Jump to content
UltravioletPhotography

Here's a question for you


Andrea B.

Recommended Posts

Are there any lenses which are NOT good for UV?

 

These days it seems like we have more and more lenses which can be used for UV work. I'm beginning to wonder if our Lens Sticky is still useful to anyone? So maybe we should try including some lenses which do not work well for UV?

 

Got any recommendations for this negative kind of list?

Link to comment
Andy Perrin
Honestly I tend to regard the ones that only pass to 380nm as “not good for UV” even though they are on the sticky.
Link to comment

But I fear I am adding too many mediocre lenses to the Sticky and should perhaps prune the listings a bit.

 

If only I had real spectrometric gear.......

....and knew how to use it.

Link to comment
Andy Perrin
If you can’t get any false color with a lens, that lens is not “good for UV” in my opinion. Ditto if there is too much chromatic aberration. If you are looking for criteria.
Link to comment

This is a great question. I have found that there is a difference between range of a lens and transmission of a lens for UV. Thus in my last test with the M43rds and 43rds lenses on my Em1, I came up with T370 value, which to me is more important.

The Olympus 35mm 43rds lens looks to have greater range into UV, but not as good a T370 value, So I would rather use the M43rds 30mm f3.5 macro as its range may be limited, but had a much better T370 value.

 

The Sigma 30mm f2.8 m43rds art lens can even beat out my UAT for T value in the UVA area and can be used to automatically focus stack in the camera.

 

I will have to retest them but I think the EL80 f5.6 has almost the same T370 value as the Wollensak 25mm f1.5 lens. Finding a fast lens is really the hard part.

 

Link to comment

I think the Lens Sticky has some value as an overview for hints of what people consider usable for UV.

 

One of the problems with the list is that how the lower cutoff wavelengths is defined.

That varies between different contributors and their methods.

Some use "Sparticle" type methods with more or less well defined filter transmissions, some use spectrometer graphs with different cut off levels and some just use hands on usability tests looking at resulting UV-photos.

 

As there is many different types of UV-photography going on, what is important for UV differ.

It is very important with a very deep UV transmission range if you try UV-B imaging, but not for Bee's vision imaging, etc.

 

The sticky would be more usable if it gradually got links to more formal ambitious posts with evaluations and tests for individual lenses. (Format TBD)

Link to comment

I think the list offering a general hint is correct.

Then individual images with the lenses tell a better story.

 

Andrea, in an other thread you asked about a speed booster. My Metabones speed booster also seems to work in UVA. This is why thinking about T370 value is important. The extra stop of light it gathers with the UAT lens, equals the stop lost due to the elements. So the exposure is the same, it just offers a wider field of view. Using lights bellow 365nm, it dramatically impacts the image and is not really usable.

 

I may or may not be critical, but I would say none of the non dedicated quartz UV lenses are going to work in UVB. Yes I can see something through a 313nm filter, just with the EL80 f5.6, the Steinheil 50mm f2.8 and my igoriginal 35mm f3.5 lenses. But none of those images are usable. The limit for what is a good image for a none quartz lenses is the 365nm area.

This is why I am pitching the idea of a T370 value, as I think that would be most useful to people. But it too would not tell the whole story as then you would need to see haze, sharpness and what out of focus blur looks like to really build a useful image.

Also when you look at the T370 value you would need to place it in contrast to the maximum aperture of the lens. My Sigma 30mm f2.8 (3.6) has almost the exact same T370 value as my Sigma 30mm f1.4 lens (3.2). But the image from the f2.8 lens is better being at f2.8 and not f1.4.

Link to comment

David - How is your measurement setup designed to determine your T370 values?

 

I agree that a T370 value for a lens could be a useful parameter to show how a lens affects the exposure time for common sunlight based UV-photography.

In that light situation it is much more important than at what wavelength the lens reaches a 3-stop attenuation, in KDS-style.

Link to comment

For my T370 calculation I did the following.

Using two 365nm LED bulbs I illuminated a subject flower. I hold the white balance constant using the custom dial in section on my my EM1, using a value of 2500.

Then add the lens and use my 370bp15 filter stacked with my Baader venus U filter to avoid the IR leak, adjust the exposure so the histogram is centered. This part I know is not ideal, calculating directly from the flower would be better, but I haven't done that yet. I mainly look for similar illumination of the flower between the reference lens and my testing lenses. Also I don't know the exact distance between the two lights and the subject, so I always use a reference (previous tested lens) to compare with.

I should just take a day and retest all my lenses in one go.

The last test I did was at f4, f5.6 and wide open on the lenses. I did this to ensure that the lens scales linearly with aperture. So that I can back calculate the T value at the wide open aperture. The Wollensak last I tested was funny here and will need to be retested. Since it had the same exposure wide open as at f2. But I have since cleaned it and it needs to be retested anyway.

 

This is the spectra for my 370bp15 filter. Not a reference spectra but for my actual filter.

post-188-0-96240200-1561895872.jpg

Link to comment

I should also explain what I think a T value is.

I assume the quartz UAT lens to be true at its aperture to T transmission at 370nm. This maybe wrong, but I don't have any other lens to compare with. Thus I assume a f5.6 exposure with the UAT to be true.

 

A fictional example:

Test lens Z at f5.6 and I see the exposure is ISO 200, f5.6, 8 seconds. But the UAT was ISO 200, f5.6, 4 seconds for matched flower illumination. Thus the T370 value is one stop less than true. So then I would report the T value as one stop less than the listed maximum aperture. So say my z lens was a 35mm f2.8 lens, I would say it has a T370 value of 4.

Link to comment

OK, thanks for the detailed description of the method.

I think I understand your reasoning behind the process.

By matching the exposure and using the exposure time as the measuring parameter you get a good indication.

 

I like the idea and in my mind I am already designing a possibly improved setup.

 

Your method is at least reasonable for a relative comparison locally for you and your lenses.

 

First of all I do not think the UAT lens transmission is true to the aperture values, indicating real T-stops.

That is normally not that important for photography. High end lenses for video are normally T-Stops instead of F-Stops.

 

If I am correct about that the UAT is not showing true T-stop values, it s not an ideal reference lens for this as it is both expensive and rare.

I am sure that it is an excellent lens, but possibly not much better for this purpose than other more obtainable alternatives.

 

Then any simple common lens with a known good UV-transmission would work almost as good as the UAT.

I suppose it would be good if others than you can use the method too and make comparable measurements.

My suggestion for a more widespread and cheap alternative might be the old version of EL-Nikkor 85mm, metal version.

 

I think I see a few possible weaknesses that might affect the measurements.

I might be wrong please do not see this as anything except my thoughts, theories and ides for improvement of the method.

It is always easier to find improvements than to get the original idea.

 

It is good to insulate the test light source with filters.

 

Both the intensity and wavelength of a LED shift.

Both parameters are affected i by the chip temperature and naturally also by the drive current.

With a simple light source like a battery powered torch, the battery voltage will strongly affect the current and light intensity.

It is a good idea to keep those constant and allow stabilisation time of the light source.

 

If the wavelength of the light source is shifting during the measurement and the peak is on a slope of filter-transmission or camera sensitivity you will see variations in exposure not only caused by the lens, but the full combination of the setup.

 

The reflectivity of a flower is not very transferable and might not be stable over longer time.

Could a piece of some kind of PTFE-material be a better alternative?

Then you will not be affected by the reflectivity variations of the flower.

Some type of matted material would be best to avoid variations caused by specularity.

 

It is difficult to say how much the result is affected in a simple measurement setup.

Link to comment
Andy Perrin
My suggestion for a more widespread and cheap alternative might be the old version of EL-Nikkor 85mm, metal version.

I like this since I have one. dabateman, maybe you could re-test a couple of lenses using the EL-Nikkor 85mm 80mm, metal and see if you get the same T370 values you got previously with the UAT?

Link to comment

Yes I have the Nikkor 80mm EL f5.6 old metal one. I think that is what you are referring to. I will compare that in my next test.

I also just remembered Jonathan post here:

https://www.ultravioletphotography.com/content/index.php/topic/2580-build-thread-at-home-measurement-of-camera-uv-spectral-response/page__view__findpost__p__19599

 

That the Nikon 105mm is better than the UAT, but they are both close. So the UAT will not be perfect.

 

Link to comment
enricosavazzi

If you can’t get any false color with a lens, that lens is not “good for UV” in my opinion. Ditto if there is too much chromatic aberration. If you are looking for criteria.

 

But I fear I am adding too many mediocre lenses to the Sticky and should perhaps prune the listings a bit.

 

If only I had real spectrometric gear.......

....and knew how to use it.

Both of the above are a good start toward finding a general solution to the problem. The criteria for an "acceptable" performance of a given lens in UV imaging are, at present, too loose, as discussed in several posts here and in other threads. One way forward could be to avoid just a binary (good vs bad) judgement, and formulate two or three subcategories within the "acceptable" category. Since we are fairly in agreement by now that we generally see one or two UV false-color in images shot with the majority of legacy lenses (UV-blue and UV-violet), or three with lenses designed for UV imaging and a few legacy lenses (UV-blue, UV-violet, UV-yellow), each of these false colors could be a subcategory (I am leaving out UV-green for the moment, which is more difficult to reproduce). This without implying any formal relationship between false color and wavelength, since in the past discussions of this relationship have proved to be divisive and have not lead to useful conclusions. I think most of us can agree that, with most Bayer sensors and the best available UV lenses, we do see at least UV blue, UV violet, and UV yellow (regardless of what these false colors mean) with the Baader U filter, which is the closest thing we have to a "standard" UV-pass filter.

 

I general, we also tend to agree on the principle of a suitable custom white balance, to be preferably created in-camera by easy means, with sunlight, and for each specific lens. The spectral composition of sunlight does change with geographic position and season, but a good number of cameras, especially modern ones, can do a reasonably good job of creating a sunlight CWB capable of letting us visually distinguish among these three false colors in the same image.

It would obviously be best to agree on test subjects for each of the three false colors. Ideally, these subjects should be of general availability and not restricted to a particular region and season (flowers have both problems, for example). Most likely this would have to be a type of material or pigment (not too expensive and of easy availability) as a "tracer" to reflect each of the false-color subcategories. We already know how bandpass filters with sharp cutoffs are difficult to find, expensive and often limited to small production batches. These limitations make filters not really suitable as broadly available test criteria. The various flavors of test PTFE targets are not helpful for this purpose, since they are designed as broadband diffuse UV reflectors, besides being easily damaged and contaminated.

 

I do have some spectrophotometry equipment, but in my opinion this is a whole can of worms that creates more problems than it actually solves. For example, I am aware of a few legacy lenses that display such an extreme amount of axial chromatic aberration (a.k.a. UV focus shift) within the NUV spectrum to make them useless for UV imaging unless a monochromatic UV illumination or a narrow-band UV-pass filter is used. Some of these lenses pass simple spectrophotometric tests of UV transmission with flying colors.

Link to comment

I do have some spectrophotometry equipment, but in my opinion this is a whole can of worms that creates more problems than it actually solves. For example, I am aware of a few legacy lenses that display such an extreme amount of axial chromatic aberration (a.k.a. UV focus shift) within the NUV spectrum to make them useless for UV imaging unless a monochromatic UV illumination or a narrow-band UV-pass filter is used. Some of these lenses pass simple spectrophotometric tests of UV transmission with flying colors.

 

I do not agree that spectrometric measurements creates more problems than it actually solves. They show a vital part of the UV capability for the lens.

However for to be a lens usable for UV photography several other aspects must be evaluated.

 

I am confident that my spectrometric relative transmission measurements are valid, but they say nothing about the absolute transmission or any focus shifts, sharpness or other aberrations.

 

We might be able to define a set of tests to be performed for evaluating the UV capability of a lens to acquire more facts beyond a simple usage-test.

I think that David's method above, possibly enhanced, can be a part of that set.

Link to comment
Thank you to all contributors so far. Great suggestions and discussions! Keep it coming.
Link to comment
I was trying to find a easy DIY method with easy to find household items for testing UV. I tried to see if a cotton swab dipped in different liquids act differently in UV and can be a sign of how well a lens works in UV. I tried different things like bleach, water, lemon juice, vinegar etc... lemon juice for example is a good UV absorbant and with lenses that transmit UV better it seemed to be more dark. But so far I could not get any false colors for different liquids, seems to me they "act" like a sharp cut filter, so below a certain wavelength they go from high to low reflection in an instant. But it would be good if we could come up with a way to test lenses with stuff you can get in your local groceries store. And then we could start ranking the lenses and maybe have a new category for lenses that transmit a bit of UV but would require long exposure and by that have higher risk of non UV "pollution"
Link to comment
Are sunflowers easily available in European grocery stores like they are here?
Link to comment

Are sunflowers easily available in European grocery stores like they are here?

well not the flowers itself, but the seed, yes. for the flowers I am guessing it could be seasonal and I believe we will find them at florist and not groceries :D

Link to comment
Andy Perrin
Hah, in the US, every kind of store tries to sell every kind of merchandise, so we end up with grocery stores selling flowers, and drug stores selling portable heaters...
Link to comment
Yes, it has gotten a little crazy here in the stores. My local grocery store sells flowers, books & magazines, gift wrapping paper, small kitchen appliances, dishes, pots and pans, toys, basic T-shirts & underwear, over-the-counter medicines, cosmetics and more!!! :lol: I think because the small stores are all gone now, that the big stores have branched out into other areas. And of course there is always Walmart or Target which seem to sell everything.
Link to comment
I have a question, of all the UV capable lenses, which ones, if any, would you consider good for UV that also have larger that 52mm filter threads?
Link to comment

Andrea's answer to that question is that she doesn't have any UV-capable lenses currently which have a front diameter larger than 52mm. Although she has one which she's not sure of - the Spiratone 75/3.5. It does not have a useable front filter thread and is probably 53-54mm if it did have. So I just hand hold the 52mm filters to it.

 

30.5 mm x 1

34.5 mm x 3

35.5 mm x 3

40.5 mm x 2

43.0 mm x 1

46.0 mm x 10

49.0 mm x 11

52.0 mm x 4

unthreaded x 7

Link to comment

Andrea you have 42 UV capable lenses!

Well I guess that is the meaning of life after all.

 

Cadmium,

I haven't tested the Lentar 400mm or 450mm ones I have yet. But you said they were good and both have large filter threads, 72mm and 77mm respectively.

 

I have a small collection:

Tamron 24mm f2.5

Wollensak 25mm f1.5

Olympus M43rds 30mm f3.5 Macro

Sigma 30mm f2.8 m43rds art lens

igoriginal 35mm f3.5

Olympus 35mm f3.5 macro 43rds

Steinheil 50mm f2.8

Nikkor 80mm f5.6 EL

UAT 85mm f4.5

 

That is more than I thought.

 

Link to comment

Please sign in to comment

You will be able to leave a comment after signing in



Sign In Now
×
×
  • Create New...