'D' is for dime...or maybe for Denver (10X objective)

[rjlittlefield]

"D" is for dime...or maybe for Denver.

In this case it's sort of both, because this particular "D" is the mark of the Denver mint, right above the "2006" on the dime.

I confess, this posting exists only to illustrate what kind of a coin picture you can get using a microscope objective on bellows. A fellow over in the Yahoo Microscope group was asking about that, and I got interested in running a test.

Somehow the picture didn't seem worthy of a gallery posting, but for what it's worth, here's the info.

Technical: Canon 300D camera, sensor size 22.7 x 15.1 mm. aus Jena 10X NA 0.25 achromat on bellows with effective 225 mm optical tube length giving 15X onto the sensor. Stacked with Helicon Focus, focus step 0.00025", using this setup. Dual fiber halogen illuminator, pingpong ball diffuser. The coin was tipped very slightly -- when it was exactly perpendicular to the optical axis, the flat surface went black because it reflected mostly the interior of the camera. Noticeable chromatic aberration removed with PTLens (Red-Cyan = -0.0029, Blue-Yellow = +0.0023).

Coinage: U.S. dime, described here. "Mint mark" described here.

--Rik
Edit: added note about PTLens.

[rjlittlefield]

As followup response to more questions, I shot the same dime with a Zeiss Luminar 16mm f/2.5 macro lens, set wide open for highest resolution.

Compared to the microscope objective, the Luminar is substantially harder to illuminate around because it has a larger diameter, a squared-off end (not tapered), and it's all black. I addressed the "black" problem by wrapping the lens in aluminum foil with a hole punched in it to look through. Diameter and shape I couldn't do anything about, so I just cut a new pingpong ball to fit snugly, then positioned the fiber heads to get the best appearance I could in a few minutes of messing around. The lighting is still pretty lame and doesn't match the previous image, but hopefully it's good enough to allow some comparisons.

Quickly summarizing... The Luminar's resolution is more uniform corner-to-corner, appearing to be a bit higher in the corners and a bit lower in the center. (This is consistent with results shown in earlier tests.) The Luminar has no detectable chromatic aberration. The Luminar's field is also very flat, versus significant curvature for the achromat objective (according to earlier tests, never posted). However, curvature of field does not matter here because of the stacking.

Bottom line -- to my eye, the Luminar image quality is not markedly superior to the microscope objective for this application. Considering the illumination issues, I think I'll be using the objective for this sort of thing.

I feel compelled again to comment about the need for stacking. This little "D" has vertical relief of only about 0.004" -- around 100 microns. That may not sound like much, but it's 12 times the 8.5 micron DOF quoted by Microscopyu for a 10X NA 0.25 objective.

More questions, observations, comments?

--Rik

[Mike B in OKlahoma]

I like to brag about doing high-magnification macro on my ants and bugs, but this puts me to shame! I've looked at a lot of dimes, and even checked out mint marks a lot in my younger days, but this definitely meets your definition of macro as showing more detail than can be seen with the naked eye!

Should have gone with the galleries for this one.

(incidentally, I agree that the microscope objective seems like the way to go here).

[Danny]

Try doing this on moving insects, its fun I tell ya

Very interesting Rik. Fantastic what the stacking can do and clean lighting.

Decided to have a look at the ratios. This is with the FZ10 and an added Nikon 35-80 set at the 35mm end. Its a copper coin and was only done with hand holding, no fancy lighting. Just an exercise for the ratio.



FZ10 @ 1296mm, 1/20th, F/8, ISO 50. Full frame.

So thats where the "D" in our copper coin joins, the horizontal bar to the curve of the "D". Yeah I know, I should set up a decent light tent and tripod to get it at a full 90 degrees But just a very quick play.

All the best and nice work Rik. I thought you might like the ratio in this, because you like the weird stuff.

Danny.

[rjlittlefield]

What does "1296mm" means? Is that 35mm equivalent, with the 3X digital zoom kicked in?

Anyway, here's another way of looking at what I think you've done. Your FZ-10 has a sensor that's 4.29 mm high x 5.76 mm wide. The lens zooms to 72 mm actual. Stick a 35 mm auxiliary on the front of that, and you have 72/35 = 2.06X lens magnification. Then that gives 5.76 mm sensor width / 2.06 = 2.8 mm field width at the subject. Kick in the 3X digital zoom (or do the same crop in Photoshop), and you're down to 0.93 mm subject width. Yep, teeny tiny, and an easy way to get it!

--Rik

[Danny]

What does "1296mm" means? Is that 35mm equivalent, with the 3X digital zoom kicked in?

--Rik

Yep, thats the cheats way of saying it Rik. I just look at the EXIF figures. So yes, thats using th 3x digital zoom. I need that to avoid the vignetting that takes place. If we use a 24mm Canon instead of the 35mm, well you can see where thats going. Darn hard on moving subjects is all I know, but its fun and challenging.

The algorithm's used on the FZ10 and the Sony CD-1000, seem to be better than cropping in software. But, the digital zoom HAS to be used on camera because of the vignetting. Thats something we can't avoid and fix too easily in software with cropping.

All the best Rick, I'll sit down one day and go through all the lenses at max. Setup a decent light tent and give it a blast to see what comes out.

Danny.