macro experiments

pdh

Legend
With the better weather upon us up here next to the Gulf Stream, and the proliferation of growing things (both fauna and flora), my intermittent interest in macro gets stoked up again ... so much so that I start hunting eBay for bargain micro-nikkors ... there are none (not bargains, anyway) ... in the meantime, I dug out the old 105/2.8 Auto-Takumar and a set of tubes ... in fact, I just bought some m43 tubes so I can use the Hexanons as well ... five minutes outside the back door produced these ... not the finest examples in the world, but I like them ... all handheld and uncropped, but very lightly toned and sharpened

5658506298_4b219b7cf4_b.jpg

macro experiments by _loupe, on Flickr

View attachment 35527
macro experiments by _loupe, on Flickr

View attachment 35528
macro experiments by _loupe, on Flickr

View attachment 35529
macro experiments by _loupe, on Flickr

E-P2, Asahi Auto-Takumar 105/2.8, f/5.6, 1/250s-1/1000s, ISO200
 
more macro experiments

this time with a 35mm Summicron ... still handheld, and the wind was gusting ... anyway, what's interesting to me is pixel-peeping the dandelion head ... much CA in the Takumar shot, none really detectable in the Sumicron ... big difference in FL of course ... also with the Summicron, the focus distance is almost impossibly close ... but all fun ...

View attachment 35573
more macro speriments by _loupe, on Flickr
1/80s, f/11, ISO800

5662047726_4b62dd882d_b.jpg

more macro speriments by _loupe, on Flickr
1/500s, f/16, ISO800

View attachment 35575
more macro speriments by _loupe, on Flickr
1/400s, f/11, ISO800
 
the fly

Shared these elsewhere, but they work well here as well.

DSC04961.JPG
Join to see EXIF info for this image (if available)


DSC04965.JPG
Join to see EXIF info for this image (if available)
 
First a shot using a telescope eyepiece as a close up diopter ( the knot was around 5mm across):
5169466095_ce84ff93c0_b.jpg
Knot macro by Mike Kanssen, on Flickr

Then one combining a microscpe with microsoft ICE, stitching to get the entire wing visible (I was able to get higher resolution versions this way):
32899381576_7c722c08e4_b.jpg
fly wing low res. by Mike Kanssen, on Flickr

and going beyond the technical definition of macro (typically quote as 1:1 to 10:1)
the foot of a fly:
5313429325_b1d34c7f3f_b.jpg
House fly foot by Mike Kanssen, on Flickr

cracked pepper:
5321001260_ce3ffdb137_b.jpg
Queen Nefertiti ? by Mike Kanssen, on Flickr
 
I've been experimenting with microscope objectives for about four weeks.

I have an Olympus 4x infinity plan objective (ebay £30), which I've mounted on a (reversed as advised on closeupphotography.com) Raynox DCR-150, which in turn is mounted on Pentax bellows unit adapted to my Olympus EM1 MK2. The point of reversing the Raynox is that it definitely reduces CA if using it to mount microscope objectives on. There is no post production correction at all for CA in this picture.

This is a stack of 60 images via Zerene Stacker. I haven't calculated magnification exactly - it's about 3-4x on the MFT sensor, at around 140mm extension from sensor plane to the front (now rear) element of the Raynox. The Olympus objective is designed to be mounted on a 180mm tube lens, so it's possible to get closer.

Top tip: Zerene Stacker does not handle Prophoto RGB at all well - you need to export your tiffs in Adobe RGB colour space unless you like horrible muddy colours.

The camera with bellows is mounted on a Manfrotto 454 focusing rail. I use an Allen key to turn the drive wheel. A full rotation moves the camera 1.25mm towards the subject. By using the Allen key as a handle/lever, I can easily judge 16 incremental steps for a single rotation, so each step is about 80 microns. The rail is mounted on my tripod, which I've clamped to the stand holding my specimen so that the camera and specimen are not wobbling around relative to each other.

Lighting is a single Godox AD200 on 1/32 power with standard reflector, suspended above the wasp with a pop-up fabric diffuser above the lens, fixed to the top of my bellows with magnets. I set the 2 sec timer on the camera, which when added to the time taken for me to move the lever, gives about 7-10 sec between flashes. I've had no overheating problems with the AD200 using this method, my biggest stack has been 300 images.

Although many photographers use multiple lights for wrap-around illumination, with small subjects, I find it easier to work with one light, and bits of white paper or tin foil to direct fill-in. It's just more precise, and by taking test shots and scrolling back and forth through them, it's easy to see the effects of your lighting/reflector changes.

I put tin foil either side below the wasp to put some fill light lower down his eyes, and hung a piece of white packaging foam on my lens for additional fill-in directly in front of his face. I've covered the bright chrome microscope lens barrel with matt camo tape to avoid unwanted reflections.

The wasp was already dead when I found it a week ago, and since my bottle of ethanol has yet to arrive, he's decomposing rapidly. You should get the idea though. There is some artefact present, but as I practice, I hope to find ways to avoid or reduce it. Not choosing hairy insects would probably be a good start, but it's winter here, beggers can't be choosers.

I'll post a picture of my set up - but I need to wait until sun-up for an iphone pic because my spare body is in transit to Olympus for servicing.

wasp_face_142mm_raynox_rev_oly.jpg
Join to see EXIF info for this image (if available)



IMG_3529.jpg IMG_3531.jpg IMG_3536.jpg
 
Last edited:
I haven't calculated magnification exactly
A reasonable starting point is (1000 mm / 4.8 diopter DCR-150) / (180 mm design tube length / 4x) = 208 / 45 = 4.6x. For more accurate figures I measure the width of the subject size with calipers and do the math with respect to the imaging area of the sensor (e.g. 17.3 mm μ43 width / 3.80 mm subject size = 4.55x).

A full rotation moves the camera 1.25mm towards the subject. By using the Allen key as a handle/lever, I can easily judge 16 incremental steps for a single rotation, so each step is about 80 microns.
As an aside, autofocus bracketing may be something to consider. While Olympus lacks zooms with long ends near 180-200 mm, the Panasonic 45-175 and 45-200 both work well and the 45-200's inexpensive. Neither's enabled for in body focus stacking with Olympus but it's my understanding autofocus bracketing is still permitted. After that it's not any different than any other Zerene stack.
 
A reasonable starting point is (1000 mm / 4.8 diopter DCR-150) / (180 mm design tube length / 4x) = 208 / 45 = 4.6x. For more accurate figures I measure the width of the subject size with calipers and do the math with respect to the imaging area of the sensor (e.g. 17.3 mm μ43 width / 3.80 mm subject size = 4.55x).


As an aside, autofocus bracketing may be something to consider. While Olympus lacks zooms with long ends near 180-200 mm, the Panasonic 45-175 and 45-200 both work well and the 45-200's inexpensive. Neither's enabled for in body focus stacking with Olympus but it's my understanding autofocus bracketing is still permitted. After that it's not any different than any other Zerene stack.
I am so new to this that I hadn't really considered magnification. To be honest, I was impressed I'd got an image.

But magnification is important because one needs to ensure that one is using the best method for a given magnification, ie, macro lens, enlarger lens, tubes, reversed w/a or microscope objective.

I'll try the calliper method :)
 
This is clever stuff (and under appreciated). One day I'll land on a decent laboratory microscope at a good price, most likely when my youngest has been through uni.
Mine is far from a decent microscope, just a cheap digital microscope with a built in camera & screen rather than eyepieces. Using panoramic stitching compensaes for the low resolution of the camera.
The scope I can fit a camera to at work is reasonable quality, but isn't the one used for the Daphnia. I don't think I've shared any shots through that on Cameraderie yet. It has the huge advantage of being able to fit samples over 2" high while the digital one is limited to less than half an inch.
Here's one through the works scope ( again stitched)
5457424324_5a88a6b52d_b.jpg
PCB detail from an old mass spectrometer by Mike Kanssen, on Flickr
 
Last edited:
But magnification is important because one needs to ensure that one is using the best method for a given magnification, ie, macro lens, enlarger lens, tubes, reversed w/a or microscope objective.
From an image quality perspective it's pretty much autofocus macro up to ~1x and autofocus bracketing through infinity corrected microscope objectives above that. Both for optical reasons and managing perspective distortion. Infinity compound objectives for ~4–100x are reasonably straightforward, as you've probably noticed, though epi illumination gets increasingly involved from 20x up.

A practical difficulty, however, is ~2x requires either big, heavy, and usually expensive common main objectives or 2x compound objectives with frequently limited working distance that are mostly quite expensive. Reversed medium format enlarging lenses can be a good compromise there if you don't mind retouching stacking artifacts from longitudinal chromatic aberration. They're quite unlikely to pixel peep as well as objectives, in my experience, though. Whether any of those are better than pushing a 1x autofocus macro up to ~2x with closeup lenses or tubes mostly depends on personal preference and how much of a ~3x capability one's looking for.

What's stuck for me out all that so far is a 75 mm enlarging lens and cropping from 1x macro to where the enlarging lens picks up. Depends what magnification translates best between the subjects one's interested in and the sensor sizes one wants to use, though.

A reasonably specific magnification is needed (along with numerical aperture) to calculate DoF and thus step size, though that's a bit more of a linear motion than autofocus bracketing thing. With autofocus bracketing there tends to be less of a penalty to extra frames in the stack, plus the step size control manufacturers provide is more vague.

One day I'll land on a decent laboratory microscope at a good price, most likely when my youngest has been through uni.
One lower cost option is to add a second rail (Manfrotto 454, NM-180, MFR-150S, whatever) for xy capability, after which it's not much different from stitching frames if you've (quasi)telecentric optics. The involved part can be getting near telecentricity. Often the subject is mounted to the xy rig rather than the camera and it's convenient to autofocus bracket in z as that avoids need for a third linear motion axis. Another option is to look for a microscope stage rather than trying to find a whole microscope. A downside either way is you end up without a microscope to use for microscope things, though.

Personally, I'm content to use a lower magnification when a larger subject size is needed. I understand xy stack and stitch as an interesting technical challenge, though, and probably do more stacking of groups of autofocus bracketed stacks to extend depth of field along the optical axis than most.
 
From an image quality perspective it's pretty much autofocus macro up to ~1x and autofocus bracketing through infinity corrected microscope objectives above that. Both for optical reasons and managing perspective distortion. Infinity compound objectives for ~4–100x are reasonably straightforward, as you've probably noticed, though epi illumination gets increasingly involved from 20x up.

A practical difficulty, however, is ~2x requires either big, heavy, and usually expensive common main objectives or 2x compound objectives with frequently limited working distance that are mostly quite expensive. Reversed medium format enlarging lenses can be a good compromise there if you don't mind retouching stacking artifacts from longitudinal chromatic aberration. They're quite unlikely to pixel peep as well as objectives, in my experience, though. Whether any of those are better than pushing a 1x autofocus macro up to ~2x with closeup lenses or tubes mostly depends on personal preference and how much of a ~3x capability one's looking for.

What's stuck for me out all that so far is a 75 mm enlarging lens and cropping from 1x macro to where the enlarging lens picks up. Depends what magnification translates best between the subjects one's interested in and the sensor sizes one wants to use, though.

A reasonably specific magnification is needed (along with numerical aperture) to calculate DoF and thus step size, though that's a bit more of a linear motion than autofocus bracketing thing. With autofocus bracketing there tends to be less of a penalty to extra frames in the stack, plus the step size control manufacturers provide is more vague.


One lower cost option is to add a second rail (Manfrotto 454, NM-180, MFR-150S, whatever) for xy capability, after which it's not much different from stitching frames if you've (quasi)telecentric optics. The involved part can be getting near telecentricity. Often the subject is mounted to the xy rig rather than the camera and it's convenient to autofocus bracket in z as that avoids need for a third linear motion axis. Another option is to look for a microscope stage rather than trying to find a whole microscope. A downside either way is you end up without a microscope to use for microscope things, though.

Personally, I'm content to use a lower magnification when a larger subject size is needed. I understand xy stack and stitch as an interesting technical challenge, though, and probably do more stacking of groups of autofocus bracketed stacks to extend depth of field along the optical axis than most.

I think I've watched pretty much all of Allan Wall's video content since acquiring a set of Pentax bellows and a 100mm bellows lens on a whim (ebay, very inexpensive). I also dropped on an Olympus 4x infinity plan lens which was equally inexpensive, and being lower magnification, usable with my Manfrotto rail.

His conclusions are exactly the same as yours. Macro lens > reversed enlarger lens > objective. His videos are highly informative, although rather long. I think he over complicates his lighting sometimes too :) - but it works for him.

I've accumulated a lot of 'stuff' over the past few weeks, and have dusted off some other things that have been lying around for ages like old filters (remove the glass and stack them and you have a 50mm spacer for your infinity plan lens) and a Rodagon 80mm W/A enlarger lens which has sat in my garage for 10 years. Through experimentation, I'm hoping to keep the things that are most useful to me, sell off the things I don't need and maybe get a decent lab lifter and good quality stage. I think a stage with lateral movements could also be useful for stitching.

I can manage with a manual rail with the 4x - anything bigger and I'll need an automated rail ($$$$) and a rig that is screwed to the table or the wall. Like my camera system, I'll take a lot of convincing that I'm getting the absolute best out of what I've got before I get 'bigger', 'closer' or 'better'.

Someone should start an extreme macro thread. Too many people in the world are currently not allowed out again.
 
Update. The 50mm spacer between my infinity objective and reversed Raynox DCR 150 completely wrecked my contrast.

A 'found' wasp. Picked him up off my cooker hob yesterday thinking he must have just dropped down dead. Looking at the state of his eyes, I think he's been dead some time and probably dropped from my cooker hood.

Stack of 120 images, Zerene, Pmax. Olympus 4x, screwed directly to a reversed Raynox DCR 150, 144mm extension. EM1 Mk2.

Lighting is a diffused Godox AD200 directly above. The wasp is glued to a pin and blu-tacked inside a white cereal bowl for a bit of wrap around lighting. Starting to get a little flare though on the limb and antenna - going to try a shallower bowl before I abandon the idea altogether.

hob_wasp_bowl_no_objective_spacer.jpg
Join to see EXIF info for this image (if available)
 
His conclusions are exactly the same as yours. Macro lens > reversed enlarger lens > objective.
Er, they're different tools for different things. So I wouldn't rank like that at all. To reiterate my previous post in a different format,
  • ≤1x: autofocus macro lens
  • ~1-4x: messy but at the moment I like an enlarger lens on an autofocus telezoom
  • ≥4x: infinity objective on autofocus telezoom
I don't see these as strict categories. With μ43's short macro lens focal lengths getting to ~1.5x to maybe 2x with tubes isn't too involved. All the enlarger lenses I've used are soft enough cropping from a conventional macro at 1x works as just well (or better) up to around "1.3x", maybe more depending on the details of how you're stacking. With a zoom as a tube lens it can be tractable to push a 4x objective down to ~3x, depending on the zoom, objective, and sensor area. Given how well the Nikon BE and E Plan 4x work for their price the 2x BM13023121 is maybe also worth considering.

I can manage with a manual rail with the 4x - anything bigger and I'll need an automated rail ($$$$) and a rig that is screwed to the table or the wall.
If that's how you'd like to go, sure. I feel like I should mention I regularly autofocus bracket in the field at 5-10x from a lightweight tripod, though, and the main reason I haven't taken that to 20–50x is the cost of the objectives. If it's not sunny I use one LED torch, handheld, for lighting with some simple acrylic diffusers (I've actually multiple torches—a phone can work too—but only use one at a time).

I suspect linear motion bracketing with a tube lens does offer vignetting advantages, mainly with 20+x objectives and with APS-C and larger formats. However, data seems scarce.

Someone should start an extreme macro thread.
That's kind of photomacrography.net. Forum activity there emphasizes pixel peeping, DSLR based linear motion, and stacking insects with Zerene. It seems to me most of the regular posters on the site have yet to fully recognize what autofocus bracketing provides or to understand Zerene's limitations compared to Helicon and Picolay (both of which also have limitations compared to Zerene). The other caution I'd make is the often homebrew optical models favored by the regulars there do have some curious internal contradictions and, sometimes, seem strangely ignorant of standard approaches in microscopy.

Those disclaimers aside, it's still probably the most useful photomacrographic resource around. I do have some things I was thinking of posting to a photomacrography thread at mu-43 before it was bought out which would be suit a similar thread here. I'm not sure they're far enough along to move out of an experimental thread just yet, though.
 
Back
Top