Laowa 2.5-5x vs Raynox MSN 202 +telephoto vs LOM 3.7 on m43

[zzffnn]

Dear all,

I am using a micro four thirds camera Olympus E-M1 mk1 for handheld macro in the field. Looking to get better single frame photos between 2x-5x on m4/3 sensor, which is about 17mm wide.

No stacking (yet) for me. Just hand-held single frame macro in the field. Some 2-3x use for insects, and some 3x-5x use for snowflakes and plants.

The following represents the single frame image quality that I am currently getting with Raynox 250 over a 300mm zoom lens (Panasonic 100-300mm f/4-5.6 mk2) zoomed to 300mm (getting about about 3x), With brutal cropping and sharpening. Snowflakes are about 3mm in size.

https://www.flickr.com/photos/zzffnn/49 ... ed-public/

https://www.flickr.com/photos/zzffnn/49 ... ted-public

I want to get better image quality. Already bought a 1.25x dedicated macro lens (Olympus 30mm f/3.5 macro) for casual close-ups. Will try it out with 3mm snowflakes and brutal cropping; though I am not expecting much.

What I am really considering is:

Option #1

Laowa 25mm 2.5-5x macro lens. I know this is a good option (IQ, WD, compactness, design, ect), though it costs $399 new.

Option #2
Raynox MSN 202 + a 40-150 Olympus zoom lens (f/4-f/5.6, the "plastic fantastic") will cover approximately the same range at 2-4.2x:

https://www.four-thirds.org/en/microft/ ... _r_olympus

This option #2 will cost me only $80 (already own the zoom lens), but working distance is only around 20-30mm. Image quality is likely inferior than Laowa?

I am guessing the depth of field at the same f stop (for example, f/8) would be roughly the same, comparing this option vs the Laowa?

Will this option be a lot harder to hand hold than the Laowa at 4.2x (assuming same f stop)? Or about the same? My camera does in-body stabilization for the Oly zoom lens, but not with the Laowa (which is fully manual on m4/3).

Option #3
I have a LOMO 3.7x objective too and can probably DIY a helicoid rig with added aperture.

But I double if it will go down to 2x and am not sure where to put the aperture behind the objective (would that be as close as to the objective back focal plane?).

Likewise, I wonder for the same f/8 aperture, will this option provide about the same depth of field as the Laowa? About the same in terms of difficulty in hand holding?

Note this objective will be used stopped down to shoot single frame macro. Maybe an overkill and frustrating to use in the field, not to mention small working distance?

In my shoes, which option will you go with?

Thank you very much,
Fan

PS: Don't worry about focusing light intensity. I am getting a twin macro flash (Yongnuo YN-24EX) for.my Oly E-M1 mk1 camera, which will provide focusing LEDs. It (being native for Canon) won't do TTL on m4/3, though I don't really need TTL for macro.

[zzffnn]

Any educated guess would be better than no response. Thank you very much!

[rjlittlefield]

I have no experience with most of those lenses.

However the DOF question is simple: all lenses will have the same DOF at the same effective aperture and same magnification. There simply is no magic in this regard, when using ordinary lenses that make a sharp image on sensor. The only way to get more DOF is to stop down and accept the corresponding increase in diffraction blur, and the DOF/diffraction tradeoff is the same for all lenses.

--Rik

[zzffnn]

Thank you very much, Rik.

Can you see obvious diffraction blur in my snowflake images above?

I was (mistakenly) using f/22 for those 3mm snowflakes.
My speedlites provided more than enough light, but I think I might have done better by using f/11.

I could not get those snowflakes to lay flat, which was why I used a small aperture. But then, I probably used too small an aperture at f/22.

My eyes could easily see diffraction blur comparing f/11 vs f/5.6, using a perfectly flat subject. Though with a 3D subject not laying flat, it become hard to tell if blur come from insufficient DoF or diffraction?

[rjlittlefield]

Well, at the maximum resolution that I can get from Flickr, 2048 pixels, I don't see any sharp pixel-level detail.

With a snowflake it's also hard to know whether absence of crisp detail is due to optics or crystal formation or melting.

But yeah, f/22 on MFT would put you firmly in diffraction territory.

--Rik

[gardenersassistant]

As requested by PM, here are some thoughts.

I cannot comment on the use of microscope objectives as I have no experience in that area.

I am open to correction and learning on any of this from people who know more about this stuff than I do (plenty of them around here!) and/or are more careful than I have been about getting the details absolutely correct. So this is my best shot, FWIW.

Here are some scene widths and working distances I just measured as best I could.


1606 01 Scene width, working distance, Laowa 25 macro, MSN-202 on 45-175 by gardenersassistant, on Flickr

As to image quality, that is difficult to assess. I have done a lot of comparisons but at these sorts of magnifications I find it difficult to get results which are consistent. I will describe what I did here and you can make what you will of the results, which I can (FWIW) make available as raw files if you prefer.

First some groundwork. Meaningful comparisons involve (IMO) using different setups to photograph, as far as is practical, the same scene, with the same framing and the same depth of field.

"The same framing" is straightforward (although it may be tricky in practice). "The same depth of field" requires a bit of thought, especially in this case where we are comparing a close-up lens setup to a macro lens setup.

Given the same framing, what matters for the depth of field (and also for the loss of detail from diffraction blurring), is the "effective f-number". This is conventionally approximated with the formula

Effective f-number = Nominal f-number * ( 1 + magnification)

where "Nominal f-number" is the f-number you set on the camera lens.

For example, if you set the camera lens to f/8 and photograph something at 2:1, the effective aperture is 8 * ( 1 + 2) = f/24. So, as you photograph macros you are using a smaller aperture than you might think.

This is true for macro lenses, extension tubes, teleconverters (the sort that fit between the camera and the main lens), bellows and reversed lenses.

However, it is not the case for close-up lenses. For a setup involving a close-up lens on a telezoom lens changing the magnification does not (in practical terms) change the aperture. The effective aperture is (in practical terms) the same as the nominal aperture. The table below gives some examples, comparing the Laowa 25 macro with an MSN-202 on micro four thirds with a camera lens similar to the one you would be using.

The table also shows what f-numbers would be needed on full frame to achieve the same depth of field. My reason for doing this is that I use multiple sensor sizes and I find it convenient to think of all of them in terms of full frame equivalents. In particular, (with small sensor bridge cameras, MFT and APS-C) I use f/45 full frame equivalent when photographing invertebrates and I know this works for my purposes (given suitable post processing) and I know that I wouldn't want to go to much higher f-numbers. So, in the table below I have highlighted the f-numbers I used for the test shots I captured, which range roughly from effective full frame equivalent f/16 to to around f/45, which I think are realistic effective f-numbers to be considering, at least for single-image captures.


1606 02 Effective equivalent apertures for Laowa 25mm and MSN-202 annotated by gardenersassistant, on Flickr

For captures for stacking it would be good to use smaller effective f-numbers so as to reduce the detail loss from diffraction. Notice that as the magnification goes up with the Laowa you are less and less able to do this. On micro four thirds you can't get below f/32 when at 5:1, and similarly but less so at lesser magnifications. This is not a problem particularly with the Laowa, it would apply to any other lens used on micro four thirds at these magnifications. However, with a larger sensor, for example with APS-C or full frame you would be able to use smaller effective f-numbers, for example f/16 on full frame (with the camera lens set to f/2.8 at 5X) rather than f/32 on micro four thirds.

With a close-up lens on a telezoom camera lens you may have access to smaller effective f-numbers. For example the 45-175 I was using goes from f/4 at 45mm to f/5.6 at 175mm, so the smallest available full frame equivalent f-number would range from f/8 at minimum magnification to f/11 at maximum magnification.

As far as depth of field goes there are two cases to consider: depth of field suitable for a single-image capture; and depth of field suitable for multi-image focus stacking. In the first case there will be a balance to be struck between depth of field and diffraction softening. In the second case the setup's optically optimum aperture can be used (assuming, as I will here, that enough light can be thrown on to the scene).

The Laowa can get to the smallest apertures I would want to use for single-image capture and far beyond. The MSN-202 setup can just about get to the smallest apertures I would want to use for single-image captures.

For multi-image focus stacking, especially at higher magnifications like those you are considering, on micro four thirds the Laowa will suffer disproportionately from diffraction softening, compared both to full frame and to a close-up lens setup like the MSN-202, both of which can use smaller effective equivalent f-numbers (larger effective equivalent apertures) than the Laowa.

In practice though, how will these pros and cons balance out? That depends on your needs, priorities, preferences etc. The best I can do is to provide you with some test images to look at.

I captured five shots of each of the combinations highlighted in the above two tables. I used the same Panasonic G80 camera for both setups. I shot raw. I used a tripod but kept my hands on the camera. I used flash. I used manual focus for the Laowa and autofocus for the MSN-202.

The images within each set of five were very variable indeed (not unusual for my testing; others who are better at this stuff may get more consistent results). I picked what looked to me the best one of each five and put it through my usual post processing workflow, the only difference being that instead of ending with a 1300 pixel high JPEG I output full size JPEGs, which are in this album at Flickr https://www.flickr.com/photos/gardeners ... 013462522/. Here is a list of the images.


1606 2a Selected test images by gardenersassistant, on Flickr

Obviously questions will arise as to whether I have biased the outcome with the particular processing I have done, so I will be happy to provide a link to the raw files if that would help.

On other matters from your post .....

As you are reckoning on working hand-held I would be prepared for extremely variable results for single-capture images. At the magnifications you are considering I would be very surprised if you could get good stacking results entirely freehand. At the very least I would think you would need some sort of support such as a pole, or perhaps a monopod, or a tripod used as a dipod, or a tripod hands-on.

I think you will find it essential to use flash at those magnifications in order to get effective shutter speeds fast enough to get sharp results (if flash is the dominant light source the effective shutter speed will be the length of the flash pulse).

Image stabilisation becomes progressively less effective as magnification increases. If you were working hand-held I doubt it will help at all at the magnifications you are considering. And if you are using flash as the dominant light source I don't think it image stabilisation would be relevant in terms of the sharpness of the captured image (I suppose it might steady what you see in the viewfinder a bit, but at those magnifications I doubt that would be significant.

[zzffnn]

Thank you very much, Nick.

Your detailed comment is very helpful to me. I am leaning very heavily on the MSN-202 solution now.

Do you remember the working distance of Raynox 250 stacked with Raynox 150 (over your 45-175)?

I lost my Raynox 150. I had it stacked onto Raynox 250 when I had it, but cannot remember the working distance. Was it something around 50-60mm?

I may go buy another Raynox 150 and Canon 500D diopter, after buying MSN-202. Though I am not sure about Canon 500D, since I already have an Olympus 30mm f/3.5 macro (1.25:1) lens.

Yes, I always use speedlites and am getting that Yongnuo YN-24EX twin macro speedlites soon.

To me, lens or in-body stabilization helps me with focusing accuracy and provides higher chance of getting sharp focus onto where I want. But you are surely correct, with fast speedlites, stabilization does not do anything extra for image quality.

[gardenersassistant]

Thank you very much, Nick.

Your detailed comment is very helpful to me. I am leaning very heavily on the MSN-202 solution now.

Do you remember the working distance of Raynox 250 stacked with Raynox 150 (over your 45-175)?

With a Raynox 250 and 150 stacked on the 45-175, my G9 will autofocus at working distances of between 55 and 65mm. It is the same whatever the focal length (and hence whatever the magnification/scene width coverage).

btw an advantage of the 45-175 is that it does not extend when you change the focal length. This means that you can change magnification/framing without moving the camera (and there is a very conveniently placed lever on the lens which can be used with one finger to change the magnification/framing). This becomes increasingly useful as the magnification increases. It can help with getting very small subjects into the frame. You can zoom to wide angle/low magnification, find the subject, centre it in the frame and zoom in on it. This is also good of course when you lose a small subject.

I lost my Raynox 150. I had it stacked onto Raynox 250 when I had it, but cannot remember the working distance. Was it something around 50-60mm?

The range of working distances for the Raynox 150 by itself on the 45-175 on the G9 is around 155 to 215mm.

I may go buy another Raynox 150 and Canon 500D diopter, after buying MSN-202. Though I am not sure about Canon 500D, since I already have an Olympus 30mm f/3.5 macro (1.25:1) lens.

I haven't used the 500D for flowers since I got the Olympus 60mm macro. I have used it very occasionally for butterflies, either on the 45-175 on micro four thirds or on a small sensor bridge camera.

Yes, I always use speedlites and am getting that Yongnuo YN-24EX twin macro speedlites soon.

FWIW I pretty much never use flash with the 500D, because what I photographed with it was flowers and larger invertebrates such as butterflies and dragonflies. I prefer natural light for flowers, and on the rare occasions I do see larger invertebrates it is generally a bright sunny day and flash is unnecessary.

[elimoss]

Full frame equivalent aperture is applied when the magnification intended for MFT is half that of FF vertically, in order that the MFT image fill the frame in the same way as the FF image.

To make a bug fill the frame of an FF image, by comparison to an MFT sensor, I need the bug to take up twice the number of pixels vertically. Thus, its magnification on the image plane must be double.

So you could maybe say that a 0.5x image on MFT at 50mm f/2.8 has an FF equivalent of 100mm f/5.6 at 1x, as you need 1x to make the same image on FF. Although, in fact it is not that simple, since effective aperture increases with magnification.

As a general rule I think one should consider the idea of 'full-frame equivalent' to be something that applies at very low, non-macro magnifications only.

[gardenersassistant]

Full frame equivalent aperture is applied when the magnification intended for MFT is half that of FF vertically, in order that the MFT image fill the frame in the same way as the FF image.

To make a bug fill the frame of an FF image, by comparison to an MFT sensor, I need the bug to take up twice the number of pixels vertically. Thus, its magnification on the image plane must be double.

So you could maybe say that a 0.5x image on MFT at 50mm f/2.8 has an FF equivalent of 100mm f/5.6 at 1x, as you need 1x to make the same image on FF. Although, in fact it is not that simple, since effective aperture increases with magnification.

As a general rule I think one should consider the idea of 'full-frame equivalent' to be something that applies at very low, non-macro magnifications only.

In my first post I used full frame equivalent effective f-number, the "effective" part taking account of the magnification.

As I work with four different sensor sizes for close-up/macros of the same scene framings, and wanting the same depth of field whatever kit I use, I find it essential to use (effective) equivalence for f-numbers. Given the differences in sensor sizes (from 1/2.3" to full frame) I don't find magnification to be a generally helpful metric (nor the 1:1 definition of "macro" either). That is why I generally refer to scene sizes, using magnification only to calculate effective (equivalent) f-numbers.