MF Macro examples

[ErikKaffehr]

Hi,

Thanks for chiming in. Checked out your article and liked it.

I have still not found out the issue OP has, but it seems he made progress and that is a good thing.

Best regards
Erik

That's not correct.

Diffraction is based on effective aperture, not marked aperture (unless your camera system makes adjustments to account for bellows-loss for the displayed aperture).

This effect becomes quite obvious when dealing with very high magnification imaging. At 18:1 magnification diffraction on an 80mp back is meaningful starting at a lens marking of f/2.8 .

Think of it from the perspective of the sensor: Here is the same sized physical opening, as seen with more and less extension:


You can read more about this in my article on Extreme Macro Photography.

[Justinr]

Hi,

Thanks for chiming in. Checked out your article and liked it.

I have still not found out the issue OP has, but it seems he made progress and that is a good thing.

Best regards
Erik

I am indeed and it's all due to the kind help I am receiving here.

I have posted the latest attempt at the lichen below, the discs are approx 1mm in diameter.

This was taken with a stationary camera and fixed lens at f8 and under exposed by half a stop or so. The subject was moved through the focal plane using a jury rigged motion table of such crudity that I daren't mention its construction here.    Anyway, 21 slices were taken at 0.8mm apart approx and this was the result of the stacking.

Edit. I've also attached an original slice.

[Hank Keeton]

Hopefully this is not too off-point for the OP, but  Doug, I'd like to congratulate you on your very systematic presentation and discussion of "Extreme" macro work. Your illustrations are immensely informative.

I'd be tempted to open a dialogue with you off-line if you like, so we don't distract from the OP's direction. I use a Hassy-555ELD with Zeiss-135+bellows, and the Sinar P2 with  SK Makro-Symmar-HM-80, Roden Macro-Sironar-120, SK Makro-Symmar-HM-180 (a superb lens in all aspects...), and Roden Macro-Sironar-N-300 (which requires an unbelievable amount of supplemental light at full extension!).

Your inclusion of the Velmex moving-table is brilliant, and one I'll pursue myself. Also, your discussion of diffraction+extension is a basic steep learning curve for the effective use of bellows. I certainly agree with your summary of stitching errors using Helicon Focus, and I'm impressed by your "several hundred images" successfully stitched! I've found I can work Helicon either closer-to-farther or the reverse, with about the same results. You're quite right about maintaining the continuity of sequencing!

And, last but not least, I appreciate your addition of the non-moving shutter, something I have not attempted with my Aptus-65, but will explore, now that I've seen your results.

All in all, a very informative expose on "extreme" macro creativity! Thank you.

Cheers,

Hank

[ErikKaffehr]

Thanks, nice to hear!

Are you using bellows, extension tubes or something?

Best regards
Erik

I am indeed and it's all due to the kind help I am receiving here.

I have posted the latest attempt at the lichen below, the discs are approx 1mm in diameter.

This was taken with a stationary camera and fixed lens at f8 and under exposed by half a stop or so. The subject was moved through the focal plane using a jury rigged motion table of such crudity that I daren't mention its construction here.    Anyway, 21 slices were taken at 0.8mm apart approx and this was the result of the stacking.

Edit. I've also attached an original slice.

[EricWHiss]

Justinr,
I'm just browsing casual snap shots I have taken with my Rolleiflex AFi-ii 12 with the same focal length you are using.   Looking at those, I am really thinking that you should be able to do a lot better than you are without even stacking.   I still think you'd do better to back off the subject some and try for a single frame.  

Here's a hand held no mirror up or anything shot with the 120mm on my camera with some crops.  This is with f/11  and a camera mounted Metz 54 flash.  Literally no work, and no adjustments in post.  I can only imagine that were I to have taken a tripod and used mirror up and stacked a dozen frames what this would look like.

[EricWHiss]

Here's another shot this time with a Rollei 6008AF body and the Hasselblad 528c back and 90mm lens at that terribly diffraction ridden f-stop of f/16.  This time on a tripod.
First full frame, then a 100% crop which you have to click and enlarge to see fully.   

My point here is that you can get plenty of detail if you back off enough to fit the subject within the DOF in a single frame. I also still wonder if you couldn't be getting sharper images.

[Bart_van_der_Wolf]

I have posted the latest attempt at the lichen below, the discs are approx 1mm in diameter.

Edit. I've also attached an original slice.

Hi Justin,

Indeed, I think we are being painfully confronted with Diffraction limited imaging, at high magnification factors.

By calculating the actual magnification factor, you'll be able to more accurately estimate the actual DOF and diffraction limitations, and required focus bracketing intervals/quantities. Even f/8 might prove to be too narrow to allow low diffraction blur levels.

Cheers,
Bart

[Doug Peterson]

1. Lighting. I note that you shoot in the dark to avoid shutter 'shake' but do you feel that extraneous ambient light or slight over exposure will reduce the clarity of the image? I slightly underexposed the bike shots to test this (although it's in no way a true comparison) for the brighter areas of the lichen shots appeared the least sharp.

"Macro" photography covers a lot of technical range.

My article was specific to macro magnifications in the ball park of 10:1. At such high magnifications and with so many frames required for the stack, ANY vibration is severely detrimental. I had to turn off the A/C, take shallow breaths, use a sturdy table (and not actually touch the table at any point) and work some zen in the way I actuated the linear slide.

At lower magnifications the requirements to reduce vibration are not nearly as onerous.

2. You use a motion table to move the subject through the plane of focus, is this to avoid having to touch or disturb the camera or do you feel that there is some optical advantage over moving the plane of focus through the subject as I do by adjusting the lens?

Both methods work, but moving the subject IMO becomes preferable as you get into very high magnification. There are pros and cons both optically, mathematically, and in practicality. For instance it's a lot easier to move the weight of a fly's eye forward and backward in even increments on a linear slide vs. moving a view camera on a linear slide, especially if its' pointed down at an angle. It's also difficult to turn the focus wheel on a large-extension view camera 100 times, in even increments, without ever accidentally bumping the camera slightly left or right. Remember that at the magnification I was shooting I had maybe 5000 pixels going across a 2mm area of subject matter, which means each pixel covered less than a micron of subject matter. ANY bump in the camera left or right, even lightly tapping on the highly-levered-due-to-extension rear standard would chance registration by several pixels.

For what you're doing I'd focus more on the basics - high shutter speed, sharp aperture, solid tripod, repeated practice to find the various pitfalls.

[Doug Peterson]

Hopefully this is not too off-point for the OP, but  Doug, I'd like to congratulate you on your very systematic presentation and discussion of "Extreme" macro work. Your illustrations are immensely informative.

I'd be tempted to open a dialogue with you off-line if you like, so we don't distract from the OP's direction. I use a Hassy-555ELD with Zeiss-135+bellows, and the Sinar P2 with  SK Makro-Symmar-HM-80, Roden Macro-Sironar-120, SK Makro-Symmar-HM-180 (a superb lens in all aspects...), and Roden Macro-Sironar-N-300 (which requires an unbelievable amount of supplemental light at full extension!).

Your inclusion of the Velmex moving-table is brilliant, and one I'll pursue myself. Also, your discussion of diffraction+extension is a basic steep learning curve for the effective use of bellows. I certainly agree with your summary of stitching errors using Helicon Focus, and I'm impressed by your "several hundred images" successfully stitched! I've found I can work Helicon either closer-to-farther or the reverse, with about the same results. You're quite right about maintaining the continuity of sequencing!

And, last but not least, I appreciate your addition of the non-moving shutter, something I have not attempted with my Aptus-65, but will explore, now that I've seen your results.

All in all, a very informative expose on "extreme" macro creativity! Thank you.

Glad you enjoyed it.

It was purely an academic self-challenge. I get to play with some cool gear and work with some stunningly good photographers and interface with some of the top engineers in the photo biz - so it's nice when I have time to write up something like this.

Nearly everything in the article was snagged from google research I did along the way. I wish I would have kept better track of that research so I could have given proper credit to the various techniques and suggestions. You know how that is though – one link leads to another to another and soon you have no idea where you read what.

The only two things I felt I "discovered"* was the pre-stack stacking in Helicon (doing small batches before combining those batches) and the no-vibration shutter-by-flash which is fairly unique to digital backs.

*and it wouldn't surprise me at all if a dozen people had done both of these things first; I just didn't happen across it in my research

[EricWHiss]

I'm a big believer in keeping things as simple as possible.  I highly doubt that you'd need to be shooting in the dark or need to use even a tripod to get a sharp image of your lichen.  I shoot hand held macro all the time with my Rollei Hy6, and I don't use mirror up either because I need to focus on the subject.   

You may still decide to focus stack, but what's the good of doing all that work until you can consistently get single frames that are sharp at least in the slice of DOF.  I still don't see that in your images and have been scratching my head as to why.  I still wonder about your 120 lens.

Here's some more casual shots taken much closer in - hand held, no mirror lockup.  Note in these I used the Leaf Sensor flex square automatic crop. These are images of some lilies on my kitchen table.  I would guess the stamen were 2-3 mm across in width?  See the individual grains of pollen on the crop when expanded.   The shot with the 50mm componon lens shows the individual cells of the flower petal and a little pile of pollen grains.   When I walk around my kitchen the table shakes and the flowers move a bit. It's not a problem if you use a small flash as this freezes things quite effectively.

I'm not saying these images couldn't be sharper either.  Were I to have used mirror lockup, tripod, held my breath, etc. they could be sharper. What I'm saying is these casual shots of mine seem sharper than what you are getting with your tripod etc so I wonder if there is something else you could be doing?

[EricWHiss]

One more image and crop of the same bunch of flowers but here one of the flower had gotten old and dropped out of the vase.

[Justinr]

One more image and crop of the same bunch of flowers but here one of the flower had gotten old and dropped out of the vase.

My apologies for dropping off the radar and not getting back to you all but have been busy elsewhere over the last couple of days.

Justin.

[ErikKaffehr]

Hi,

That's OK, but it would be interesting to know what magnification you have, that is how large is your subject?

That would help us to calculate effective aperture and perhaps give a better understanding of your shooting situation. I agree with Eric Hiss that better detail could be expected, but it is pretty impossible to judge if all parameters are not know. What is your shooting distance, measured from sensor plane?

Best regards
Erik

My apologies for dropping off the radar and not getting back to you all but have been busy elsewhere over the last couple of days.

Justin.

[Bart_van_der_Wolf]

What is your shooting distance, measured from sensor plane?

Hi Erik,

That won't allow to accurately determine the magnification factor, except for infinity focus. The only reliable way is to measure the size of a known object size (e.g. tape measure or calipers) as it is rendered on the sensor (e.g. 'x' sensels with a sensel pitch of y microns).

Cheers,
Bart

[RobertJ]

I think you should back up and just take some normal shots, and evaluate the plane of focus, even if it's not where you actually focused.  I mean, it shouldn't be that hard to get sharp detail.

EricW, the dollar shots you posted show f/20 being sharper than f/16.  Are you seeing the same thing?  Is there any reason why it doesn't just keep going downhill after f/16?  Or maybe I'm seeing things wrong...

[EricWHiss]

I think you should back up and just take some normal shots, and evaluate the plane of focus, even if it's not where you actually focused.  I mean, it shouldn't be that hard to get sharp detail.

EricW, the dollar shots you posted show f/20 being sharper than f/16.  Are you seeing the same thing?  Is there any reason why it doesn't just keep going downhill after f/16?  Or maybe I'm seeing things wrong...

Good Catch!  For some reason the sharpening was turned up to 511 on that one file and I didn't notice it.  I'm sure when I did the test I was playing with this to see if I could tighten it up and forgot to reset it.

Here it is again with normal sharpening.  But still that was pretty cool that f/20 could be looking pretty sharp with that 5.2um 80mp back right?

[EricWHiss]

ps. T-1000
I agree with you and have been suggesting to back up and try to get one good frame with the whole subject within DOF.

Erik K - have a look at Justin's picture of his set up, his comment about the lichen disk's being approx 1 mm across, and also the picture with the ruler.  You'll get an approx idea of magnification. Not that it really matters, anyhow, since magnification differences at this level is mostly going to impact the size of the DOF and not how sharp the sharp part of the DOF is. 

In every shot there should be one slice that's sharp. I'd suggest working on getting that part sharp before worrying about stacking or other stuff.  Also Justin you might be having focus issues. I'm still wondering about the 120.   Another thing to look at is your lighting.  Small flash units work great for macro work since they can be placed quite close to the subject and are relatively large in scale to tiny subjects.   The most important part is the flash duration is quite fast compared to studio strobes. This really helps mitigate camera or other vibration.

I use a pocket wizard on the camera and one on the flash to get the flash close to the subject and off camera for more interesting lighting.

[ErikKaffehr]

Hi,

I don't know how the images he post are scaled and cropped. He posted a slice where DoF was pretty thin, and that may indicate large magnification.

I would also think that in a slice which contain a part that is in focus there would be optimum sharpness in some part.

Some other reflections, he perhaps could sharpen quite a bit more. Illumination can be flat and result in a soft image.

On reflection of my own tests, I feel that what is lost to diffraction can be regained in part, but fine low contrast structures will be lost, I rechecked the images you have posted, and checked out the f/8 and f/16 images. I took to the freedom to cut and paste from your postings. The f/8 image shows a paper structure while the f/16 image shows very little paper structure. I have just copied f/8 and f/16 image side by side. This is pretty much in line with what I would have expected from my own experience. I hope you don't mind that I use your images in this way, I feel it is proper use.

Best regards
Erik

ps. T-1000
I agree with you and have been suggesting to back up and try to get one good frame with the whole subject within DOF.

Erik K - have a look at Justin's picture of his set up, his comment about the lichen disk's being approx 1 mm across, and also the picture with the ruler.  You'll get an approx idea of magnification. Not that it really matters, anyhow, since magnification differences at this level is mostly going to impact the size of the DOF and not how sharp the sharp part of the DOF is.  

In every shot there should be one slice that's sharp. I'd suggest working on getting that part sharp before worrying about stacking or other stuff.  Also Justin you might be having focus issues. I'm still wondering about the 120.   Another thing to look at is your lighting.  Small flash units work great for macro work since they can be placed quite close to the subject and are relatively large in scale to tiny subjects.   The most important part is the flash duration is quite fast compared to studio strobes. This really helps mitigate camera or other vibration.

I use a pocket wizard on the camera and one on the flash to get the flash close to the subject and off camera for more interesting lighting.

[Bart_van_der_Wolf]

On reflection of my own tests, I feel that what is lost to diffraction can be regained in part, but fine low contrast structures will be lost, I rechecked the images you have posted, and checked out the f/8 and f/16 images. I took to the freedom to cut and paste from your postings. The f/8 image shows a paper structure while the f/16 image shows very little paper structure.

Hi Erik,

And that is part of the problem at hand, but I suspect there is more (maybe a lens issue, maybe lighting, maybe camera vibration, etc.). Diffraction will lower the high spatial frequency MTF response more than for low spatial frequencies. High spatial frequencies already have lower MTF response, partly due to the sampling with area aperture sensels. That combined response starts to visibly reduce micro-contrast when the aperture is still relatively wide.

In practice, I notice it for high contrast micro-detail when the diffraction pattern exceeds 1.5x the sensel pitch. For a 9 micron sensel pitch, that happens at f/10 and narrower (smaller sensel pitches at proportionally smaller apertures). Combined with the relatively flat lighting, that dictates apertures that are (significantly) wider than f/10, at small magnification factors(!).

When we increase the magnification factor by focusing closer than infinity, we also increase the apparent diffraction blur diameter. At 1:1 we already need f/5.0 or wider to not affect visible sharpness of high contrast micro-detail, low contrast micro-detail will suffer a lot more already.

Cheers,
Bart

P.S. Here is an example of how the Diffraction to Sensel pitch diameter (marked as "Dsd") ratio influences the visibility of various levels of (micro-)contrast:

Click on the image for actual 100% zoom size

Here is another (animated) representation of the DOF/resolution trade-off, starting at f/5.6 there is a visible reduction of the higher micro-contrast detail:

[EricWHiss]

Don't forget that while my $2 bill shots were taken with the mamiya and 120mm lens, the Aptus ii 12 back used has 5.2um sensor pitch, but Justin's back has larger. I have noted diffraction starts to be noticeable between f/8 and f/11 on the 80mp backs, however its not that noticeable and some lose can be 'mitigated' through software.

My lily pictures are clearly affected by diffraction but still usable - they look quite fine a normal print sizes even though I present a 100% pixel peeping crop.