Using video for botanical close-up stacks

A forum to ask questions, post setups, and generally discuss anything having to do with photomacrography and photomicroscopy.

Moderators: rjlittlefield, ChrisR, Chris S., Pau

palea
Posts: 40
Joined: Mon Dec 17, 2018 8:09 pm

Post by palea »

gardenersassistant wrote:I have used the 45-175 a lot (on G3, G5, G80) with Canon 500D (+2 diopters), Raynox 150 (+4.8 diopters), Raynox 250 (+8 diopters), Raynox 150+250 or 250+250 stacked, Raynox MSN-202 (+25 diopters) and (testing only) Raynox MSN-505 (+32 diopters).
Cool, I've been considering the same configurations. How do you find the working distance with the higher diopter options? Raynox specifies the 202 and 505 at 32 and 18.5mm at infinity focus and I've seen 18.5mm also reported for the 150+250 combination. A thin lens model suggests the slab depth available from autofocus would decrease from 5mm with 150+250 to 1mm with the 505. Given sufficient working distance for field subjects to receive light, I was thinking a low cost, manual Arca Swiss screw drive rail might suffice for collecting multiple autofocus subslabs to stack. The thin lens model suggests variation in magnification across the focus range might be too great for this to work well, which I'm not sure of accepting but does have some support from tests I've made at diopters below the 150. (Also, not sure the tripod I have or some of the substrates I work on are sold enough to accurately return to position after such handling, but those are separate problems.)

Also curious as to your experiences with image quality. Rik has suggested against stacking close up lenses on a few occasions and also mentioned the 202 might degrade with departure from infinity focus, though the latter was in the context of a macro lens focusing to 310mm rather than the 45-175's 900mm. It seems like the 150 and 250 hold up well and I've found a few reports of the 202 improving with closer focus. Presumably there's some interaction with the specific behavior of the autofocus lens's entrance pupil. As a zoom with a fixed length body the 45-175 might be unusual in that regard.
dolmadis wrote:Is the alternative of running a 4K stream on a focus rail manually a do-able option?
This has been demonstrated by a couple of members here, showing some loss of image quality compared to frames collected without camera motion. The relation between rail feed, video frame shutter speed, and image quality doesn't appear to have been explored much. My rough math suggests depth rates of hundreds of microns to low single digit mm/s appear feasible. In principle, with a slow enough feed and smooth enough motion camera motion blur could be reduced to negligible.

The approach could also be used with 6k video, given a GH5 or certain more expensive bodies.

gardenersassistant
Posts: 190
Joined: Sun May 31, 2009 5:21 am
Location: North Somerset, England

Post by gardenersassistant »

palea wrote:
gardenersassistant wrote:I have used the 45-175 a lot (on G3, G5, G80) with Canon 500D (+2 diopters), Raynox 150 (+4.8 diopters), Raynox 250 (+8 diopters), Raynox 150+250 or 250+250 stacked, Raynox MSN-202 (+25 diopters) and (testing only) Raynox MSN-505 (+32 diopters).
Cool, I've been considering the same configurations. How do you find the working distance with the higher diopter options? Raynox specifies the 202 and 505 at 32 and 18.5mm at infinity focus and I've seen 18.5mm also reported for the 150+250 combination. A thin lens model suggests the slab depth available from autofocus would decrease from 5mm with 150+250 to 1mm with the 505. Given sufficient working distance for field subjects to receive light, I was thinking a low cost, manual Arca Swiss screw drive rail might suffice for collecting multiple autofocus subslabs to stack. The thin lens model suggests variation in magnification across the focus range might be too great for this to work well, which I'm not sure of accepting but does have some support from tests I've made at diopters below the 150. (Also, not sure the tripod I have or some of the substrates I work on are sold enough to accurately return to position after such handling, but those are separate problems.)

Also curious as to your experiences with image quality. Rik has suggested against stacking close up lenses on a few occasions and also mentioned the 202 might degrade with departure from infinity focus, though the latter was in the context of a macro lens focusing to 310mm rather than the 45-175's 900mm. It seems like the 150 and 250 hold up well and I've found a few reports of the 202 improving with closer focus. Presumably there's some interaction with the specific behavior of the autofocus lens's entrance pupil. As a zoom with a fixed length body the 45-175 might be unusual in that regard.
I'm afraid you have lost me completely with the technicalities. For example, I don't know what "slabs" are, or how to calculate with the thin lens model. Also I don't do much (or by some people's standards any) high magnification stuff, with my invertebrate stuff usually being full body or more distant shots of not very small animals captured hand-held with single images using minimum, sharpness-crushing apertures, mostly these days, as with all but the last of these examples, using a bridge camera (the last one used the 45-175 on a G5) . This sort of thing.

Image
1376 01 1358 03 2018_08_13 P1490474_DxO RAW SP7 LR7 1400h by gardenersassistant, on Flickr

Image
1376 02 1374 13 2018_09_01 P1530201_DxO RAW SP7 LR7 1400h by gardenersassistant, on Flickr

And when I do photograph smaller subjects like this I take the same approach.

Image
1226 03 P1360282_DxO SP7 LR7 2018 1300h by gardenersassistant, on Flickr

Image
0775 29 P1080062 LR by gardenersassistant, on Flickr

This is all a long way from stacking using sweet spot apertures with tripod and rails, so I'm afraid my thoughts about the sharpness of these close-up lenses are not likely to be useful to you.

As to working distances, they are very short with the more powerful setups, but I think that is the case whatever kit you use. Within the limits of my measurement capability, the 202 and 505 distances are what I get with them, irrespective of focal length/magnification, on a 45-175 on a G80.

I measured the working distance of a 150 + 250 on a G80 as around 75mm at 45mm focal length when focused at infinity, and around 70mm at 175mm.
Last edited by gardenersassistant on Sun Jan 20, 2019 5:04 pm, edited 1 time in total.
Nick

Flickr
Blog
Journey since 2007

Rework and reposts of my images posted in this forum are always welcome, especially if they come with an explanation of what you did and how you did it.

ChrisR
Site Admin
Posts: 8668
Joined: Sat Mar 14, 2009 3:58 am
Location: Near London, UK

Post by ChrisR »

I don't think I can remember seeing a shot with TWO flying insects nearly that big in the frame, that sharp; and the glob springtail is very impressive for something so darned small!
Well done!
Chris R

palea
Posts: 40
Joined: Mon Dec 17, 2018 8:09 pm

Post by palea »

gardenersassistant wrote:I just measured the working distance of a 150 + 250 on a G80 as around 75mm at 45mm focal length when focused at infinity, and around 70mm at 175mm. This is in line with what you would expect for 12.8 diopters (1000 / 12.8 = 78 ).
Thank you! Much appreciated; I was skeptical of the 18.5mm claim as it's inconsistent with the stacked diopters I've measured as well as Johan's Raynox calculator.
gardenersassistant wrote:I don't know what "slabs" are, or how to calculate with the thin lens model.
Oh, sorry, I took the bit about "I have used the 45-175 a lot [with diopters]" at face value. You can easily look up the close up lens and focus breathing calculations and build a spreadsheet or whatever if you want, but I'd say the modeling's only important if one's trying to estimate tradeoffs about what kit to buy. If you've already got stuff to measure that's going to be much more accurate.

Substack slabbing is mainly used for avoiding depth errors which result in improper hair crossings and such. My understanding is it's common in high magnification insect work, hence the Bugslabber utility for scripting Zerne. The method is simply that instead of stacking, say, 100 frames at once, you stack them in groups of 10 (or 20) so and then stack the resulting 10 (or 5) slabs to form the final image. In practice, the slabs should overlap some percentage but, as I've only done it a couple times, I don't have a good sense of the requirements. Others can comment more accurately, but a practical application might be more like combining 9 subslabs of 20 frames, each overlapping the next by 5 frames.

In principle, one can do the same thing if more depth is needed than an autofocus bracket can provide. For example, if a diopter's compressed the 900mm to infinity focus range of the 45-175 to run the 5mm from 65-70mm but you need 10mm of depth you could do something like post focus, move the camera 2.5mm, post focus again, move another 2.5mm, and do a third post focus. In theory, the result would be the desired 10mm of depth. However, it could be done with a pretty basic manual rail (or even a nodal slide) rather than hauling an automated rail and controller into the field.

I haven't seen that anyone's actually tried this but it also seems improbable I'm the first person to have though of it.
ChrisR wrote:I don't think I can remember seeing a shot with TWO flying insects nearly that big in the frame
22.6mm f/8 1/800 Panasonic FZ330 1/2.3" bridge camera. Sometimes I think we get too caught up in ILC sensor sizes and their associated limitations. Discussion above in this post is perhaps a case in point.

gardenersassistant
Posts: 190
Joined: Sun May 31, 2009 5:21 am
Location: North Somerset, England

Post by gardenersassistant »

ChrisR wrote:I don't think I can remember seeing a shot with TWO flying insects nearly that big in the frame, that sharp; and the glob springtail is very impressive for something so darned small!
Well done!
Thanks. We had a wasp nest in the garden last year and another one several years ago. Both times I spent many hours with wasps flying around my head to get around the quite large diffuser and in and out of the hole. I think it was 7,000ish captures over several days last year, mostly empty frames - wasps move very, very fast. Catching them in the frame was a matter of chance. If I waited until I saw one in the frame and then pressed the shutter button the frame would be empty.

Here's another one with two in the frame. Not filling the frame so much this time, because of wanting the third OOF one in the composition. This one used the 45-175 on a G80 with two Marumi 330's stacked, one reversed. I used it as an opportunity to compare a lot of camera/lens/close-up lens combinations.

Image
1364 06 2018_08_23 P1110730_DxO RAW SP7 LR7 1400h-2 by gardenersassistant, on Flickr
Nick

Flickr
Blog
Journey since 2007

Rework and reposts of my images posted in this forum are always welcome, especially if they come with an explanation of what you did and how you did it.

gardenersassistant
Posts: 190
Joined: Sun May 31, 2009 5:21 am
Location: North Somerset, England

Post by gardenersassistant »

palea wrote:
gardenersassistant wrote:I don't know what "slabs" are, or how to calculate with the thin lens model.
Oh, sorry, I took the bit about "I have used the 45-175 a lot [with diopters]" at face value. You can easily look up the close up lens and focus breathing calculations and build a spreadsheet or whatever if you want, but I'd say the modeling's only important if one's trying to estimate tradeoffs about what kit to buy. If you've already got stuff to measure that's going to be much more accurate.

Substack slabbing is mainly used for avoiding depth errors which result in improper hair crossings and such. My understanding is it's common in high magnification insect work, hence the Bugslabber utility for scripting Zerne. The method is simply that instead of stacking, say, 100 frames at once, you stack them in groups of 10 (or 20) so and then stack the resulting 10 (or 5) slabs to form the final image. In practice, the slabs should overlap some percentage but, as I've only done it a couple times, I don't have a good sense of the requirements. Others can comment more accurately, but a practical application might be more like combining 9 subslabs of 20 frames, each overlapping the next by 5 frames.

In principle, one can do the same thing if more depth is needed than an autofocus bracket can provide. For example, if a diopter's compressed the 900mm to infinity focus range of the 45-175 to run the 5mm from 65-70mm but you need 10mm of depth you could do something like post focus, move the camera 2.5mm, post focus again, move another 2.5mm, and do a third post focus. In theory, the result would be the desired 10mm of depth. However, it could be done with a pretty basic manual rail (or even a nodal slide) rather than hauling an automated rail and controller into the field.

I haven't seen that anyone's actually tried this but it also seems improbable I'm the first person to have though of it.
Thanks for the explanations. That is very helpful. I might have a go at slabbing.
palea wrote:
ChrisR wrote:I don't think I can remember seeing a shot with TWO flying insects nearly that big in the frame
22.6mm f/8 1/800 Panasonic FZ330 1/2.3" bridge camera. Sometimes I think we get too caught up in ILC sensor sizes and their associated limitations. Discussion above in this post is perhaps a case in point.
I have some interesting discussions about my approach sometimes. Equipment is one side of it - I have m43 with macro lens and extension tubes (in fact that is what I use all the time now - macro lens not extension tubes - for botanical close-ups) and for APS-C I have macro lenses, extension tubes, teleconverters and reversing setups. But I prefer using close-up lenses. (I almost forgot - I have full frame too now which I could use with the macro lenses, extension tubes and teleconverters I got for use with APS-C, but I haven't tried full frame yet for invertebrates. There haven't been any around since I got the A7ii. Haven't found any use for the A7ii yet actually. I thought it might have a "special something" for botanical close-ups, but it seems that it doesn't, or if it does I can't find it.)

What often surprises people is that, unlikely as it sounds, with invertebrates I get the same image quality using close-up lenses as with macro lenses etc. That is because I use minimum aperture whatever kit I'm using, and that is f/45 full frame equivalent (for example f/8 on 1/2.3"), and so diffraction is dominant, pulling whatever I use down to about the same level of poor sharpness. (So why do I do that? I like the DOF). It isn't quite as bad as it sounds because with close-up lenses the effective aperture does not change with magnification, so it stays at f/45 equivalent whatever magnification I'm working at. But I do get some tut tutting about it.

The other thing I like about using close-up lenses is that I can use autofocus, which is reliably accurate, and fast (no hunting). I use autofocus most of the time. Not for the wasps in flight. Like my one and only flies in flight session I had to use prefocus, but apart from that about the only time I use manual focus for invertebrates is when a spider is on the other side of its web, or visible but inside and veiled by a silken nest.

Oh, and I can use up to 1/2500 sec with flash with my bridge cameras, which is nice on bright days when I want to be sure that the flash is the dominant source of light to reduce motion blur/ghosting (as was the case with the flies in flight for example).

Image0723 034 2015_05_13 P1710606 LR 1300h by gardenersassistant, on Flickr

Image0723 049 2015_05_13 P1710689 LR 1300h by gardenersassistant, on Flickr
Nick

Flickr
Blog
Journey since 2007

Rework and reposts of my images posted in this forum are always welcome, especially if they come with an explanation of what you did and how you did it.

palea
Posts: 40
Joined: Mon Dec 17, 2018 8:09 pm

Post by palea »

gardenersassistant wrote:I have some interesting discussions about my approach sometimes.
I'm curious about the reversed Marumi 330. What was the setup like and how'd that work out? Achromat pairs are an option I have on my list to look into but I haven't done more than even sort of glance at the math yet. Assuming it's the middle close up that's reversed the difference between such an arrangement and a reversed lens seems like more one of degree rather than kind.
gardenersassistant wrote:(So why do I do that? I like the DOF)
Yeah, I encounter a lot of emphasis about sharpness when it's often DoF, subject motion, or some other factor which is limiting. As mainly a plant photographer I've the luxury my subjects are usually fairly cooperative when it comes to holding still, even when working distances are low.
gardenersassistant wrote:I have macro lenses, extension tubes, teleconverters and reversing setups.
I like my 1x macro but it doesn't compete with a close up on a telephoto for working distance. I've long found tubes and teleconverters unwieldy compared to close ups. Reversing too, though my efforts in that direction have been ad hoc and theefore not as elegant as using something like a Componon.

gardenersassistant
Posts: 190
Joined: Sun May 31, 2009 5:21 am
Location: North Somerset, England

Post by gardenersassistant »

palea wrote:I'm curious about the reversed Marumi 330. What was the setup like and how'd that work out? Achromat pairs are an option I have on my list to look into but I haven't done more than even sort of glance at the math yet. Assuming it's the middle close up that's reversed the difference between such an arrangement and a reversed lens seems like more one of degree rather than kind.
Here is how it looks on the FZ330.

Image
1426 5 2 x M330 on FZ330 with KX800 by gardenersassistant, on Flickr

It is the outer 330 that is reversed. I use a UV filter reversed as a lens cap.

Image
1426 6 Reversed M330 on M330, UV filter as lens cap by gardenersassistant, on Flickr

The magnification is between the Raynox 150 and 250. Most of what I do with invertebrates is with the Raynox 150 on the FZ330. I don't often have to put the Raynox 250 on it. With the 45-175 on m43 it is a different story. I am frequently having to change between the 150 and 250, sometimes for the same subject. (I like to have further out "environmental" shots as well as full body, or occasionally closer, shots.) That is the main reason I prefer to use the FZ330 rather than one of my G series cameras for invertebrates. The FZ330 also gives a wider range of magnifications than the 45-175, especially with the twin 330s because, being larger diameter than the Raynox 150, they have less vignetting at the wide end (no vignetting with the Raynox 150 or 250 on the 45-175).

The working distance with the twin 330s is a bit short at around 140mm on the FZ330 (vs the calculated 1000 / 6 = 167) compared to the 150 at around 210mm. As far as I recall the sharpness/detail/texture capture is very good (better than the 150 I think) at lower magnifications, but falls off compared to the 150 as magnification increases. That is especially the case with the two 330s connected normally. Reversing one of them helps with that. Here are some screenshots of comparisons from a series of test shots I did with a 10 UKP note last year, quite a lot of them.

Image
1426 1 List of banknote comparisons by gardenersassistant, on Flickr

These are with the two 330s connected normally (larger at Flickr)

Image
1426 2 2 x Marumii 330, normal connection, on FZ330 by gardenersassistant, on Flickr

These are with the outer 330 reversed.

Image1426 3 2 x Marumii 330, outer M330 reversed, on FZ330 by gardenersassistant, on Flickr

Here is a comparison with normal connection at the top, reversed at the bottom. The two sets of captures were not aiming at exactly the same part of the subject but aligning them puts one of them out to the edge, which is not a fair comparison, so I kept it with both in the centre of the frame. I think the difference shows up all the same.

Image1426 4 Normal 330 connection top, bottom reversed by gardenersassistant, on Flickr

As far as I recall (I did so many of these comparisons and I have to go and do other things now so I can't check) a single 330 has this fall off too, possibly worse, unlike for example the Canon 500D (which in turn I found to be very different from the Canon 250D, both copies of which that I have tested have been very poor).
palea wrote:
gardenersassistant wrote:(So why do I do that? I like the DOF)
Yeah, I encounter a lot of emphasis about sharpness when it's often DoF, subject motion, or some other factor which is limiting.
Absolutely.
palea wrote:As mainly a plant photographer I've the luxury my subjects are usually fairly cooperative when it comes to holding still, even when working distances are low.
You are fortunate. I live in a notoriously breezy location and plants tend to move around a lot here, especially of course the longer, thinner stemmed ones.
palea wrote:
gardenersassistant wrote:I have macro lenses, extension tubes, teleconverters and reversing setups.
I like my 1x macro but it doesn't compete with a close up on a telephoto for working distance. I've long found tubes and teleconverters unwieldy compared to close ups.
Yes, that is all consistent with my experience.
palea wrote: Reversing too, though my efforts in that direction have been ad hoc and theefore not as elegant as using something like a Componon.
I haven't done much reversing. Didn't like it. I have an APS-C reverser that maintains the electrical connections, so you still have control of aperture, but it caused ... not sure what the word for it is, flare maybe, anyway it wrecked images even after I tried flocking the recess at the front.
Nick

Flickr
Blog
Journey since 2007

Rework and reposts of my images posted in this forum are always welcome, especially if they come with an explanation of what you did and how you did it.

palea
Posts: 40
Joined: Mon Dec 17, 2018 8:09 pm

Post by palea »

gardenersassistant wrote:I live in a notoriously breezy location and plants tend to move around a lot here, especially of course the longer, thinner stemmed ones.
My hope is the present obsession with making AF-C capable of kids, dogs, and wildlife will eventually produce a camera which can handle a grass seedhead. Based on present and historic rates of AF improvement I'm not terribly optimistic about this, but such technology happening before I die isn't entirely implausible.

Mostly it helps I tend be working with low plants whose short stature limits wind forces and movement, though the downside is that also increases vibration frequencies and complicates frame alignments. There also tends to be a dawn lull in the windier locations where I am, though sometimes the window of having enough light before much wind starts just stays closed. I get up early a lot and tend to follow a priority order from more exposed to less exposed locations and taller to shorter. It mostly works most of the time.
gardenersassistant wrote:The working distance with the twin 330s is a bit short at around 140mm
I've only one 330, so can't duplicate the measurements exactly, but mine measures +3.12 (though don't take the last digit too seriously). That puts my calculated estimate for the working distance of a pair with one reversed at 145mm, fairly close to your measurement, and gives an effective focal length for the pair in the vicinity of 165mm.

Since repeating an element backwards is a standard technique for reducing aberrations it makes sense reversing one of two nominally identical close up lenses would yield improvement. (There's also a slight increase in EFL compared to non-reversed due to the reversing ring increasing the separation between the diopters but it may be overridden by the changes in principal plane placement from reversing.)

One could extend this to 250+reversed 250 sort of an alternative to the 202. More flexible and more compact but generally more expensive options are the 25mm Steinheil triplets and 30mm diameter matched achromat pairs Edmunds and Thor sell, which offer potential for longer working distances and maybe better correction as they're built for the purpose. Analogously, I just did some quick testing of equivalent asymmetric combinations with my Nikon close ups but the conclusion is the differences appear much smaller than with the 330s and I'd need to buy some rings and set up a better test target to do a proper job of it.

gardenersassistant
Posts: 190
Joined: Sun May 31, 2009 5:21 am
Location: North Somerset, England

Post by gardenersassistant »

palea wrote:My hope is the present obsession with making AF-C capable of kids, dogs, and wildlife will eventually produce a camera which can handle a grass seedhead. Based on present and historic rates of AF improvement I'm not terribly optimistic about this, but such technology happening before I die isn't entirely implausible.
You've got me curious. What is the problem with autofocusing on a seedhead?
palea wrote:Mostly it helps I tend be working with low plants whose short stature limits wind forces and movement, though the downside is that also increases vibration frequencies and complicates frame alignments. There also tends to be a dawn lull in the windier locations where I am, though sometimes the window of having enough light before much wind starts just stays closed. I get up early a lot and tend to follow a priority order from more exposed to less exposed locations and taller to shorter. It mostly works most of the time.
You are very, very much more analytic and organised about your shooting than i am! I just go out when I feel like it, wander around and see what catches my eye.
palea wrote:
gardenersassistant wrote:The working distance with the twin 330s is a bit short at around 140mm
I've only one 330, so can't duplicate the measurements exactly, but mine measures +3.12 (though don't take the last digit too seriously). That puts my calculated estimate for the working distance of a pair with one reversed at 145mm, fairly close to your measurement, and gives an effective focal length for the pair in the vicinity of 165mm.

Since repeating an element backwards is a standard technique for reducing aberrations it makes sense reversing one of two nominally identical close up lenses would yield improvement. (There's also a slight increase in EFL compared to non-reversed due to the reversing ring increasing the separation between the diopters but it may be overridden by the changes in principal plane placement from reversing.)

One could extend this to 250+reversed 250 sort of an alternative to the 202. More flexible and more compact but generally more expensive options are the 25mm Steinheil triplets and 30mm diameter matched achromat pairs Edmunds and Thor sell, which offer potential for longer working distances and maybe better correction as they're built for the purpose. Analogously, I just did some quick testing of equivalent asymmetric combinations with my Nikon close ups but the conclusion is the differences appear much smaller than with the 330s and I'd need to buy some rings and set up a better test target to do a proper job of it.
Interesting. I've used 250's stacked, very occasionally. I'll see if I can work out how to reverse one of them and do a comparison.
Nick

Flickr
Blog
Journey since 2007

Rework and reposts of my images posted in this forum are always welcome, especially if they come with an explanation of what you did and how you did it.

palea
Posts: 40
Joined: Mon Dec 17, 2018 8:09 pm

Post by palea »

gardenersassistant wrote:What is the problem with autofocusing on a seedhead?
If it's swinging much in wind at any sort of close up-ish magnification it's moving over different focus points and fairly out of focus most of the time. I've yet to have a body which can acquire and track one with any reliability. It seems strange to me that what's essentially simple harmonic motion or even simply stationary objects so often give AF-C implementations fits but I can understand finding the right something that's really blurry isn't easy, particularly if the lens that's on is a macro and therefore has slow AF.
gardenersassistant wrote:I'll see if I can work out how to reverse one of them and do a comparison.
Cool, I'll be curious how it works out if you get to it. At least where I am 49mm male-male reversing/coupling rings for reversing a front Raynox 150 or 250 on a middle 150 or 250 are easy to source. For reversing the middle one I'd go 46m-49f + 49m-49m and then 43f-43f (not so easy to find). In your case I suppose that might also start with a 52-49 step down.

gardenersassistant
Posts: 190
Joined: Sun May 31, 2009 5:21 am
Location: North Somerset, England

Post by gardenersassistant »

palea wrote:
gardenersassistant wrote:What is the problem with autofocusing on a seedhead?
If it's swinging much in wind at any sort of close up-ish magnification it's moving over different focus points and fairly out of focus most of the time. I've yet to have a body which can acquire and track one with any reliability. It seems strange to me that what's essentially simple harmonic motion or even simply stationary objects so often give AF-C implementations fits but I can understand finding the right something that's really blurry isn't easy, particularly if the lens that's on is a macro and therefore has slow AF.
My Sigma 105 macro is too slow to be useful in a breeze, but I find my Olympus 60mm macro is not too bad. Achromats on telezooms are faster, which is what I use for invertebrates, including ones that are moving. I find a macro lens is better for the botanical stuff though.
palea wrote:
gardenersassistant wrote:I'll see if I can work out how to reverse one of them and do a comparison.
Cool, I'll be curious how it works out if you get to it. At least where I am 49mm male-male reversing/coupling rings for reversing a front Raynox 150 or 250 on a middle 150 or 250 are easy to source. For reversing the middle one I'd go 46m-49f + 49m-49m and then 43f-43f (not so easy to find). In your case I suppose that might also start with a 52-49 step down.
I cobbled this together. It gives a big gap between the lenses.

Image1427 10 Raynox 250 + reversed Raynox 250 by gardenersassistant, on Flickr

Before I did the test shots I ordered another 49 to 58 to get them closer. That was a bit premature though because looking at the test results I'm thinking again of an MPE-65, but I do so little of the small stuff that I probably won't bother.

I used a G80 with 45-175 and KX800 twin flash. I did shots at 45, 100 and 175mm, with two 250s stacked normally, two 250s stacked with one reversed, and with an MSN-202. I put the raw files through DXO PhotoLab and Lightroom, output as full size JPEGs, which are in this album at Flickr.

I'm not impressed with any of them. They all get worse as the focal length increases. The reversed 250s look softer in the corners than the non-reversed pair. And the 202 looks horribly soft in the corners. (As I recall it's not just softness with the 202, but distortion. I remember applying a large amount of barrel/pincushion correction - don't recall which - to get one of my few MSN-202 images to stop looking very weird.)

To be honest I'm surprised I have got the results I have using stacked 250s and 202. If I had done these tests first I probably wouldn't have bothered. Probably best that I didn't - some of the results have been not completely horrible. But I would have thought that anyone wanting to work more seriously at these magnifications would want to be using better optics. I'd be very interested to see what a somewhat similar MPE-65 image looks like, to see just how much better it would be.
Nick

Flickr
Blog
Journey since 2007

Rework and reposts of my images posted in this forum are always welcome, especially if they come with an explanation of what you did and how you did it.

ChrisR
Site Admin
Posts: 8668
Joined: Sat Mar 14, 2009 3:58 am
Location: Near London, UK

Post by ChrisR »

I'd be very interested to see what a somewhat similar MPE-65 image looks like, to see just how much better it would be.
I think you might be waiting a long time, cos the MP-E doesn't stop down far enough! (f/16).
Physics of diffraction rules, but you've got perfectly acceptable sharpness in these web images. Ergo, you're using a better tool for the job.

I wouldn't "fix" it, unless you turn to static subjects and big prints people will walk up to scrutinize.

The second wasp pic is a cracker too, but I agree the composition is more satisfying in the first.

gardenersassistant
Posts: 190
Joined: Sun May 31, 2009 5:21 am
Location: North Somerset, England

Post by gardenersassistant »

ChrisR wrote:
I'd be very interested to see what a somewhat similar MPE-65 image looks like, to see just how much better it would be.
I think you might be waiting a long time, cos the MP-E doesn't stop down far enough! (f/16).
But at 5:1 nominal f/16 is effective f/96, and with the MP-E on m43 it would be f/196 FF equivalent. On the other hand the effective aperture of a close-up lens like the MSN-202 doesn't change with magnification (I've never found this documented anywhere but I have illustrated it in this video.)

The linked banknote shots were at f/16 on m43, so FF equivalent f/32, well within reach of the MP-E at those magnifications, even on full frame. (The scene widths for the linked shots were around 15mm at 45mm focal length, 7mm at 100mm and 4.5mm at 175mm, 4.5mm of course being the scene width for the MP-E at 5:1 on Canon APS-C.)
ChrisR wrote:Physics of diffraction rules, but you've got perfectly acceptable sharpness in these web images. Ergo, you're using a better tool for the job.
The wasps and shield bug were low magnification, done with a Raynox 150, which is what I use most of the time. The uncropped image for the globular springtail was around 10 to 11mm wide. I don't know what the scene width was for the larger springtail, but it was shot with the 45-175 at 45mm. So none of the shots embedded in this thread is in the range that I'm finding troubling with the 202 and stacked 250s in the linked album. It is the 100mm and especially the 175mm shots that don't look good to me. Here is the MSN-202 shot at 175mm, f/16, scene width 4.5mm. Would the MP-E (at effective FF equivalent f/32) be no better than this? (Genuine question. I don't know. But I'd be very interested to find out.)

(Full size at Flickr)

Image1427 09 P1660237 R202_PLab LR by gardenersassistant, on Flickr

I'm trying to get what I can with single shots. I suppose for stacking one uses a smaller effective aperture; let's see, f/2.8 at 5:1 gives ... oh ... f/16, which is f/32 FF equivalent on m43. Hmmm. Oh well, I'd still like to know what an MP-E shot of this sort of subject would look like.

Hold on a minute, what have I misunderstood here? It seems that with the MP-E on m43 the maximum aperture at 5:1 would be effective f/16 (in m43 terms now, not FF equivalent). But with the MSN-202 on the 45-175 the maximum aperture would be between f/4 to f/5.6 depending on focal length/magnification. I wonder how that would affect comparative sharpness between the two setups?

This can't be right. Or if it is surely it can't be relevant, for some reason. I'm out of my depth. Help!
ChrisR wrote:I wouldn't "fix" it, unless you turn to static subjects and big prints people will walk up to scrutinize.
Neither of those for me. :)
ChrisR wrote:The second wasp pic is a cracker too, but I agree the composition is more satisfying in the first.
Thanks.
Nick

Flickr
Blog
Journey since 2007

Rework and reposts of my images posted in this forum are always welcome, especially if they come with an explanation of what you did and how you did it.

palea
Posts: 40
Joined: Mon Dec 17, 2018 8:09 pm

Post by palea »

gardenersassistant wrote:I'm not impressed with any of them. They all get worse as the focal length increases.
Well, with the 45-175 at f/16 they're 2.2 MP effective and aren't going to pixel peep that happily. Downsizing to around 1700 x 1300 offers a more fair view in some ways and the histograms show the 202 is relatively overexposed, so it compares more favorably with the black point moved up as partial compensation. It also looks to me like the alignment tolerance is in the upper tens to a couple hundred of microns, which exceeds the depth of field and favors variable softness in the corners. I'm also unfamiliar with how wear smoothing's influenced the surface texture of the note. (There's also a potential confound if there's meaningful interimage variability in flash power.)

Within that, I believe this is the most controlled test of these combinations that's even been published. So that's very cool. It looks to me like 250 + reversed 250 might be better corrected for aberrations than 250 + 250 but have more field curvature.
gardenersassistant wrote:I wonder how that would affect comparative sharpness between the two setups?
I think effective MP is most commonly calculated by dividing sensor area by Airy first null disk size at 520 nm. Probably the most important bit is this measure of the Airy disk width exceeds the 3.75 um pixel pitch of a 16 MP u43 sensor at f/5.9, which is equivalent to stating your G80 is diffraction limited for single shots or focus bracketing beyond f/6. At least when the optics deliver a spot size smaller than the Airy disk.

The math can be done a number of other ways depending on which Rayleigh, Strehl, or Airy criteria one likes with what colors of light and how one wishes to account for the Bayer pattern on the sensor.

It can be a helpful exercise to explore this in a spreadsheet, R, or whatever tool you like. I made a table comparing u43, APS-C, and 135 and end up going back to it every so often.

(4k post focus---and, to a lesser degree, 6k on the G9---are a bit more complicated to analyze due to video sampling and compression.)
gardenersassistant wrote: (I've never found this documented anywhere but I have illustrated it in this video.)
Maybe try this thread. I'm not convinced the model described there is entirely correct but it at least captures some of the most significant factors. As a starting point, it's my understanding the Raynox 150 and 250 have a 37mm clear aperture, which makes them f/5.6 and f/3.4, respectively. 150+250 is f/2.1 and 250+250 is f/1.7. About a two stop advantage over most enlarging lenses and six stops compared to microscope objectives. I haven't seen a clear aperture for the 202 but my guess is 31mm, which makes f/1.3, though you can get a better number measuring yours.

The piece I suspect might missing be from Rik's model, as I understand it in the link, is calculation of the combined EFL of all of the lenses used and checking it doesn't produce some other limiting EA. This is unlikely to be significant at the >1x that discussion is concerned with but it might matter around 0.5x where's possible (though unlikely) to put f/2 or faster primes behind the f/2.8 or so provided by 150. It's also possible the model's entirely correct and it's just that the corner case is sufficiently obscure no one's ever noticed minor limiting incurred by the close up lens aperture.

I'd feel better about this if the model had more thorough discussion but, from what I can tell from what's currently available, it's only when maxing out the Raynox 202 that the EA of Raynox + 45-175 combinations might be limited by the close up lens rather than the 45-175. If it does happen the model predicts a 0.04 stop slowdown, so it'll be fussy to measure.

Post Reply Previous topicNext topic