UV Lens Test: What should be included?

[Andrea B.]

Most UV lens tests just show the lens works for UV.

That is not enough!

There are other factors which need to be examined: chromatic abberration, focus shift, corner vs. center sharpness, diffraction, ergonomics, hotspots, flare etc.

 

If I could put together a nice suggested Test Plan for this, it could be useful. I am an impatient tester myself, so it would certainly help keep me on track should I plan to test any of my current bin full of 35/3.5s. "-)

 

Test for Hotspots: How best to do this? Shoot an aperture series, of course. But of what? The sky? A wall? Monochrome or colour?

 

Test for diffraction set-in: Shoot an aperture series. Of what? Turn off all sharpening in-camera and in-editor?

Can Hotspot test aperture series be combined with the Diffraction aperture series? Should the aperture series be made in monochrome or colour?

 

Test for Chromatic Abberration: How best to do this to induce it if the lens is prone to it?

 

Test for UV-capability: Needs must be informal unless one owns a mono-chrono-thingie. Need a standardized test scene. Include what? Use UV-flash? Use "strong sunlight"? Landscape versus close-up useful?

 

Test for sharpness in corners/center: Shoot a brick wall. Monochrome or colour?

 

Added 05 Sept 2014

Test for UV focus shift: Newspaper on the wall. Focus for Visible light. Shoot. Then add UV-pass filter and shoot again without re-focusing. Are the letters still sharp??

 

Ideas welcomed !! And thanks for any input.

The tests need to be standardized and repeatable.

[igoriginal]

Test for UV-capability: Needs must be informal unless one owns a mono-chrono-thingie. Need a standardized test scene. Include what? Use UV-flash? Use "strong sunlight"? Landscape versus close-up useful?

 

Ideas welcomed !! And thanks for any input.

The tests need to be standardized and repeatable.

 

For those who don't have access to a properly set up and competently-used spectrometer testing method ... then may we suggest using the "Sparticle" narrowband-pass filter array testing method, as an alternative? (Per the UV community's own Steve Smeed's original testing idea).

 

I know, for a fact, that this is the "poor man's spectrometer", so to speak. :D But it works reliably well, for all general intents.

 

(Enricho Savazzi uses this test for some of his own published results, as well).

 

Granted, it may not provide very specified and highly-detailed transmission curve graphs ... but, I think that the overall testing results (if standardized by the one implementing it, and the testing procedures published so that they may be reproduced) would be sufficient and satisfactory for a formal post.

 

And as for knowing how one would gauge "standardization" across such a test, may I propose using a "control lens" (with known UV-A transmission ranges) as a "master" for the other lenses in the test. Meaning, first adjust all camera / lens / lighting conditions until results with the "control lens" match the known transmission of said lens. Then, we would have a "standardized" test for all of the other lenses in the testing batch.

 

For example: With the Kyoei / Kuribayashi / Orikkor 35mm F/3.5, which is known to transmit UV (practically) down to about 320nm or so ... we simply adjust all lighting / shooting conditions and camera settings, until the narrowband filter array shows light in the 320nm range. That would be our "control" or "master."

 

Then, we would be set to test the remaining lenses (with unknown transmissions), under the same lighting / camera setting conditions used for "calibrating" the test to match the transmission results of the known lens.

 

(I understand that this "poor man's spectrometer" is not exactly elegant nor "institutional"-sounding in protocal, but we have to be realistic about the fact that not everyone owns, or has easy access, to an optimal spectrometer test.)

 

Thoughts?

[Andrea B.]

A filterboard is useful somewhat. It does put us at the mercy of the variations in sunlight. So it might provide different conclusions depending on time of day, time of year, changing clouds and so forth.

 

Then there is the problem of fitting a filterboard with filters which don't leak IR. Even though the lens under test is IR-blocked, how do you prevent the IR in sunlight from wrecking the UV transmission of a given filter on the board? I suppose each filter in the board could also be fitted with an IR blocker. Begins to get expensive, but temporarily let's assume money is no object.

 

I wonder if Bjørn's Broncolor could be rigged & triggered to fire in back of a filterboard to provide the UV illumination through the filters?

 

I don't mind lack of elegance whatsoever. That is totally unimportant. But there must be repeatability of results for a test to be valid and useful. But illumination is a big problem in filterboard tests. Everyone has a different UV flash. Sunlight is too variable. And so on. At best the filterboard gives is not un-useful, but you must take the results with a grain of salt.

 

********

 

Anyone want to define a Hotspot test? I think I recall that you are supposed to shoot the sky or a blank wall in the aperture series. But is it best to shoot close or distant. And not shoot into the light - which isn't usually done anyway in ordinary documentary photography.

[igoriginal]

A filterboard is useful somewhat. It does put us at the mercy of the variations in sunlight. So it might provide different conclusions depending on time of day, time of year, changing clouds and so forth.

 

Yes. With outdoor conditions, this is true. But, I was referring to testing conditions which can be standardized (those that are stable).

 

And, in the case of lighting, of course, I am talking about a more tightly-controlled, indoor environment.

 

In fact, speaking of which ... I have been coming up with some less-traveled roads for alternate indoor UV lighting: That of implementing reptilian terrarium UV-A bulbs. (Standard E27 / E26 "Edison" screw base design, in my case.)

 

(In fact, I have been getting some very incredible results, going this route, lately. Both, with indoor reflective and indoor fluorescence work! And I plan to post my results on UVP, within a week or two, including my set-up and procedures. As well as what brands of bulbs to use.)

 

Unlike LED "torches", which do not emit evenly or fully across the entire UV-A bandwidth (but rather "spike" within a narrower range), the reptilian terrarium UV-A bulbs emit a very consistent, smooth, and evenly-transmitted energy across the entire spectral range! (Just like natural sunlight.)

 

Moreover, there is very little to no risk of eye or skin damage if used responsibly, because these bulbs have a very reliable intensity / distance ratio.

 

(What I mean by the "intensity / distance ratio", is that the transmissions of these bulb brands have a published UV-A and UV-B intensity "fall-off index", which is physically proportionate to the distance from the bulb. Stand at a sufficient distance from these reptilian bulbs, and the UV-A and UV-B fall-off will drop to a level where there is nominal risk to the experimenter. Which is why they are specifically designed for reptilian terrariums. Good energy at short range, but not a lot of errant energy outside of the terrarium.)

 

Thus. the camera can be released on a timer or wireless trigger, after setting up the subject / camera, and the reptilian / terrarium bulb can be "clicked" on / off with a long enough power strip from a safe distance.

 

Finally, I have come up with a way to mount filters directly in front of the bulbs, themselves, by obtaining non-expensive parts on the market. This includes various tubes (with filter threads at the end), bulb holders, and even parts for providing ample ventilation for the bulb assembly / tube, itself ... so that the entire apparatus does not overheat. (Although it is already a big help that these are CFL bulbs, and not filament or halogen-based. Thus, the heat generation, by design, is already somewhat marginalized. Thankfully, so.)

 

Being able to mount filters directly in front of the bulb is especially critical for UV-fluorescence work, of course. (These bulbs also emit some weak visible-energy ranges, after all). My tube design, in effect, blocks out all other errant light from the mounted bulb, except for what is permittied to directionally escape from the front of the tube barrel - and any added filter working as a spectral modifier.

 

Because of this design, my reptilian terrarium UV-A bulb holder / mount / tube design also makes the reptilian terrarium UV-A bulb highly directional (UV "spot light" or "flood"), so that only the intended subject is in the "line of fire", while there is no significant UV-A energy being cast at any other direction within the room.

 

More on this, later.

 

(This week, I am experimenting with spraying the entire inner surface of the tube with chrome-colored spray paint ... as to make the entire inner wall and inner rear end of the tube highly-reflective, so that even more of the UV energy emitted by the bulb inside of the tube is not wasted, but rather conserved, and shot out from the front.)

[Andrea B.]

The last photo in Enrico's post is made with reptile bulb:

http://www.ultravioletphotography.com/content/index.php?/topic/981-where-does-red-come-from/

 

It is not clear that such bulbs have no spikes??

 

...because these bulbs have a very reliable intensity / distance ratio.

How are you confirming that? Do you have a UV meter?

[igoriginal]

Good question. Yes, as a matter of fact, I do.

 

I am using Solarmeter Model 5, for my tests. Granted, it's not the most sophisticated UV-A / UV-B meter out there, but it seems to be accurate enough, since the measurements consistently reflect the published fall-off index ratings included with the bulb literature.

 

At a distance greater than roughly 8.5 feet or so from the bulb, the UV meter drops to 0 (although it seems to pop in and out of a flat 0 and some fractional amounts, when within a distance of 6 to 8 feet. Thus, the CFL is probably "cycling", somewhat. And not steady like incandescent / halogen.)

 

At a distance of just 4 to 6 inches from the bulb, the readings seem to approximate a typical reading outside, mid-day, and clear sky. So, drop-off from this bulb is quite abrupt, to say the least. Which would work well for a home terrarium where the pet owner would not want UV to reach far outside of its intended use environment.

 

Wow. I didn't know that Enrico was trying some of the same things. Neat!

[JCDowdy]

Andrea,

I would prefer something less technical coming from someone like you. I am less interested in you trying to show a Sparticle Bandpass Array than I am how you appreciate the differences and usability. By that I mean, you might shoot some different scenes to show what in your experience are relative strengths and or weaknesses of the lenses. Obviously you would control for the necessary variables of lighting and perhaps field of view (although FOV is a difference on can also compare). That sort of real world approach from someone of your experience is as valuable if not more so than some lab test.

[JCDowdy]

Iggy,

 

The better quality reptile lamps my in fact offer a convenient and affordable standard illumination source. Some of them have good CRI and so might enable good Vis to UV focal shift comparison. However they are not a smooth spectral transition like sunlight. I can tell you this with some assurance since I actually test lamps for a major reptile lamp manufacturer. This is in no way a flaw of the lamp, just how any compound phosphor UV-Vis lamp works. Also there are Hg lines and the spectrum sometimes confuses good in camera white balance somehow. I am not sure why that is, but I have also encountered intermittent anomalous behavior of my software for calculating CRI on these lamps. I think is might have to do with how the camera WB algorithm and/or spectroradiometer CRI calculation handles the UV tail of the visible as a discontinuity.

 

A better source for a smooth spectrum from Vis into UV might be a quartz envelope tungsten filament lamp. Problem with these is of course heat and lots of IR to contend with.

[Andrea B.]

I didn't know that Enrico was trying some of the same things.

Oh we've all tried everything long since while trying to find the best way do illuminate our long-suffering subjects in UV photography. :D

 

I think the big lights like the Broncolor (or anything similar) have produced the best UV lighting I've ever seen. But they are a bear to handle. Unless handled by the Bear. I should get one, but currently for floral documentation I'm happy with my much lighter SB-14 Mod which has its limitations but which would not break my toes off if I dropped it while working.

 

*******************

 

...Hg lines and the spectrum sometimes confuses good in camera white balance somehow.

 

Strictly speaking, however, in-camera WB is not necessary. Although very nice. So the question becomes: can the mercury-spiked WB be repaired in the converter/editor when you fail to fully achieve it in-camera? I'd would think so, but if not please I would be curious to know.

I have had some occasional white balance failures from mixing UV torchlight with sunlight. And not very good recovery in the editor - stuff looks all blotchy. (....mentioned this before somewhere sometime....)

[JCDowdy]

Strictly speaking, however, in-camera WB is not necessary. Although very nice. So the question becomes: can the mercury-spiked WB be repaired in the converter/editor when you fail to fully achieve it in-camera? I'd would think so, but if not please I would be curious to know.

I have had some occasional white balance failures from mixing UV torchlight with sunlight. And not very good recovery in the editor - stuff looks all blotchy. (....mentioned this before somewhere sometime....)

 

One click in PhotoNinja and all was OK!

 

Mixing UV-LED and sunlight spectra also adds another variable of different exposure geometry.

[Andrea B.]

...real world approach...

 

That has certainly been Bjørn's philosophy over the years. And one reason why his write-ups have always been so widely read around the world.

Real world testing is also important because we need to know filter sizes, helicoid sizes and such-like stuff. Kitting up a lens is not always so easy, whew!

 

For UV-capable lens testing we do need to add hotspot testing at the minimum. Too important in our little game. And I don't see anyone explicitly doing that test. Although, granted, hotspots usually tend to pop up right away if they are going to - with only the occasional exception.

 

Anyway I was thinking that if I do one Brick Wall Aperture Series then I could kill a few birds with one 35/3.5 rock.

A BWAS could reveal hot spots, show diffraction setting in and reveal corner blur.

Then I could shoot a landscape and a close-up of the usual Rudbeckia and CC Passport under Sunlight-only and Sunlight + SB-14.

 

Chrom abb is one I don't know how to do. How do you induce chrom abb if a lens is prone to it? Shoot a tree? Zoom in on the branches?

 

****

 

Just another brick in the wall......

[JCDowdy]

I have purchased a few of the ISO standards for lens testing. None of them cover CA but I do not yet have ISO 15795:2002 Optics and optical instruments -- Quality evaluation of optical systems -- Assessing the image quality degradation due to chromatic aberrations or ISO 12233:2014 Photography -- Electronic still picture imaging -- Resolution and spatial frequency responses in my standards library.

 

But that does rather drift away from the real world approach.

 

So in keeping with that, I would ask in what practical conditions have you encountered problems with flare or hot spotting and where do you normally observe CA with lenses that have these issues?

 

If I understand the brick wall test for example, it would be dependent on the solar angle relative to the direction the wall faces. If the sun is "setting" over a low wall and not to far out of the frame vs behind the camera, that would obviously make a big difference. You might not normally shoot that way but my answer to how you induce flare & hotspot is to do what you might otherwise know to avoid.

 

If you pick a lens that you know will have CA, like an uncorrected quartz lens, the CA is most noticeable along high contrast edges. I think branches are used in a CA example on PhotoNinja website, so that seems a good real world scenario to me.

 

So you run and you run to catch up with the sun but it's sinking. Racing around to come up behind you again.

[nfoto]

Flare is usually pretty much obvious. Sometimes a narrow lens hood can improve the situation. CA may or may not be equally obvious and spherical aberration factors itself into that equation as well. If you have axial colour (longitudinal CA) then there will be different colours on either side of the focused plane. Stopping down tends to lessen the CA issue. Most of the 35 mm f/3.5 lenses we employ for UV do have some chromatic and/or spherochromatic aberrations.

 

Hot spots are not usually a problem in UV work. I have only encountered it once with an ordinary lens (the 13.5 cm Nikkor-Q in S mount, if anyone wonders). The Coastal 60 mm f/4 APO does show hot spotting commencing around 1:3 magnification and apertures smaller than f/8. A specially crafted lens shade does remove the hot spot issue, though. Special contraptions such as stacked lenses can produce hot spots as well.

 

All UV set ups are highly susceptible to stray light leaks (in filter pack(s), adapters, focusing extension, read-out windows, etc.). It can sometimes be real detective work to track down and eliminate such leaks.

[Andrea B.]

I haven't quite sorted out how to detect purple fringing in a purple UV foto. :D ....j/k.....

 

 

As has been mentioned somewhere, a narrower-band UV-pass filter could also help against either axial CA or magnification CA when using an uncorrected lens.

 

...what practical conditions have you encountered problems with flare or hot spotting...

The usual stuff. Shooting against the sun. Stacked filters (fairly rare, though). Reflections off other UV-reflective stuff (fairly rare, though). Not placing the pink shiny side of the UV Baader towards the subject (also, yes, fairly rare).

 

But also note that UV-capable lenses which are not dedicated UV-lenses may have some odd internal materials which can bounce the UV light around. I once had a perfect image of some interior Novoflex 35/3.5 "rings" register in a photo. Only that one series, never since then. I have NO idea what I was doing that caused that. Wasn't shooting into the sun. Guess it was just one of those lucky alignments of all shooting factors.

Can't remember other examples of this kind of internal materials problem which have been reported, but shining a UV torch into a lens and looking around for reflections is good practice when a new lens has arrived for testing. Similarly look for IR reflections as we all know what those can do to a photo.

 

Stray light leak when doing UV photography has bitten me more than anything else except the blood-thirsty Maine mosquitoes. Add open port doors to Bjørn's list above.

[enricosavazzi]

[...]

(This week, I am experimenting with spraying the entire inner surface of the tube with chrome-colored spray paint ... as to make the entire inner wall and inner rear end of the tube highly-reflective, so that even more of the UV energy emitted by the bulb inside of the tube is not wasted, but rather conserved, and shot out from the front.)

A clean aluminium surface has a high and practically flat reflectivity from NIR to UV-B or UV-C. This should be the easiest type of UV reflector to use without risking "strange" wavelength-dependent effects. Paint might work, but it is anybody's guess what the organic binders might do to incident UV.

 

In my UV "studio", I use small flat aluminium panels as reflectors, cut out from sheets originally designed for use inside fluorescent or halogen light armatures (one of my local construction materials and plumbing store sells 1 m by 30 cm pieces). The type I use is embossed with an irregular pattern to break up and diffuse radiation, and has a shiny and a more matt side. Aluminium kitchen foil, as long as not waxed or covered with a plastic film, can also be used to "wallpaper" the inside of a lamp casing. The matt side is better to avoid concentrated hotspots.

[Andrea B.]

Great advice, Enrico, thanks. Aluminum panels as reflectors for UV - this could also help with outdoor work too. I could just cover a small canvas with Al foil.

 

I was also worried for Igor that paint inside a tube might be flammable or something. However, I'm sure he's careful about stuff like that.

 

************

 

And so I think I will go clip a handful of Rudbeckias and run a trial Lens Test. Then you can let me know if it is useful or not and what should be refined.Much of this effort is for me to learn more, you know !!

[JCDowdy]

I second Enrico, front surface Al mirrors are what I use for UV beam steering.

 

What is useful for me to learn are the field conditions that might then be modeled in a standardized laboratory test. What will be more difficult to quantitate is the subjective observable, how does one measure flare for example? One is likely to more easily measure it effect as degradation of some other directly measurable parameter perhaps, but which?

[igoriginal]

Thanks for the advice, folks!

 

Yes, I am aware of using aluminum and other shiny / reflective surfaces for "mirroring" or "paneling."

 

However, the reason I am experimenting with chrome "spray-on", is because the tube designs I am using are not easily "wall paper-able." There are many tapers and extruded points along this tube, which would create uneven and jagged creases and wrinkles along the inside of the tube when trying to "wall paper" something onto it. And that could be potentially counter-productive. (Some of the surfaces I am expeimenting with terminate in a parabolic, funnel, or even pyramidal shape. I am definitely trying a lot of shapes out.)

 

Yes, Andrea, I am quite careful about looking into the testing parameters and use indications / contra-indications of this metallic spray-on. My father, in fact, is a retired industrial engineer and professional welder (he worked for the automotive parts re-manufacturing industry, too). Thus, I am going on a lot of his own suggestions and prior experience. This "spray-on chrome" is quite durable, non-flammable, and has a high-temperature point indication, for industrial uses. (The finished surface can even be welded with!)

 

Thus, I am experimenting with this ... not because I am unaware of the common other alternatives ... but because I am trying to see if there is a more efficient and uniform application process in adding a reflective surface without trifling with "wall-papering" things.

 

Not sure if the light output would be as efficient as some other surfaces, of course, but that's why I am trying something different.

[Alaun]

Igor, what type of spray on chromium is this? I googled a bit (because I did not know that stuff) and found many of these products are based on silver actually... and there is a good reason, as chromium can be quite unhealthy ...

 

What I know what is used on big furnaces is Aluminum paint, which is made of tiny Al-flakes and linseed oil (as far as I know). This color withstands even very high temperatures. What I do not know is, how the reflectivity is with regard to UV, but I think it could also work as some kind of nice reflecting (not mirroring) surface.

 

With respect to the tubes and bulbs, my experience is, that even bulbs from the same supplier, with the same item no. coming from different factories can show a different spektrum (I have a set with 4 tubes and had to replace one, Phillips tubes).

 

Werner

[igoriginal]

@ Alaun,

 

It's a special "powder-coating" type paint, designed by Alsa Corp. There are no actual chromium substrates used in this highly-reflective, mirror-like, "chrome"-like paint. It simply simulates the look.

 

But, these paints are considered "exotic", and they are NOT cheap. One also is required to purchase the proper application apparatus (spray gun, high-pressure system, dispensing canisters, etc.) in order to implement it correctly. The process requires curing time. And the finished, mirror-like surface is highly durable, can withstand a constant high-heat environment up to 400 degrees, and is resilient to UV radiation, too.

 

It's also anti-corrosive and salt-resistant (these paints are also being used on boats, and not just automobiles!)

 

Here is the info, if you want to read up on it:

 

http://alsacorp.com/portable_powder_coating_system/index.html

 

Now, whether this experiment will yield satisfactory results in terms of UV spectrum reflective photography, I do not know. I am curious in experimenting with it, though. To see how UV-reflective it is.

 

And since my father already has some of these supplies left over, from prior work ... then I see no harm (or added cost) in trying it.

[Andrea B.]

Well, write us up a nice report when you are done, Igor. It will be interesting!! :D :D

 

 

 

Here comes the sun and I say, "It's all right".

[igoriginal]

I plan on it! (Like so many other projects I am overly ambitious in juggling at the same time. Hehe. :D )

 

But, hey. It might turn out to be an epic failure, or an illuminating (no pun intended) success. Either way, experimenting is the backbone of our UV community. We won't know, until we try. Right?

 

Thank you, Andrea!

[enricosavazzi]

I tried to post on this thread yesterday, but something ate my post.

 

Basically, my suggestion about what to test in a UV lens includes a relatively simple test for chromatic aberrations (of both kinds). This is not new, but it can be standardized to make results more easily comparable.

 

Choose a subject with plenty of detail but not too much 3D relief, and not easily blown around. Place the camera on tripod and remote control, out of wind, air currents and strong floor vibrations. Use a wide or fully open lens aperture and a multispectral light source (preferably UV-enabled electronic flash, sunlight may do unless weather is very variable).

 

1- Focus in live view with Baader U (because just about everybody among us has this filter, so it is kind of a standard).

2- Take a first shot.

3- Change filter to a green filter (or red, we may need to discuss and test which is most likely to show chromatic aberrations with respect to UV). Do not refocus.

4- Shoot a second image.

5- If inclined to do so, change to NIR-pass filter, do not refocus, shoot a third image.

6- Convert all images to BW.

7- Combine the two/three images together as separate color channels into a single image.

 

Chromatic aberration of both types should show as an obvious mismatch in the borders between light and dark areas in the three color channels. The result might be easier to see with a subject that has a high contrast at all wavelengths but reflects evenly throughout the used radiation bands. A perforated aluminium sheet or net on a dark cloth background comes to mind.

 

PS - This test is designed to evaluate what is sometimes called "UV focus shift", which is actually a form of chromatic aberration. Focus shift, properly speaking, is not wavelength-dependent but aperture-dependent, and more closely related to spherical and a few other aberrations.

[nfoto]

While the test as such is OK, it might not be very informative.

 

Firstly, virtually all lenses will fail to pass this test, even such UV workhorses as the UV-Nikkor and the Coastal 60, if IR is included. The Coastal will *almost* pass, but there is a tiny difference in image magnification (

 

Secondly, as most of our current UV work is conducted by focusing in the actual spectral band using excellent band-pass filters and Liveview, focus shift between visible and UV is no longer keeping the practical importance it formerly had. If one does multispectral composites, however, I agree focus shifts should be clarified and quantified.

 

Thirdly, we may observe axial colour issues also in the narrower UV range, but spherical aberration apparently is much more destructive and for non-UV designs might not be eliminated even when stopping the lens well down.

 

So the basic aspect remains: if the lens is able to deliver an adequate performance in UV (in terms of UV response and image sharpness), we can use it, if not, use something better suited for the task. As each has his or her own standards for image quality, I agree testing according to some carefully described experimental regime is beneficial. I'm personally left with the impression that a lot of lenses can do adequate UV work and in particular if one is aiming to record UV floral signatures and have sufficient UV illumination.

[Alex H]

I tried to post on this thread yesterday, but something ate my post.

 

Basically, my suggestion about what to test in a UV lens includes a relatively simple test for chromatic aberrations (of both kinds). This is not new, but it can be standardized to make results more easily comparable.

 

Choose a subject with plenty of detail but not too much 3D relief, and not easily blown around. Place the camera on tripod and remote control, out of wind, air currents and strong floor vibrations. Use a wide or fully open lens aperture and a multispectral light source (preferably UV-enabled electronic flash, sunlight may do unless weather is very variable).

 

1- Focus in live view with Baader U (because just about everybody among us has this filter, so it is kind of a standard).

2- Take a first shot.

3- Change filter to a green filter (or red, we may need to discuss and test which is most likely to show chromatic aberrations with respect to UV). Do not refocus.

4- Shoot a second image.

5- If inclined to do so, change to NIR-pass filter, do not refocus, shoot a third image.

6- Convert all images to BW.

7- Combine the two/three images together as separate color channels into a single image.

 

Chromatic aberration of both types should show as an obvious mismatch in the borders between light and dark areas in the three color channels. The result might be easier to see with a subject that has a high contrast at all wavelengths but reflects evenly throughout the used radiation bands. A perforated aluminium sheet or net on a dark cloth background comes to mind.

 

PS - This test is designed to evaluate what is sometimes called "UV focus shift", which is actually a form of chromatic aberration. Focus shift, properly speaking, is not wavelength-dependent but aperture-dependent, and more closely related to spherical and a few other aberrations.

 

I would like to note here that the thickness of the filters as well as optical properties of the glass used in them can affect the focus shift observed, especially at higher magnification, or when rear-mounted filters are used. Here is an example, shot using UG5 filters of different thickness: 1mm, 3mm and 1+3mm (stacked). The light source was Nichia 365nm LED. Naturally, nothing was changed between the shots, except for filters.

 

 

So the filters used should be made such that the focus shift they introduce for their respective main wavelength is the same and does not influence the results of the test.