28mm and shorter lens for UVA - what and why?

[diant]

After some 2-3 days of thorough studying UVP pages I've obtained a some more or less clear view about 35mm lens, suitable for UVA imaging. There are a good many (well to very well) UV-performers in 35mm class. But 28mm (or shorter) class remains in the dark.

 

Enna München 28mm f/3.5 Lithagon - as we can see

here - is far from 3,5/35mm Soligor&Co; from 3,5/35mm Prinz Galaxy&Co; from 3,5/35mm Enna München Ennalyt&Co etc in its UV-capability.

 

For example Enna München 28mm f/3.5 Lithagon has only 13% T_350nm, while 3,5/35mm Soligor has there 50% and even more (if we look at KA-Soligor).

 

My question is the next: Hasn't it been discovered till now (2021) any UV-capable lens in class 28mm or shorter, which can compete with our best 35mm old-fashion UV-performers in range of "Normal" (non-dedicated) lens?

[enricosavazzi]

That depends in part on which sensor size you intend to use. For small sensor formats like GoPro and webcams, and sometimes 4/3 and Micro 4/3, quite a few legacy wideangle cine lenses are usable, although most of them are limited to about 370-400 nm. For larger sensors there are very few, if any, good choices. The Nikon AI Nikkor 24 mm for full frame has been discussed on this site as well as on my web site (http://savazzi.net/p...i-nikkor24.html). It is better than the Lithagons but not a direct competitor with the various 35 mm f/3.5 lenses in terms of UV transmission, though. Other lenses have been discussed here, but I have no first-hand experience with them.

 

I did waste some money on Enna Lithagons and similar lenses (http://savazzi.net/p...naLithagon.html), but with little to show in the matter of results. The 28 mm produces some false color, the 24 mm none. It is not difficult to find lenses that allow imaging at 385 nm and longer wavelengths, but they cannot produce UV false color other than a monochromatic image.

 

Quite a long list of lenses is mentioned in the UV lens sticky, but most of the quotations lack a quantitative measurement of UV transmission (with a spectrometer) and the qualitative definition of "acceptable performance" is very "elastic". Therefore, they should be taken with a grain of salt. E.g., try to find, or ask for, actual UV image samples taken in sunlight with these lenses, before you spend your money on these. Test images shot with 365 nm LED torches are monochromatic in any case, and not useful to judge false-color performance.

[dabateman]

Sensor size is critical factor.

Also how much reach do you need?

 

The Tamron 17mm adaptall lens with inbuilt filters is surprisingly good for UVA. It covers a full 135 format sensor, discussed in this thread:

https://www.ultravioletphotography.com/content/index.php/topic/4521-the-quest-for-ever-wider-breaking-the-20-mm-barrier/page__view__findpost__p__44606

 

A 67mm to 77mm step up ring fits it and I have used 77mm ZWB2 with BG39 2mm thick filters.

 

My APS-C Sigma 10-20mm f4/5.6 and Sigma 17-70mm f2.8/4 both in Canon EF mount work in UVA with cut offs in 370s.

 

As for C-mount lenses I scanned and talked about a bunch here:

https://www.ultravioletphotography.com/content/index.php/topic/4099-spectral-scan-of-some-c-mount-lenses/page__view__findpost__p__38683

 

 

These tests are promising with other 135 format lenses:

https://www.ultravioletphotography.com/content/index.php/topic/4483-uwa-for-uv-a/page__view__findpost__p__44134

[ulf]

There are technical design reasons for this lack of wider than 35mm lenses that are good for UV.

 

I assume we are talking about lenses for a 24 x 36 mm format and look for lenses as good as the Kuri/ soligor 35/3.5.

There are several wide lenses with marginal UV transmission that David mention above.

 

35mm is about as wide you get before the design gets more complicated and demanding.

Shorter focal length need a retro focus design with more lens elements.

Most if not all accidentally discovered lenses that reach deeper into UV-A are old, designed around 1960 ± a decade.

 

They are also rather slow, allowing for few and thin lens elements.

Faster lenses from that period are worse in UV-performance.

The ones we can use for UV were the cheap low end lenses that was built with optical glass types that has a good UV-transmission.

The AR-coating if present was not that advanced and allowed more UV to pass.

Modern AR-coatings normally cut UV more efficiently.

That is not by intentional design, but a result of making a better AR for the visual band.

 

To make a reasonably good wider lens demands more lens elements to handle all design challenges.

Some of those elements must also be thicker. => More glass for the light to pass and more AR-coated surfaces that limit the UV.

 

Modern lenses also use more optical glass types that has not so good UV-transmission.

 

I am not optimistic about that we ever will find a full frame-lens, 28mm or wider, with a similar UV-transmission as the good 35/3.5-lenses.

I think there are none to be found.

[nfoto]

"Why" is the easiest part of the question. "which" lens is very difficult to answer with any certainty of success. The wider the angle of view the less UV you get is a good rule of thumb, but of course there are some exceptions if you accept the less-than-perfect qualities of these candidates.

 

There are a number of pretty similar-looking lenses around 19-20-21mm with max.aperture approx. f/3.5-4.5. 'They tend to have a very large, curved front element and a tiny rear section. The large front size prevents many filter alternatives, however one can often manage to implement rear filtration either just behind the lens itself or in an adapter (mirrorless cameras). I have had quite good results from Panagor 21mm f/4 and Tamron 21mm f/4.5. I'm sure other members of this lens group can do service as well. You will not get into deep UV, but flower signatures can clearly be appreciated and some of these lens will do focus to infinity as well.

 

The SUNEX 5.6mm f/5.6 is a fisheye-lens that has been used by e.g. Ulf and myself to get some good UV (and IR) footage.

[diant]

Enrico, first of all I must thank you very much for your site and your work. Some day ago I faced here with a link on your site and so it was discovered by me with a great pleasure and benefit for my present explorations.

 

I'm personally not maintaining the aim of full frame coverage. I spoke yet that I'm using 1" sensors and even smaller.

Nikon AI Nikkor 24 mm has " 9 elements in 7 groups" (I quote your site) and so it's doubtful that it can be the competitor with the various 35 mm f/3.5 lenses in UVA. Two glued elements - this sounds like a verdict (in mass fabrication no one will use a special optical UV-transparent glue, and the more so there is no need in it).

 

Enna Lithagon 28/3,5 is good lens but not the best as I can judge.

If we look here, we see that Industar-69 28/2,8 is 1,5x faster in 380BP20 filter than Lithagon 28/3,5. And it almost twice faster than Lithagon 28/3,5 in 345BP25 filter.

 

So Industar-69 28/2,8 now is obvious leader in 28mm class, but it is rather inconvenient in work, even if we glue a ring adapter 30 to 49mm to its aperture ring. (It block its distance ring etc.)

 

The real lack of a quantitative measurements of UV transmission in =<28mm class is a problem, I agree.

But as concerning myself it is not so impassable obstacle for me - I can compare UVA transmission different lens with my hand-made spectrograph till around 330-340 nm (may be some shorter, with different grates, I've not yet tried all my "compendium"). I never have time for making spectrograph with reflective grates, which would allow me to widen available diapason to 300nm.

 

Dabateman, thank for your links. You've provided me a good job for today :) But I'll study they all nevertheless. Even if my head bang and ring.

 

Also how much reach do you need?

340-350nm is my little aim, 320-330nm - a great. I need no more shorter rays.

 

Ulf, I agree with all your thoughts and your conclusion, sad as it is, looks as a most probable one.

But my situation give me a chance - I don't limited a full frame demand.

 

I have had quite good results from Panagor 21mm f/4 and Tamron 21mm f/4.5. I'm sure other members of this lens group can do service as well. You will not get into deep UV, but flower signatures can clearly be appreciated and some of these lens will do focus to infinity as well.

Birna (I'm not confusing the name?), if I'll not get into deep UVA with thess lenses, they have not a lot of interest for me. I want to obtain as much as possible UVA coverage in one lens, picking it from "normal" lenses class. It is impossible in FF size, but may be (I guess) possible in small size format...

[colinbm]

The old Wollensak 25mm / 1" is a C-mount lens & reasonable & covers APS-C sensor.

[ulf]

If a wider FOV is needed than possible with available lenses, sometimes it is possible to do image stitching, from a few images.

 

Many programs can do that nowadays, more or less well.

Hugin is a powerful freeware package that give very good results.

http://hugin.sourceforge.net/

For this to work it is important to have sufficient overlap.

 

For absolutely best results the camera should be rotated around a point inside the lens.

There are several panorama heads that can be used for that.

Many times it works with handheld images, but that is normally not an option for UV so from tripod to tripod with a panorama head is a shorter step.

[dabateman]

Well a 1" sensor will give you some advantages and disadvantages with lenses.

Not all the best C-mount lenses have been discovered yet. I don't rank the Wollensak 25mm f1.5 lens as a great lens. Mine doesn't reach well into UV, its really a 28mm focal length and its only really an F2 lens.

 

I think there will be some simple designed C-mount lenses that should be cheap.

 

Also the UKA (Pentax) 25mm f2.8 C-mount quartz lens is fairly cheap at round $600 to $1000 depending on the name stamped on it.

 

 

[diant]

Ulf, I know about frames stitching, but my way requires a one-shot images.

 

Dabateman (David? Is it right?), I've studied your C-mount spectrums' cut-off and I see that there is a some "mystical" cause that forces every modern C-mount lens has its cut-off point more to the right then "350nm" landmark. What the cause for it, how do you think?

 

I think there will be some simple designed C-mount lenses that should be cheap.

Yes, there are exist. I have some of them - Avenirs from them. (9 flares only).

But two days ago I tested it. Result - "350nm" landmark withstanded with Avenir 1,4/25mm.

I have also simple 4/4 Fujian 1,6/35mm (1,7/35mm too), but not tested they yet.

 

Also the UKA (Pentax) 25mm f2.8 C-mount quartz lens is fairly cheap at round $600 to $1000 depending on the name stamped on it.

Cheap, but not for me :)

However one of these days I'll probably restore my possession of two quart-fluorite lenses, I worked with some 15 years ago. UFAR-12, but they have 41mm focal distance...

[Andrea B.]

There is *lots* of good information in this topic !!! Thanks everyone for your excellent contributions.

 

Something also to mention here is the tendency for wide-angle circular discolorations (due to either vignetting or aberrations) that can occur with wide-angle lenses used with dichroic filters. And sometimes even absorptive filters have a bit of a problem over wide-angle lenses. A possible mitigation is to rear-mount the filter, IIRC? Also a circularly discolored photo can be processed as a monochrome, monotint or split-tone to get rid of the discolorations. There's an example on UVP somewhere, but I think you can mental image that processing and immediately see that this is a nice workaround for UV false color rings.

 

Something not mentioned above is the use of wide-angle enlarger lenses for UV work. Enlarger lenses as a group may not reach too far, but I've made some good UV photos with them. And note that some enlarger lenses do have a good reach into the UV. (See Ulf's measurements. El-Nikkor 105/old and 85/old come to mind.) Of course, any enlarger lens needs an added helicoid and step-rings. These are not such expensive add-ons these days. I have a 28mm enlarger lens which I haven't tried yet. I'll post about when I get some photos. (I've had trouble finding a front ring for it, but there are other ways to mount a filter.)

 

In addition to Birna's/Ulf's Sunex fisheye, I think there are some other fishie examples too.

 

*******

 

I have the Nikkor 24/2.0 AIS and 28/2.0 AIS.

Enrico, Birna, Ulf, any advice on whether those might be UV-capable??

Apologies for not knowing, but I don't.

I'd be happy to run them through a basic UV-capability test if it would be useful? (We are a bit cloudy currently, but that won't last. I'll be able to get out tomorrow for sure.) As I write this, I decided I will test these even if no request to do so, because I'm CURIOUS. :lol:

[diant]

Andrea, I agree with you, an enlarger lenses may keep more then one discovers! Some days ago I did UVA-test for some of my lenses and decided to pick up from my stores one such enlarger lens EL-Nikkor 4/40mm. And what we see? A shallow tail of UV transparency up to 340-350nm surely. Though it is not considered to be among most UV-capable EL-Nikkor lenses (as 80mm and 105mm).

 

 

Of course, I understand, that I can't judge upon these spectra along about its real curves - I can only compare lenses between each other.

For example: some about 15 years ago I passed through spectrophotometer Shimatzu many Large Format lenses, and among them measured transmission of small triplet lens T-43. So I know surely, that its cut-off point is 314 nm. But I can't see that on my present hand-made spectrograph simply because... fade out halogen lamp light intensity at UVA short margin; fade out grate transmission there; fade out CMOS sensibility there.

 

...As I write this, I decided I will test these even if no request to do so, because I'm CURIOUS. :lol:

Bright! :) To know all that try hide itself from us - is it not our main motive to go on and on and on in life

 

BTW. David, "350nm" landmark with modern (even simple) CCTV lens stands unshakable.

[nfoto]

Andrea, are you certain 'wide-single' enlarger lenses really are *wide* and not just adapted to a smaller negative format?

[Andrea B.]

Nope, not sure. Just going by how things look.

[Andy Perrin]

Halogen light is surely not a good way to test for UV cutoff?

[dabateman]

Not relevant to your 1inch sensor needs. But generally relevant to the subject title is the Sigma 19mm f2.8 smooth barrel lens in either only E-mount or m43rds. I have one and quite happy with the Autofocus and sharpness in UVA.

 

As to your 350nm question. It come down to type of glass used. Many cheap Lomo and C-mount lenses use glass type that have transmission down to 350nm and the coating are simple not to effect that much.

 

My glass extender lens for my NONS camera, which allows for full Fuji instax film coverage and the glass Lomo lens for my lomo cameras cut off there. Plastic lens typically cut off at or around 400nm. So not great for UV.

 

Than many modern lens coatings are designed for visible wavelengths and cut off UV transmission to avoid the purple blob problem. Or problems with lateral chromatic aberrations.

 

If a simple coating and a lens made from K9 or B270 or BK7 glass could be found. I am hoping one of the newer Chinese lenses from Ttartian or 7artisn or Mitakon or Loawa might fit this group. Than we would have a cheap lens with 300nm cut off.

[ulf]

I assume the quest for wider lenses is for getting a wider FOV.

For small sensors there might be some lenses that might work, but they are not among the old good lenses like the 35/3.5, with a rather deep UV-A reach.

 

A 35mm lens on a full frame sensor show a similar FOV as a 21mm lens on a APS-C sensor.

Maybe getting a converted full frame camera can be a better path to reach a wider FOV?

That is what I did.

 

Ifs and hope about coatings and glass-types, for modern lenses might come true some day, but I would not hold my breath for that.

[diant]

Halogen light is surely not a good way to test for UV cutoff?

What is a better light for this task at home, Andy?

 

David, I think that 350nm-limit for a modern lenses have its root no so in crown glasses' problems but rather in flint glasses, which now are totally replaced by Pb-As-free glasses (with Ti-Sb instead Lead and Arsenic). If we look into Schott's optical glass catalogue and follow only one parameter of modern N-flints - Lambda80/Lambda5 - and compare it with L80/L5 for old F2 or F5 flints - all things become clear. Modern MC, of course, also plays its role, but it can't push the transmission into the total zero level.

 

Ulf, I prefer to work with cooled and full PC-controlled cameras, so converted full frame ones will be inconvenient for me.

I freely use C/CS, M39/M42, Nikon, Canon etc. lenses on my cameras - it is quite comfortably. Moreover I can (and often do it) to mount on them some unthinkable lens, as the working distances of my cameras are only 12,5 and 17,5mm.

[Andy Perrin]

Ask Ulf what kind of light he uses for doing lens tests with his spectrometer in UV. I think he is using mercury bulbs but he can tell you best. Halogen is dominated by visible and infrared.

[Cadmium]

I forgot to mention the Bushnell 21mm/4.5 (about the same size/shape as the Nikon 18mm/4). Clark has one.

 

 

 

[ulf]

Ask Ulf what kind of light he uses for doing lens tests with his spectrometer in UV. I think he is using mercury bulbs but he can tell you best. Halogen is dominated by visible and infrared.

No Andy, I am not using mercury bulbs for my measurements.

They are way too uneven and spiky in their spectra.

 

I am using a special industrial illumination unit from Hamamatsu, LC8 that normally is equipped with a mercury bulb and was used for UV-curing, before high power UV-LEDs were available.

https://www.hamamats...8_TLSZ1008E.pdf

My unit is a more rare type with a short arc low noise Xenon bulb (L8253) and an internal mirror system for a wide band light output.

The type model of the full unit is L9588-04.

This is absolutely not anything for normal home usage.

 

The unit delivers rather much light, also in the UV-range.

I need that much light, as measuring absolute lens transmission properly requires an integrating sphere that give a very low output signal to the spectrometer.

 

With different sets of light conditioning filters, I can measure absolute transmission of lenses from ca 270nm to 1000nm.

The accuracy of the measurements at the 10-15nm ends of this range starts to be a bit fussy due to low signal levels and noise.

 

With a collimated beam system I can normally measure a relative transmission even down to 200nm to 210nm.

[Andy Perrin]

Why is the xenon bulb less spiky than mercury? Don't all gases make a line spectrum?

[diant]

I forgot to mention the Bushnell 21mm/4.5 (about the same size/shape as the Nikon 18mm/4). Clark has one.

The limit at 350nm here! New lens, new glasses with Ti-Sb flintes and MC coating.

 

Ulf, very seriously way! All my best wishes to you and your work.

[ulf]

Why is the xenon bulb less spiky than mercury? Don't all gases make a line spectrum?

I do not know why.

I lack the physics background for that.

There are some spikes riding on top of the even spectra.

https://en.m.wikiped...amp_profile.png

 

I have similar effects from the deuterium lamp, but they are mostly canceled by the filtering system between the Deuterium and Halogen lamps, at around 400nm

https://en.wikipedia...terium_arc_lamp

 

The Mikropack DH-2000-BAL Deuteium/Halogen light source is magnitudes weaker that the LC8 and useless for measurements with integrating spheres.

[Stefano]

Why is the xenon bulb less spiky than mercury? Don't all gases make a line spectrum?

I don't know if this is enough to explain it, but at higher pressures the lines become broader and can overlap.