Jump to content
UltravioletPhotography

Lenses' UV photography suitability


dunksargent

Recommended Posts

dunksargent

I am new to UV photography and await return of a G1 camera currently being converted to full spectrum before commencing UV experiments .

 

Regarding lens suitability: Is it possible to ascertain a lens' UV photo. suitability without using a camera by passing a UV light through it (UV torch) and measuring wavelength emitted with a UV meter ?

 

If so, can anyone recommend a suitable UV light meter and UV torch?

 

Thanks

 

dunk

Link to comment

We note in the Lens Sticky that lenses good for UV typically have few elements and no coatings. (Dedicated UV lenses may vary from this, of course, because of use of quartz or fluoride elements.)

 

It is possible that you might get some idea of a lens suitability for UV photography by shining a UV torch through it and measuring what wavelength or how much is emitted. UV torches themselves vary a lot in bandwidth and efficacy, so that might be a problem. I think you would want to test with something like a high-power Nichia 365nm UV-LED. Not all Nichias are the same.

 

For the record, I've never tried this myself so I'm not really providing a thorough answer to your question - only suggesting to choose the torch with care if you try this. :unsure:

Link to comment
  • 2 months later...

let me quickly bump up this topic (instead of opening a new one with quite the same question)

 

So I really wonder how do you test a lens for UV suitability. I am having a method on my own but I am quite sure it's not the possible best :D

 

So for the tests I am using a black light bulb (all I know it emits UV-A light... not sure of the bandwidth) and a plain £5 lamp from a local hardware store. I put it near the fridge (about 1m) and I use a magnet to put a white paper on the side of the fridge, illuminated by the bulb. I use my Sony Nex 6, first without any lens (just a NEX-M42 adapter) and I hold the filters (Hoya U-360 + Schott BG40 combo) close to eliminate any leakage and take a photo of the paper (making sure to not cast any shadow). Then I take a shot with the lens I want to test, aiming to have the same brightness on the paper than before (spot metering at center). Once I got the two shots I compare the shutter speeds and withdraw the EV lost by the lens aperture (I assume the filter only version is f1.0). Unfortunately I don't have any of the lenses mentioned in the sticky list as a good performer and I think there is absolutely no way for me to justify the price of a true UV lens to my wife, so I can't really compare yet with known good performers.

 

So far the best UV performer for me surprisingly a modern lens! The Sigma 30mm f2.8 for Sony E mount is giving me the best results! I am thinking to get one more and use some elbow grease on it* to further improve it.

 

But again... what I find as a good result you might be laughing at as poor performer. So that's why I am asking if anyone else have any good DIY method to test a lens for UV capabilities?

 

Notes:

* - (If you're not a native English speaker like me, do NOT go in a hardware store asking if they have any elbow grease on stock... :D unless you wanna see funny reactions from the store keepers...)

Link to comment

If you just want to compare lenses that you have already, why not shoot the same scene, under same light condition, using same silter set, lens set to same aperture, camera with same ISO, object at same magnification ratio, and see which lens produces best quality images and fastest shutter speed. That would be simplest.

 

Othervise, you can get one of the narrow band filters from eBay for not too much (with the peak anywhere between 330-350 nm and bandwidth of about 10-15 nm). Thank you can sort lenses by how "deep" in UV your lens does.

Link to comment
enricosavazzi

Hi Timber,

 

funny that you should mention the Sigma 30 mm f/2.8. I have one in Micro 4/3 mount always attached to the photo tube of a Zeiss photomicroscope, where it works very well, but never thought of testing it for UV (the photomicroscope is the only reason why I purchased this lens).

 

Below a simple test with Baader U and UV-enabled electronic flash. The first lens is the CoastalOpt 60 mm, the second the Sigma 30 mm, both at f/16. Everything remained the same between tests, except the lens and the distance between camera and subject.

post-60-0-35436300-1409078001.jpg

post-60-0-87633600-1409078001.jpg

The Sigma 30 mm obviously transmits less. The tone is also more strongly violet, which means it is transmitting mostly in the 380-400 nm range. I did not make quantitative measurements, but I guess the overall difference in transmission is between one and two stops. In conclusion, it requires an exposure between 2 to 4 times longer with continuous light sources like sunlight. 380-390 nm should be enough to show a good bullseye pattern with common flowers (not in high contrast like typical with good lenses and Bader U, but more like Shuler UV, Astrodon Uvenus and Schott UG5).

 

What is really interesting, though, is that this is the first autofocus lens I know of that is usable for UV photography with mirrorless cameras, which may have its uses.

 

The vignetting in the second picture is simply due to the fact that I am using a lens shade optimized for the CoastalOpt 60 mm. The Sigma 30 mm itself shows no significant vignetting.

Link to comment
enricosavazzi

I cannot seem to be able to add pictures when editing my preceding reply, so I am making a new one.

 

Olympus 12 mm f/2, same shooting conditions.

post-60-0-03877800-1409079992.jpg

 

UV transmission is even poorer, and restricted to the longest UV wavelengths. However, It can still be usable with long exposure times or strong UV source. So far I did not know of any autofocus 12 mm wideangle usable anywhere near the UV.

 

I was holding a 2" Baader U in front of the lens with my fingers, so there is a little uneven flare.

Link to comment

Enrico, Sony SEL 16mm 2.8 autofocus lens for NEX cameras can be used for UV, except for the furthest corners. It's been discussed long ago on another forum. Here are two examples:

 

http://www.holovachov.com/img/s1/v49/p106352348.jpg

 

 

http://www.holovachov.com/img/s1/v46/p302885299.jpg

Link to comment

I only tried the little olympus 45/1.8 , sunny outside, all hand held, very nice autofocus, a bit of wind with Panny GH3uvir and Baader U, all at 1.8 ISO 1600 and 1/15 to 1/25s:

 

 

with respect to possible sharpness:

post-21-0-86852900-1409081105.jpg

 

blue:

post-21-0-32823800-1409081120.jpg

 

yellow:

post-21-0-04595800-1409081134.jpg

 

 

Werner

Link to comment

my original question is how could you measure "home made" how well a lens transmits? Like let's say "A" lens is taking off 2 EV in UV and "B" lens is taking 5 EV.

 

The method I've tried is to put an M42 - NEX adapter and no lens on the camera and then I shoot a light source. I take this as my reference, I assume it as being f1.0 as there is no glass element (I only use the M42 adapter to make light a bit more "directional"). Then I shoot with the lens with the histogram on and trying to keep the same luminosity of the "scene" (I don't focus the lens so the blur will help me to achieve more "smooth" lighting) and then check the EV difference and withdraw the EV for the aperture (like f2.8 is 3EV).

 

So for an example:

without lens: 1/400 (f1.0)

with lens (an uncoated -polished- Jupiter 8 2/50): 1/40 (f2.0)

Then it's roughly 1.5 EV (the difference is 3.5EV but f2.0 is 2EV different from f1.0)

 

Now I know it's very un-scientific as it is but this gives me a rough idea on which lens worth disassemble and polishing to achieve better results. As you said same light settings, same scene and better shutter speed means better transition, but I would like to measure, and since I don't have any of the UV excellence lenses I just don't have a standard to compare other lenses. That's why I thought to have the no-lens option as the base standard. Sorry, I just like to be over-technical data geek :D

 

But if anyone else have a better way to test a lens then please let me know :) Some lenses take insane amount of time to get the coating off :)

Link to comment

The method I've tried is to put an M42 - NEX adapter and no lens on the camera and then I shoot a light source. I take this as my reference, I assume it as being f1.0 as there is no glass element (I only use the M42 adapter to make light a bit more "directional").

 

Why do you assume that the camera without the lens is equal to F/1.0? There are lenses that are faster than that. F/0.95, F/0.75, etc.

 

But if anyone else have a better way to test a lens then please let me know :D Some lenses take insane amount of time to get the coating off :)

 

If you just want to compare lenses that you have already, why not shoot the same scene, under same light condition, using same silter set, lens set to same aperture, camera with same ISO, object at same magnification ratio, and see which lens produces best quality images and fastest shutter speed. That would be simplest.

 

Othervise, you can get one of the narrow band filters from eBay for not too much (with the peak anywhere between 330-350 nm and bandwidth of about 10-15 nm). Thank you can sort lenses by how "deep" in UV your lens does.

Link to comment
enricosavazzi

[...]

But if anyone else have a better way to test a lens then please let me know :D Some lenses take insane amount of time to get the coating off :)

I don't think you can assume that no lens = f/1 lens. Take for example pinholes, they have no lens but do have a finite aperture very different from f/1. The front diameter of your lensless adapter functions like a giant pinhole, and the distance from the front opening of the adapter to the sensor is equivalent to its focal length. Most likely diameter/FL is not 1. If something else is limiting the amount of light reaching the sensor, then also this rough approximation of the speed of the no-lens will be off. Even the NA of the microlenses on a Bayer sensor may become the limiting factor of the effective aperture with a giant pinhole.

 

Basically, if you want to use the camera as a photometer (which it is not designed to be), it has to be used as a relative photometer, and you need a reference lens. It does not need to be a very good UV lens in terms of image quality. A fused silica meniscus lens cannot be too expensive, should be a very good reference lens in terms of UV transmission, and once you know focal length and diameter (which you can change by putting a circular aperture in front of it) you can easily work out its f/ratio. A non-colored Pyrex meniscus element can also be used if you are only interested in VIS and UVA.

 

You might even be able to use a moderately small pinhole (f/16 or f/22) as a reference lens that transmits 100% of UV. It is certainly going to be more accurate than an open adapter, and you can directly compare the results with real lenses also stopped down to f/16 or f/22.

Link to comment

I think I will go out and try a UV pinhole foto.

 

Bjørn has always said that if you can record a known UV signature of a subject, then the lens is UV-capable.

(We usually choose Sunflowers, Black-eyed Susans or Dandelions for our known UV signature subject.)

The rest is about your personal preference regarding how easy or difficult it is to record the UV photo with a particular UV-capable lens.

 

For lens comparisons I would probably pick from my lens set the lens that offers the shortest recording time of a particular well-illuminated sunlit scene and compare the remaining lenses to that base lens in the same sunlit scene. Over time as new lenses are acquired, the selection of the 'best' lens may change, but that is ok as all the measurements are simply relative.

 

As to how deep into the UV a lens can record, that is harder to determine. The non-laboratory method is to use a filter board to give an idea of a lens recording range. But even that requires collecting some filters. However, most generic UV photography does not require a lens which shoots under 365nm or 350nm.

 

So, if you can record the bullseye on a Dandelion in an f/8 exposure in under 4 seconds in strong sunlight, then you are good to go for basic UV shooting.

 

Maybe I should make that a general statement somewhere?? I'd have to decide whether it should be f/8? f5.6? 4 seconds? 2 seconds?

Link to comment

For a pinhole, "focal length" is taken to mean distance from the aperture to the recording surface. Thus, for a Nikon camera with no lens, the "free opening" is about 40.5 mm and the "focal length" is 46.5 mm. So the 'effective aperture' is approx. f/1.2 on-axis. As the blur circles are humongous, no sharpness can result at all. In fact, a pinhole trades enormous depth of field by no actual sharpness at all, since the resolution can not be better than the blur circle size. However, when the recording format is large enough, you can get a pleasing rendition even with blur circles around 1 mm in diameter. 'Large' here meaning say 4x5" or more. I shot pinhole on 4x5" in the old days and got pretty impressive results (that is, until you looked too close as no details actually were sharp, they just looked that way).

 

pinhole_F990929887.jpg

 

I searched for examples of UV pinhole, but in vain. The captures might not have been scanned yet.

Link to comment

Wow! Thanks for the answers... most of them really eye opening especially the pinhole idea is amazing :D How comes I didn't think about it? That will give me the most exact starting point for standards... :)

 

I only assumed the f1.0 because I needed a starting point... I could assume f1.4 or f0.8 but since I can't really measure it seemed to me the easiest to go for f1.0, which I am aware is nowhere near the real value. I just really like to categorize everything (ask my wife about it... she will complain for 2 days non-stop! :D) like "this lens is 2.3EV UV performer and this one is 5.8EV... I understand that instead of EV I could use "potato" :) Sorry for not being 100% serious and academic... but hey, life is (or should be) about having lots of laughs and joy isn't it? :) And UV photography is definitely brings a lot of excitement and joy to one's life. If I annoyed someone with my questions / brainstorm please forgive me :D

 

A fused silica meniscus lens cannot be too expensive

I am looking into this one.. maybe it's time for me to study how to make a camera lens :) Sounds really exciting... I already disassembled a few lenses (mostly for cleaning) and some of them seem to be very simple design (especially the old russian ones).

Link to comment

Timber, questions are never annoying. We all learn from them. Sometimes trying to answer someone's question helps me to better figure out things.

Trying to classify the UV-capability of lenses and cameras is something we all try to get a handle on. You would probably enjoy learning how to perform lens transmission measurements with a spectrometer (and monochromator and integrating sphere). Wish such equipment were not so expensive. Be sure to take a look at our Lens Sticky to see what we have listed so far.

Link to comment

Without an integrating sphere for a spectrophotometer, we have to fall back on using narrow-angle 'optics' (whether refractive or pinhole type) for comparisons. Putting a pinhole aperture of 1.1 mm diameter (drilled hole in lens cap for example) on a PN-11 or similar will make an f/90 pin lens with equivalent 100 mm focal length. This is well away from the optimalised aperture so don't expect any kind of sharpness, but that is not the point here. We simply need something that has a narrow field of view and blurs the subject.

 

We now have a 'reference system' in relative terms. Make an exposure with the pinhole 'lens' + the UV bandpass filter, then set your candidate lens + UV bandpass filter to give an exposure with similar response in terms of pixel brightness. You should not focus your lens perfectly just slightly defocused. The difference in EV indicates the relative response of your lens (+ camera).

Link to comment

You should not focus your lens perfectly just slightly defocused.

 

For some reason I can't quite figure out why to leave it slightly defocused?

Link to comment
enricosavazzi

To make it respond "similar" to the pinhole version of course.

Exactly, it makes it easier to compare the two images. We cannot force the pinhole to be sharp (at apertures similar to those available on a lens), but we can force the lens to be fuzzy.

Link to comment

Please sign in to comment

You will be able to leave a comment after signing in



Sign In Now
×
×
  • Create New...