80mm lens shootout - for coin photography

Have questions about the equipment used for macro- or micro- photography? Post those questions in this forum.

Moderators: rjlittlefield, ChrisR, Chris S., Pau

karlmera
Posts: 16
Joined: Sun Feb 05, 2012 12:37 pm

Post by karlmera »

austrokiwi1 wrote:The OLy 80mm F4 bellows lens is optimised for 1-1 . Its not a great lens as it suffers badly from Lateral CA.

As far as Coins go 1-1 is applicable to only small diameter coins. For Thaler/Crown/Silver-Dollar sized coins on a full frame sensor about 0.6X is used.
Yes, the Olympus and the macro-symmar, and others?

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

Post by ChrisR »

I'm sure Ray doesn't need anyone to explain for him, but he tests a lot of lenses. :) This set just happens to be 80mm.
I thought there was one specifically at 1:1 , but this one is close if your interest is there :
0.3x ...1.2x
Chris R

ray_parkhurst
Posts: 3412
Joined: Sat Nov 20, 2010 10:40 am
Location: Santa Clara, CA, USA
Contact:

Post by ray_parkhurst »

karlmera wrote:
austrokiwi1 wrote:The OLy 80mm F4 bellows lens is optimised for 1-1 . Its not a great lens as it suffers badly from Lateral CA.

As far as Coins go 1-1 is applicable to only small diameter coins. For Thaler/Crown/Silver-Dollar sized coins on a full frame sensor about 0.6X is used.
Yes, the Olympus and the macro-symmar, and others?
I believe the 80mm Leica bellows lenses (Milar, Summar, and Photar) also include 1:1 in their ranges. I don't have adequate documentation for any of these so am not 100% sure.

edited to add:

This 1:1 issue is actually something that makes APS-C sensors seem more suitable for coin photography. To fill the FF sensor (actually I usually do 95% to leave a little space for cropping) with a US Dime (18mm) or worst case a US Trime (Silver 3c Piece, 14mm) you need to go beyond 1:1. This means that FF sensors cannot make use of dedicated macro lenses to fill the 24mm vertical height of the sensor with small coins, since to do this requires going well beyond 1:1 (~1.3:1 for Dimes, or up to ~1.6:1 for Trimes). Largest US coins are ~40mm, so down to ~0.5:1 magnification. This bodes well for those duplication lenses which are optimized around 1:1 such as the 75ARD1, or the 105PN. It's also a perfect range for the 85MV or the 80MS without requiring the lens to be removed and reversed like most enlarging lenses would.

On the other hand, corresponding magnifications for APS-C are ~0.8:1 for Dimes, or worst case ~1:1 for Trimes. Cents are at 0.72:1 (where this shootout was shot). Largest coins (ASE $1) are at ~0.35:1, inside the optimum range for most enlarging lenses. So for APS-C, enlarging lenses can work well without having to swap as different size coins are shot.

austrokiwi1
Posts: 350
Joined: Sun Sep 14, 2014 10:53 am

Post by austrokiwi1 »

This 1:1 issue is actually something that makes APS-C sensors seem more suitable for coin photography.
I think this comment shows a clear coin collectors bias to one particular coin series. In regards to coins it can not be taken as a general across the board comment. I believe it focuses on the specialist collecting area of the writer whose focus appears to be US 1 cent ( small variety). I repeat my self 1-1 is usually only used with small coins. RMPRMPS seems to focus on one of the smallest diametre coins ( US 1 cent) with his photography. For such coins 1-1 works. However in America the backbone of the collecting market is the MorganDollar ( aprox 39mm in Diametre). In Europe it is often Thaler sized coins( 41mm) My coin collecting focus runs predominantly to Thalers but I have coins as small as 2-3 mm in Diameter in my collection. I therefore find the "Sweeping" comment That APS-C sensors are more suitable for coin photography is incorrect. I have two cameras that I use for coin photography a FF and a MFT. IF anything the squarer shape of the MFT Sensor is better suited to imaging Coins.

For me the advantages of a MFT sensor are that a lower magnification is required; often not more than 0.5X times and sometimes as low as 0.1X is needed. The MFT sensor low mag requirements sees a much greater DOF, and stacking is very rarely required( I have never used stacking with coins and my MFT camera). The trade off is greater noise levels and for me some times that recommends strongly against MFT.

Its been a long time since I have used a dedicated APSc camera so I am not going to comment on that format. As far as full frame goes, it is my preferred format for coins. I research and write articles ( my last was in the July 2017 Numismatist "Bubble and Burst" page 65. Note the print quality degraded the coin image quality).

For me the advantages of FF are the resultant print size. Many coin collectors only deal with digital images therefore FF is mostly an overkill as FF images usually exceed the posting limits but a large margin. I even note that RMPSRMPS reduces his FF images for posting, effectively saying the images look better( a Comment I have never got my head around, so I probably don't understand it) For me the greater Dynamic range and better overall resolution result in much better prints, than, based on my skills, I can achieve with APSC and MFT. That said the issues with FF are online posting. Often I have to compress my FF images so much there are very poor facsimiles of the original image.

MY rant may be a bit over the top and perhaps may seem to lack direction but I take serious issue when any one says "this is the best Camera/sensor size for coins( or any subject matter) . For me such narrow recommendations suggest the person stating it is assuming there is only one subject type or only one correct way to take images. I can't comment in other areas but with coins it really depends on what coin you are photgraphing and the purpose you have for the image. If it is a small coin ( US 1 Cent)then APSC with a PN 95mm is going to perform very well. If its a Silver Dollar sized coin a FF sensor with a PN 105mm is going to be one of the best options. It comes down to knowing your equipment and matching lens and Camera( if you are so lucky tio have a choice) to your photographic needs. So lets make it clear 1-1 magnification and APSC is not necessarily the best coin photography option!!!.

For me Karlmera's question about which of the lenses are 1-1 resonantes. For me I rarely use 1-1 magnification. I suspect insisting on just a 1-1 magnification has handicapped some lenses in the test and hidden their true potential. For example in my experience the MInolta CE 80mm produces images of similar quality to the Agfa repromaster( but at different magnifications. I believe the Repromaster is optimized for 1-1 through 2-1 MAg range)
Still learning,
Cameras' Sony A7rII, OLympus OMD-EM10II
Macro lenses: Printing nikkor 105mm, Sony FE 90mm F2.8 Macro G, Schneider Kreuznach Makro Iris 50mm , 2.8, Schnieder Kreuznach APO Componon HM 40mm F2.8 , Mamiya 645 120mm F4 Macro ( used with mirex tilt shift adapter), Olympus 135mm 4.5 bellows lens, Oly 80mm bellows lens, Olympus 60mm F2.8

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

Post by rjlittlefield »

austrokiwi1 wrote:The MFT sensor low mag requirements sees a much greater DOF
Sorry, but I have to correct the technical point about DOF.

Using a smaller sensor in fact does not increase DOF, except indirectly by influencing how wide of an entrance cone people set their lenses for.

Yes, if you frame the same subject and set the "same" f/8 on large and small sensors, you'll get more DOF on the small sensor.

But that's because the smaller sensor demands a shorter lens, which at f/8 will have a smaller entrance pupil. It's the smaller entrance pupil that produces the greater DOF, and in exchange you also get proportionally greater diffraction blur with respect to the subject.

If instead you stop down the lens on the larger camera to give the same DOF, then it turns out that you also get the same diffraction blur.

It's the angles that count. With the effective f-numbers scaled to match the sensor size inside the camera, then outside the camera the entrance cones will be the same width, and that's the condition that causes same DOF and same diffraction blur.

Mathematically, this corresponds to using effective f-numbers that are directly proportional to the sensor size, for example effective f/16 on a 24mm sensor height versus f/8 on 12mm sensor height.

Of course this simply reinforces your point that no one sensor size is universally better than another, even for the same size subject.

--Rik

ray_parkhurst
Posts: 3412
Joined: Sat Nov 20, 2010 10:40 am
Location: Santa Clara, CA, USA
Contact:

Post by ray_parkhurst »

austrokiwi1 wrote: I therefore find the "Sweeping" comment That APS-C sensors are more suitable for coin photography is incorrect.
My conclusion that APS-C is more suitable for coins was not based on just a single (small) coin size. Perhaps you could have avoided this (33 sentence) rant if you had simply read my (9 sentence) post edit.
austrokiwi1 wrote: For me the advantages of a MFT sensor are that a lower magnification is required
Hopefully you can see that this statement is the height of hypocrisy. You are stating the advantage of MFT over FF for the exact same reason I stated the advantage of APS-C over FF: lower required magnification.

Another factor making APS-C (or MFT) more suitable than FF that I did not state was that lens coverage is less critical. Many of the lenses in my test did not even cover APS-C well, so going to FF would have been futile. I had held-off doing this shootout for nearly 2 years while I waited for a FF camera that met my earlier stated criteria. Interestingly, it was RobertOToole's publishing of his 4x Lens Test series that finally convinced me to give up. He also used APS-C for his series since it is so difficult to find lenses that cover FF sensor at 4x mag. I suspect both his test and my shootout would have consisted of very few lenses had we used FF.

edited to add:
austrokiwi1 wrote:I even note that RMPSRMPS reduces his FF images for posting, effectively saying the images look better( a Comment I have never got my head around, so I probably don't understand it)
Hmm, I don't know where you came up with this factoid. Other than the fact I don't use FF, I reduce my images for posting because the 5184x3456 (18MP) images would be far too large to post on most forums, not because they "look better".

ray_parkhurst
Posts: 3412
Joined: Sat Nov 20, 2010 10:40 am
Location: Santa Clara, CA, USA
Contact:

Post by ray_parkhurst »

rjlittlefield wrote:
austrokiwi1 wrote:The MFT sensor low mag requirements sees a much greater DOF
Sorry, but I have to correct the technical point about DOF.

...

--Rik
Rik...I think what AK is saying is simply that if you shoot to fill both sensors with the same coin, the magnification for MFT is much lower than for FF, so DOF is larger on MFT than for FF due to the lower magnification.

edited to add:
From your stacker DOF table, if shooting at f/8, a 40mm coin on MFT has a DOF of ~2mm, while on FF it is ~1mm.

karlmera
Posts: 16
Joined: Sun Feb 05, 2012 12:37 pm

Post by karlmera »

ray_parkhurst wrote: if shooting at f/8, a 40mm coin on MFT has a DOF of ~2mm, while on FF it is ~1mm.
Is the diffraction the same?

ray_parkhurst
Posts: 3412
Joined: Sat Nov 20, 2010 10:40 am
Location: Santa Clara, CA, USA
Contact:

Post by ray_parkhurst »

karlmera wrote:
ray_parkhurst wrote: if shooting at f/8, a 40mm coin on MFT has a DOF of ~2mm, while on FF it is ~1mm.
Is the diffraction the same?
No.

Feff = Finf * (m+1)

For MFT case, m=0.35

for Finf = f8, Feff = 10.8

For FF case, m=0.6

for Finf = f8, Feff = 12.8

So for MFT, you get a larger effective aperture as well as more DOF than you do for FF when shooting to frame the same subject with same aperture. APS-C is between the two, closer to MFT.

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

Post by ChrisR »

I was expecting it to be "a wash"
but running some numbers it doesn't look that way. Could always be the wrong numbers.
Taking here a 12mm wide subject, so M is 2 for DX and 3 for FX

Image

It looks like the "M + 1" effect. If you take M up to 20, the difference in DOF is insignificant
Last edited by ChrisR on Sun Feb 18, 2018 8:42 am, edited 1 time in total.
Chris R

ray_parkhurst
Posts: 3412
Joined: Sat Nov 20, 2010 10:40 am
Location: Santa Clara, CA, USA
Contact:

Post by ray_parkhurst »

Yep, numbers are less dramatic at 2:1 vs 3:1 but still make a difference.

Why did you use larger pixels on the FF vs APS-C? Cameras today seem to have small pixels for FF as well, with large pixel count. Mfrs seem to be converging around 4um at the current technology nodes so even the FF should have 4um pixels IMO.

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

Post by ChrisR »

I use 24MP for both sizes of sensor, to compare like with like. I think that's appropriate for the comparison.

If you're looking at DOF then you need to know the CofC, which you can't change at the same time in a comparison.

If you have more pixels in the sensor, then the CofC becomes a smaller percentage of the sensor dimension.

If you're displaying the picture o the same size screen, then again, % of picture width is the useful one.
Granted if you have a more pixels then you'd choose different numbers.

Note the reversal of magnitude of lines 20 & 21, there's a crossover in which s more significant.

If you assume a smaller, "one inch" sensor at 18mm wide, compared with FX, sure you get more DOF with the smaller sensor, but you've also hit the diffraction limit: With a 36mm subject -


Image

So, why is the aperture significant?? Who cares, you can change it, but... tbc...
Last edited by ChrisR on Sun Feb 18, 2018 9:23 am, edited 1 time in total.
Chris R

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

Post by rjlittlefield »

ray_parkhurst wrote:From your stacker DOF table, if shooting at f/8, a 40mm coin on MFT has a DOF of ~2mm, while on FF it is ~1mm.
ray_parkhurst wrote:
karlmera wrote:Is the diffraction the same?
No.

Feff = Finf * (m+1)

For MFT case, m=0.35

for Finf = f8, Feff = 10.8

For FF case, m=0.6

for Finf = f8, Feff = 12.8

So for MFT, you get a larger effective aperture as well as more DOF than you do for FF when shooting to frame the same subject with same aperture. APS-C is between the two, closer to MFT.
With constant Finf, you do get larger Feff on the smaller sensor, but not enough larger to make up for the change in sensor size. Feff=10.8 has a lot more diffraction blur relative to the MFT sensor than Feff=12.8 does to the FF sensor. As a result, when you make same size prints, the MFT sensor gives a more blurred image to go along with the greater DOF.

Again, the requirement for same diffraction, same DOF, is that Feff must scale in proportion to sensor size.

In the Stacker DOF table (at https://zerenesystems.com/cms/stacker/d ... romicrodof), the simplest place to see the proper relationship is in table 2-B, with Feff(FF) = 2*Feff(MFT) so that the effective f-numbers are proportional to the sensor size. For example, MFT would be m=0.35 and Feff=8, while FF would be m=0.7 and Feff=16, both of which have DOF=1.1 mm.

--Rik

karlmera
Posts: 16
Joined: Sun Feb 05, 2012 12:37 pm

Post by karlmera »

rjlittlefield wrote: Feff=10.8 has a lot more diffraction blur relative to the MFT sensor than Feff=12.8 does to the FF sensor. As a result, when you make same size prints, the MFT sensor gives a more blurred image to go along with the greater DOF.
But if we want the same sharpness, we should take the relative aperture, p.e.

eff. 8 for MFT and eff. 16 for FF? And the dof is the same, isn't it?

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

Post by ChrisR »

....continued,

lo and behold, if you alter the aperture so they're both equally diffraction limited, then you get this - check the effective aperture.
Image

That raises a bunch of the issues which we've considered before: does your lens work at the apertures you have to set it at? Does diffraction blur look like OOF blur, and how do they add up - or stack up? If you're having to use a lens at its maximum aperture, how does aberration blur add in, or even, is lens performance overwhelming anyway? Etc!


Credit - it was originally Rik's spreadsheet, of course.
Chris R

Post Reply Previous topicNext topic