rmpsrpms

The differences are noticeable even at the smaller image size, but any one image in isolation looks "OK". If you compare a larger image the differences are more obvious, but for publishing on the web at 800x800 pixels any of the lenses can do an OK job.

I did a shootout/comparison of 75mm lenses a while back (see here: http://www.coincommunity.com/forum/topic.asp?TOPIC_ID=98494&SearchTerms=75mm). I chose 75mm because it is just long enough to make work with most bellows on Dollar size coins, and short enough to keep the setups compact. It is my preferred size for use on the small microscope stand setups I favor. But some folks prefer larger setups, and some bellows have minimum extensions that are too long to work with Dollars. Both these situations require longer lenses.

I am an avid El-Cheapo enthusiast, and have watched eBay for deals on various lenses for a long while (see here: http://www.coincommunity.com/forum/topic.asp?TOPIC_ID=102384). I also have picked up other more expensive but interesting lenses along the way. Many are in the range of 100mm so I decided it would be good to compare them. Keeping in the 86mm-110mm range, I have 22 lenses to compare. I did not include the 95mm or 105mm Printing-Nikkors although I may add them later.

The lenses included in this shootout range from 86mm to 110mm; from Japan, Germany, France, and Russia; from enlarging, duplication, and macro applications; and from $15 to $250 on the used market. Many are seen regularly on eBay, while others come around only infrequently. The list includes, in order of increasing focal length:

86mm f5.6 Tomioka E36C duplication lens (fixed-aperture)
86mm f4 Tomioka E36 duplication lens
89mm f5 Rodenstock Scitex S-2 duplication lens
90mm f4.5 Perfex Anastigmat enlarging lens
90mm f4.5 Wollensak Enlarging Raptar enlarging lens
90mm f4.5 Elgeet Colorstigmat enlarging lens
90mm f4 Tomioka E36 duplication lens
90mm f6.3 Perfex Anastigmat enlarging lens
90mm f4.5 Roussel Trylor enlarging lens
90mm f4.5 Lens Made In Japan (no brand) enlarging lens
94.1mm f4 Tomioka E66 duplication lens
100mm f4.5 Lomo Mikroplanar macro lens
100mm f4.5 Kodak Enlarging Ektar enlarging lens
101mm f4.5 Wollensak Enlarging Raptar enlarging lens
105mm f4.5 Vivitar enlarging lens
105mm f4.5 Tomioka Tominon macro lens
105mm f5.6 Rodenstock Rodagon enlarging lens
105mm f5.6 Nikon EL-Nikkor enlarging lens
108mm f5.6 Rodenstock LFOV duplication lens (fixed aperture)
110mm f4 Industar N-100-Y enlarging lens
110mm f5.6 Rodenstock Scitex S-2 duplication lens
110mm f5 Rodenstock Scitex S-3 duplication lens (fixed aperture)

I used a 1954-S Lincoln Cent as subject for the shootout. I took overall photos using each lens, with white balance set for Tungsten, so that the color presentation differences of the lenses can be seen. No sharpening was done in-camera or in post-processing. Levels were adjusted equally on all images.

Lighting was kept constant for all shots. Of course the lighting will appear different for different focal lengths due to perspective changes, but only slightly within this narrow range.

Aperture was held constant at f5.6, or at either the minimum aperture (eg f6.3 for Perfex) or at the fixed aperture of the lens (eg f5.0 on Scitex). This is a wider aperture than I used for the 75mm shootout, where I chose f8. The wider aperture exposes more issues for each lens, and allows a better comparison. The f8 in 75mm shootout somewhat equalized the lens performance, while f5.6 will show more of the differences. In some cases f5.6 is the maximum aperture, while in others it is stopped-down by 1 stop.

Critical focus was performed on the middle portion of Lincoln's neck. This is approx in the middle of the topography of a Lincoln Cent, so focusing here gives best balance of focus between fields and tops of the devices.

I cropped the center and bottom edge of each shot. The center is over the Neck area where critical focus was done. Bottom edge shows two effects: lens coverage; and field flatness. I only considered out to the edge of the coin, since that's how these lenses would be used, so "corner" sharpness is not really relevant for coin photography.

I used my HRT2i, APS-C, 18MP camera for these tests. A full-frame camera would be more revealing for edge sharpness issues, but most coin photographers will likely use APS-C.

Now the catch...you need to visit my website to see the images and results! Take a look at:

http://www.macrocoins.com/100mm-lens-shootout.html

So you're not actually changing the angle of the object (coin) or the subject (camera) - the program is extrapolating from the varying depths of focus what the different angles would look like.

That's correct. All done with software. It would be extremely tough to get this level of clarity with a tilted coin. The stack depths would be excessive and the result would likely not be very good.

That's amazing. How do you compose the 3D pics?

The process is the same as focus stacking, ie taking a series of photos at different focus planes and then using software to create a composite image using only the in-focus parts. There are several software packages that will do this such as Zerene Stacker, Helicon Focus, even Photoshop. The software creates a "depth map" as it moves through the stack of images, so it knows the relative depth of the in-focus parts it includes in the final image, and can then use this depth map to create variable-angle perspective representations of the image.

Here's a 1949-S Lincoln Cent DDO#3 presented with a couple of techniques.

Unprocessed image using 5x magnification, 25-stack of images:




Unprocessed image using 10x magnification, 36-stack of images:




3D processed 10x image but looking straight on to compare with unprocessed:




3D processed 10x image looking at 45-deg from 9:00:




3D processed 10x image looking at 45-deg from 2:00




I think this technique is useful to help visualize these kind of variety details in a similar way to viewing then with a microscope, at least the way I view them. I'm not so sure the 3D versions of these is all that useful, but some may like them. I find them kind of gimmicky, and while they can be sort of cool they don't offer any more information than the 2-D perspective views above. If you disagree, let me know. It's not too hard to include 3D versions of these.

Ray

Duplicate post

I haven't posted here for a while so I thought I'd post something a bit different...

I've been searching for a long time for a nice quality example of this famous Nikon 4x objective. As Mark Goodman states in his review, "this lens is the standard by which all others are compared in this magnification range".

You can read Mark's review at:

http://coinimaging.com/nikon_4apo.html

As a test of this objective I took a picture of a 1955-S RPM#1 from a group of 7 I found in a recent roll search. The objective performed every bit as well as expected. Here is the image:



Note that this image is better, but not tremendously better than what can be achieved with a 0.13NA or even 0.10NA objective. Its higher NA (0.2) makes it sharper, but it also means more shots are required for the stack. The image above required 31 shots!! A 0.1NA objective would require half or fewer shots, but each one would be a little less sharp due to the smaller aperture causing diffraction.

I've had good luck with microscope slides for axial lighting of cents. I take a small piece of rigid black foam and cut a 45-deg angle in it, then tape the slide to the foam. You need to make sure you put a non-reflective black surface behind the reflector to avoid having it re-reflect up to the sensor...Ray

I know what you mean about the variability using different viewers. Some of my images are stacked using CombineZP. The source images look better in CZP than in any other viewer I have used. No idea why. All my editing is done with Nikon ViewNX/NX2 and it seems to be fairly balanced in presentation. I usually check the images using Firefox as the viewer before posting to the web, figuring that many people use it or some other similar web browser that all have similar jpg viewing engines.

By the way, the ISO for the above images was 100, so that's also not a factor.

I am actually learning to like those "sparklies"...knowing where they come from and why helps a lot and they add to the local contrast and luster presentation.

This image was slightly sharpened ("1") after downsizing it to 800x800. In my case the "specklies" are due to the LED lights being very small point sources. They go away if you diffuse the lights a bit...Ray

PS I just looked back and realized I posted this coin twice...sorry

How about this one...Ray

Here's a toned Lincoln Wheat Cent. This is using a Tominon E36 86mm f/4 lens instead of the Apo-Rodagon-D. Not as crisp as the ARD but came out pretty nice...Ray

Bravo. In addition to color and luster, you've also captured the fine detail of the coin. An exceptional photograph. Do you care to expand on the setup for this image (types of lights, light position, f-stop, exposure, etc).

Thanks very much! I would be glad to share the description of the setup:

Rodenstock Apo Rodagon D 75mm f4 M=1 set at f5.6

Modified Microscope Stand

Pentax Auto Bellows

Nikon D7000 in Live View, Aperture Priority Mode (not sure the shutter speed, but faster than 1/100sec)

Two Jansjo LED lights at 10:30 and 1:30, 75mm above coin surface, approx 10deg off vertical, no diffusion

...Ray

Here's a Nickel (actually, 5 Centavos) showing good color and luster simultaneously:

I shoot mostly with axial lighting, but I do vary the light angle to achieve the effect that I want. I think that is true with any light setup. You find the basic setup, then you start experimenting depending on the coin and desired effect.

Please explain, because I think of axial lighting as having only one possible angle, 0-degrees from the viewing axis, ie straight down from the lens. Do you mean that you vary the angle of the coin to change how the light reflects off the coin surfaces? Or do you have an axial setup that allows you to adjust the light angle in some way? I generally shoot fairly wide open so am always conscious of DOF and try to shoot with the coin as flat as possible. Any tilt causes focus issues with the lens wide open. I don't want to have to stack images to get a good coin pic, and most modern coins have low enough relief that as long as the coin is flat I can get acceptable DOF at the sharpest aperture for the lens.

Excellent photo and great choice of background.

 

Not quite as good but another example of bronze/copper on a black background:

Actually your photo is very nice, and very sharp. What camera/lens do you use?

Nice effect. Have you explored axial lighting?

I've tried axial lighting in the past but it doesn't give a good representation of a real-world look from my experience. It's very pinpoint in nature, and thus creates too high contrast and direct reflection back to the sensor. Do you have any different experience with it? ...Ray

It's hard to get both color and luster to show up in a single shot. Color requires high angles and diffusion, while luster requires medium angles and pinpoint sources. Here is my latest attempt using a new lighting setup with a medium-lustrous, highly-toned, raw 53-S Lincoln. The setup uses a horizontal aluminum reflector just above the lens, with 2 LED lights shining UP to the reflector, which then both reflects and diffuses the light going to the coin. Disadvantage is that there is less light getting to the coin, so exposures are longer. But the results so far look pretty good, with reasonably even illumination, color definition, enough luster but not too much, and decent resolution of surface details.