:1 ratio of green-sensitive pixels to blue-sensitive pixels) may make green light a better option in spite of its slightly longer wavelength.
Using blue or even violet light should give you a noticeable increase in resolution. To work around the sensor issue, just increase your optical magnification by a factor of 2 or so, so as to spread details across enough blue-sensitive photosites to capture all the resolution that is available in the optical image.

--Rik

Ecki
Posts: 775
Joined: Wed Aug 13, 2008 2:04 am
Location: Cape Town, South Africa
Contact:

Post by Ecki »

Welcome on board!

Could you elaborate a little bit how you prepare plants and thin sections for SEM imaging, please?

Best,
Ecki

Lou Jost
Posts: 5944
Joined: Fri Sep 04, 2015 7:03 am
Location: Ecuador
Contact:

Post by Lou Jost »

Rik, thanks for your comment on my comment. The dragonfly-eye technique you linked to is ingenious.

I use slabbing, but have not yet tried it on my translucent specimens. I will be giving them a try soon. Thanks for offering to take a look at them if they don't work!

dlg
Posts: 26
Joined: Wed Nov 04, 2015 11:47 pm

Post by dlg »

Lou: I sure know about paperwork! Just offering to lend a hand. Flowers prepared for SEM are not "destroyed" at all. They are simply dried, and now attached to a SEM stub. I have all my own flower samples and all my Kew preps available for re-examination.

I think you could over-stain the flowers and make them opaque that way. Overstraining is a potential problem in histology, but in your case could be used to your advantage. Methylene-Toluene blue 1:1 aka Richardson's stain is a good dual action, blue-blue stain. Just do a sloppy de-stain.

The slabbing is what I alluded to as well. More labor intensive, but it can produce good results.

Looking forward to seeing your blue/green light experiments.

Ecki: Re SEM prep of flowers, you first dehydrate them in several changes of 100% EtOH, then they go in critical point dryer (I have a Tousimis 815 A). In the pressurized chamber, the EtOH is exchanged for liquid CO2. The liquid CO2 is heated to above 40 C at ~80 bar, so the CO2 goes beyond critical point into supercritical phase (a hybrid state with density of liquid and viscosity of gas), then the CO2 is slowly bled off. You end up with a bone dry flower that retains its natural shape (i.e., no "beef-jerky"). The cell surface structures look like fully hydrated.

The flowers are not sectioned, but mounted whole. Because of the greater depth of field inherent to SEM, there is no slicing involved, but it is single image, whole mount imaging. You could do image stacking, if you want, but usually not necessary.

If you want to do true sectioning in SEM, that gets a bit more involved and I have never done that. Requires rather pricey equipment. You can do sectioning for TEM, but that's completely different. Did it back in grad school class, but have not done it since. I've done some semi thin sectioning (2 µm) of plastic embedded material for transmitted LM, but TEM is generally 80–100 nm thin, so different beast altogether. For semithins, I use an ultramicrotome with freshly broke glass knives; for TEM you almost have to have a diamond knife.

Lou Jost
Posts: 5944
Joined: Fri Sep 04, 2015 7:03 am
Location: Ecuador
Contact:

Post by Lou Jost »

Flowers prepared for SEM are not "destroyed" at all. They are simply dried, and now attached to a SEM stub. I have all my own flower samples and all my Kew preps available for re-examination.


That's very interesting! I did not know that these could be saved reliably.
Methylene-Toluene blue 1:1 aka Richardson's stain is a good dual action, blue-blue stain.
I'll try that. Great suggestion. The blue will give nearly the same resolution advantage as using blue light. Make the subject monochromatic (almost) instead of the light!

Why do you recommend using a double stain? Less monochromatic...

Lou Jost
Posts: 5944
Joined: Fri Sep 04, 2015 7:03 am
Location: Ecuador
Contact:

Post by Lou Jost »

I suppose any blue fabric dye would work. Fabric is just cellulose, same as the cell walls of an orchid specimen. Though I suppose they may be soluble in alcohol. Fabric dyes are easier to find here than technical chemicals.

Post Reply Previous topicNext topic