scales as a test-target

[Adalbert]

Hello everybody,
Does anybody know what can be used for the testing of the old lenses?
I’m looking for any idea :-) I have already tried with the following subjects:
http://www.photomacrography.net/forum/v ... highlight=
How can I recognize the flaws of the lens on the picture?

https://farm5.staticflickr.com/4807/444 ... 2c0b_o.jpg
BR, ADi

[Lou Jost]

That particular species is one of the hardest to judge. Lighting is often a bigger factor than the lens because those scales are so reflective.

A less reflective species has lots of levels of details that you can inspect for CA and resolution. But you have to look much more closely than we can look in a web-sized image.

I especially like to include the white scales along the edges of many butterfly and moth wings (including the Sunset Moth you used). These scales form dense overlapping layers that will reveal the effects of LoCA on a stack. These scales provide a better test of the effect of LoCA on a stack than flat targets.

Also, let your wings get dusty. Flecks of dust reveal astigmatism.

[Chris S.]

Agreed that the metallic-looking scales of the sunset moth are hard to use for a lens test. But there are quite different scales on other portions of this moth that I’ve found to be very good for testing of objectives in the 50-100x range. These scales are found in the very black regions of the moth’s wing, and are not reflective--on the contrary, they are practically black holes when it comes to lighting them. These black scales are longer than the metallic ones, and have regularly-spaced holes in them that vary in size, but range around the resolution limits of these objectives. Importantly, these black scales will lie flat if deposited on a microscope slide, whereas the metallic scales bend at right angles and stick up off the slide.

To use these black scales for lens testing, I’d recommend scraping a few off the wing into some alcohol. Swirl the alcohol to get the scales into suspension, then use an eyedropper to deposit alcohol and scales onto the slide, and dry it down. Van der Waal’s forces (I presume) hold the scales pretty solidly—I’ve been using the same tests slides for years. For lens testing, I backlight the slide by placing a white card several inches behind it, and lighting that card evenly.

An example of such a test is in this thread, starting with the third image down.

For lower magnifications, and for initial tests at higher magnifications, I use laser-printed test targets glued to microscope slides.

This said, I’d recommend extreme wariness about judging your lens on any target without an identical comparison stack made with a known-good objective under the exact same conditions. There are way too many variables that affect the image. I’ve tested a lot of lenses, and feel confident testing/judging an optic only if the test optic is in my hands, and I test it against a known-good optic also in my hands, and if the only thing that differs between the two image stacks is the optic. This done, the only method of comparison I really trust is to lay the test stacks on top of one another in Photoshop layers, and flash between them.

--Chris S.

[Lou Jost]

The scale strew that Chris suggests is a very good idea. I use it too, and I especially like to backlight these so that, when comparing lenses, the light stays exactly the same. Lenses that are different distances from the subject can partly shade light sources that are placed above the subject. No such problems with backlighting.

However I still find three-dimensionality is a useful feature of a test subject. After all, that's what we are going to be photographing in real life. The effect of LoCA on a stack can be tested very well by challenging the lens with the thick layer of very finely detailed white edge scales.

And again, let the dust fall on it!! These are free astigmatism generators.

Of course these suggestions are only useful if you can take the time to make a short or partial stack. If you have many lenses to test, it may only be practical to take single shots. In that case, as Robert has clearly shown, wafers are an excellent target.

[Macrero]

Butterfly/Moth scales are good lens testing target. Though as already mentioned, those particular scales are not the best chioce.

Like Lou, I too like to make some of the tests with "3D" subjects, usually insects, since that's what I shoot in real life.

About the "recognizing the flaws" part, as I said in your other thread about lens testing, in order to be 100% sure the lens/objective performs flawless, you have to know how that particular lens/objective should do.

If you don't know the lens, and it does well in your tests and you are happy with the result, I would say don't obsess with perfection and just use it to take some nice pictures.

If you have paid a lot of money for an expensive lens you haven't owned/used before, and you want to be sure it performs flawless, the only way is to make a proper comparison with an actually flawless unit.

Best,

- Macrero

[Lou Jost]

I should add that wafers have some other advantages, especially their repetitive patterns that give you a very accurate way to judge the differences between center, edge, and corner. Robert's tests are examples of how well this works. It is much harder to make such close comparisons using natural objects that are not exact duplicates of each other.

[Chris S.]

We might do well to subdivide the concept of "lens testing" into a couple of different activities. One such activity is characterizing new-to-our-community optics to see how they perform in the photography of small things; this is a broad question, and answering it involves evaluating quite a few characteristics of the lens, including longitudinal aberration. Another--quite different--activity is checking an unknown specimen of a well-understood optic, to see if it is a "dud" that should be returned, or a "keeper." This is a much narrower question, and much more easily answered with more limited testing.

The three-dimensional test subjects that Lou and Macrero advocate are no doubt useful for characterizing unknown optics, such as Lou, Robert, and some others have done so much of. I've done rather little of this, but have tested quite a few Mitutoyo objectives to separate the good from the bad. In my experience, the differences between good and bad Mitties mainly involve resolution and contrast, though in particularly bad specimens, lateral CA sometimes also jumps out. For this particular, limited task, I think a black and white laser print, and (with higher magnifications) a wing scale with interstices ranging around the expected resolution minimum, are sufficient to do the job. This task, I think, is what Adelbert is trying to perform with his new Mitty 20x.

Adi, can you shoot identical stacks with your Mitutoyo 20x, and your Nikon 20x LU, and layer full-size images in Photoshop for comparison? If you upload such a file somewhere and provide a download link, we can better address your question about, "Are there any flaws in the image from my lens?" (Though all we can do is compare it with your Nikon 20x, of course.)

--Chris S.