• Ultraviolet Photography
  •  

A perspective on lenses for UV photography

UV Lens
12 replies to this topic

#1 nfoto

    Former Fierce Bear of the North

  • Owner-Administrator
  • 3,254 posts
  • Location: Sørumsand, Norway

Posted 24 May 2017 - 12:13

Tinkering in the field of UV photography for some times usually means one has a lot of "UV lenses" floating around. That certainly applies to me. Yet I tend to actually use only very few of them for any ordinary UV work. For years the work horses have been Noflexar 35/3.5, Coastal Optics 60/4 APO, and the legendary UV-Nikkor; all of which except the Coastal I have several samples of to distribute the wear and tear.

Recently I purchased a funny fisheye lens, the SUNEX 5.6 mm f/5.6 (no aperture control, fix-focus but one can alter the focused zone in field by adjusting three set screws on the lens barrel). I have tested the SUNEX on Tussilago and Taraxacum and yes, it does UV fairly well and has 185 degrees field of view too. It suits my Nikon D3200 (internal Baader U) eminently.

Now, the arrival of the SUNEX and earlier a Tamron 21 mm f/4.5 made me think of making the standard 'fixed perspective shot using various focal lengths' lens makers are so fond of presenting in their pamphlets. I could in theory at least cover the 5.6-1000 mm range with the lens I own. These would be lenses all of which focus to infinity (+) in UV. The fixed perspective simply means the camera and subject positions remain the same throughout, a setup easily implemented in the field.

As the chilly spring in my part of the world gave way to a more warm May, spring flowers here caught up on the delay in seasonal development and excellent UV subjects such as Marsh Marigold (Caltha palustris) came into bloom.

The lenses I selected for the fixed perspective test are as follows;
  • SUNEX 5.6 mm f/5.6 Fisheye lens
  • Tamron 21 mm f/4.5 (early Adaptall model, rear filter removed)
  • 35 mm f/3.5 Noflexar
  • 60 mm f/4 APO Coastal Optics
  • 105 mm f/4.5 UV-Nikkor
  • 180 mm f/5.5 Tele-Megor
  • 210 mm f/9 (UV-Nikkor with 2X Petri-based TC)
  • 250 mm f/4 Nikkor-Q (as 25 cm f/4; for Nikkor S mount; used with N-F adapter to fit F-mount)
  • 300 mm f/5.5 Petri
  • 400 mm f/6.3 Petri
  • 1000 mm f/11 Reflex-Nikkor (first version, with filter wheel in which one of the filters is removed to better enable UV)
As a sanity check, I also included the Zoom-Nikkor 50-300 mm f/4.5 ED AI as it rides permanently in my car anyway.

Earlier I have used the old 18 mm f/4 Nikkor for UV shots required to have the "wide" look, but the Tamron albeit projecting a narrower view tends to perform somewhat better. Sometimes I use the old 16 mm f/3.5 Nikkor Fisheye, but the UV rendition is not very great due to the built-in filters. The 16 mm f/2.8 has filtration in the rear that can be removed, but UV response remains largely at the same modest level nonetheless. I decided to include only the SUNEX as the 'wide' lens in this experiment.

#2 nfoto

    Former Fierce Bear of the North

  • Owner-Administrator
  • 3,254 posts
  • Location: Sørumsand, Norway

Posted 24 May 2017 - 12:13

Overall views Part I. Entire frame shown.

Images taken at f/11 unless otherwise specified, ISO 400. Weather was sunny with occasional clouds, but a slight breeze required upping ISO to 400 to avoid longer exposure times than around 1/2 sec. Distance from camera to the main clusters of Caltha flowers was around 4.5 m.

Focusing was conducted with a Zacuto loupe on the LiveView of the D3200. In the case of the SUNEX Fisheye, the low overall magnification of detail made focusing difficult and in addition tiresome, since the three set screws had to be opened, a focus attempt done, then tightening the screws to reduced the otherwise bothersome tilt of the optical axis.The better focus in UV is much easier to attain when working with the lens on the more typical close-up.

All RAW files were processed using my standard work flow with Photo Ninja and UV white balanced against PTFE standards (not shown). The 100% crops are 'untouched' in the sense that neither sharpening nor noise reduction has been performed on them.

I had the 35/3.5 Noflexar with me at the time, but overlooked it in the sequence. Human error influences any experiment.

SUNEX 5.6 mm f/5.6 @f/5.6 (only available setting)
Attached Image: J1705233380.jpg

Tamron 21 mm f/4.5 (rear filter removed)
Attached Image: J1705233359.jpg

60 mm f/4 APO Coastal Optics
Attached Image: J1705233357.jpg

105 mm f/4.5 UV-Nikkor
Attached Image: J1705233361.jpg

210 mm f/9 lens (UV-Nikkor 105/4.5 + 2X TC built from Petri parts)
Attached Image: J1705233363.jpg

#3 nfoto

    Former Fierce Bear of the North

  • Owner-Administrator
  • 3,254 posts
  • Location: Sørumsand, Norway

Posted 24 May 2017 - 12:14

Overall views Part II. Entire frame shown

I had to forego the 180/5.5 Megor as the extension I had brought for it was insufficient. I either could focus it at more distant scenes, or closer, but not to the actual scene itself. I should have considered this problem in the planning stage but didn't. Human error again.

The 1000/11 Reflex-Nikkor has 8 m near limit, which was insufficient for the scene. I did add extension to it, but then the image largely was obscured by the secondary mirror so everything went black. Thus no UV images from that optic here. I do have used it for UV landscapes earlier, though.

250 mm f/4 Nikkor-Q (with NF tube). No extra extension added as this lens natively focus to approx. 3.2m. Attached Image: J1705233370.jpg

300 mm f/5.5 Petri. Some extension (K-5 ring) added to make it go to 4.5m, it stops at 8m natively.
Attached Image: J1705233375.jpg

50-300 mm f/4.5 Zoom-Nikkor ED AI @300 mm f/4.5. No added extension, it has 2.5 m as near limit.
Attached Image: J1705233393.jpg

400 mm f/6.3 Petri. Around 6 cm extension added (PN-11); its ordinary near limit is a longish 9m.
Attached Image: J1705233377.jpg

#4 nfoto

    Former Fierce Bear of the North

  • Owner-Administrator
  • 3,254 posts
  • Location: Sørumsand, Norway

Posted 24 May 2017 - 12:15

100 % crops of centre of frame to show UV performance better in detail. All lenses at f/11 unless otherwise specified.

SUNEX 5.6 mm f/5.6 @f/5.6 (its only setting)
Attached Image: J1705233380_100pct.jpg

Tamron 21 mm f/4.5 (rear filter removed)
Attached Image: J1705233359_100pct.jpg

Coastal Optics 60 mm f/4 APO (slightly offset crop as that particular shot wasn't properly focused)
Attached Image: J1705233357_100pct(corner).jpg

UV-Nikkor 105 mm f/4.5
Attached Image: J1705233361_100pct.jpg

210 mm f/9 lens (UV-Nikkor + 2X TC from Petri parts)
Attached Image: J1705233363_100pct.jpg

#5 nfoto

    Former Fierce Bear of the North

  • Owner-Administrator
  • 3,254 posts
  • Location: Sørumsand, Norway

Posted 24 May 2017 - 12:17

100 % crops of centre of frame to show UV performance better in detail. All lenses at f/11 unless otherwise specified.

250 mm f/4 Nikkor-Q (for Nikkor S mount, used with NF-tube)
Attached Image: J1705233370_100 pct.jpg

Petri 300 mm f/5.5
Attached Image: J1705233375_100pct.jpg

50-300 mm f/4.5 Zoom-Nikkor ED @300 mm f/4.5
Attached Image: J1705233393_100pct.jpg

400 mm f/6.3 Petri
Attached Image: J1705233376_100pct.jpg

#6 nfoto

    Former Fierce Bear of the North

  • Owner-Administrator
  • 3,254 posts
  • Location: Sørumsand, Norway

Posted 24 May 2017 - 13:10

Summing up

It should be fairly obvious from the examples posted earlier that a number of lenses can do adequate service for UV work. This is not a new observation by all means, yet it serves to put the UV lens question into a perspective (sic) one should keep in mind. The purpose of the UV captures will dictate what capability the lens(es) need to answer to.

That being said, there are caveats. Firstly, just observing the LiveView on the D3200 I was struck by the much higher image clarity and contrast projected by UV-specialists such as Coastal Optics 60/4 APO or the UV-Nikkor. The RAW output from these lenses also manifested themselves in warm orange-red hues compared to the red or even cold purple of the other tested optics. In particular the 50-300 ED RAW files emerged almost blue straight off Photo Ninja.

Secondly, although exposure times were largely similar* (except for the 50-300), all files apart from those from CO 60 and UV 105 required far more massaging of the data in Photo Ninja during the "UV White" step and not surprisingly were quite a lot more noisy as well. Thus the finer details are less decisively resolved. There are in addition chromatic fringes, in particular in blue, quite prominently displayed for the longer focal lenses. The f/11 selected here sufficed to keep fringing well under control for the 210 lens (UV-Nikkor + 2TC) and the 250/4 Nikkor, a little less for the 300 Petri, and the 400Petri ought to be at f/16 as expected from previous test shooting. However, fringing is first and foremost a problem for larger prints not so much for web page and similar use.

The really poorly performing lens for UV was, quite as expected, the 50-300 ED. The lens itself is pretty sharp, but its performance breaks down in UV and resolution and contrast suffer a great deal. Plus one needs several stops more overall exposure to get anything out of the lens at all. Thus it is a prime example of the usual saying that almost any optic may deliver some kind of UV image provided enough exposure time is applied, but that in itself does not guarantee the results are useful.

The top tier performers are of course CO 60 and UV-Nikkor, anything else would be hardly be imaginable. However, the Tamron 21 once again showed itself to be a very fine performer with remarkable sharpness for a non-specialist UV lens. Amongst the longer lenses, the 250/4 Nikkor came out above the 300 Petri thanks to better contrast and detail resolution, but the difference is not as big as the price difference would indicate :D in particular if one lacks the adapter from S- to F-mount.

Last among the crowd are the 210 mm combination with UV-Nikkor and 2X TC, probably due to the fact that the master lens doesn't really like any TC at all; and the 400 Petri that although quite responsive to UV as such lacks the ultimate sharpness for this kind of photography. I have to rank the SUNEX lens on the lower tier of performance as well. It does much better for UV close-ups though.

The 50-300 ED, whilst a very respectable performer for visible photography, is clearly not well suited for UV. Hardly surprising. Images are very noisy and details are blurred, plus required exposure times are several stops longer than for the other lenses in this test.

The final lesson is that all lenses, including the unlikely lens for UV 50-300 ED, did show the flowers to have the same UV signature. That is a welcome food for further thought.



*Remember, this was a field experiment with non-controlled lighting. Thus a similar exposure time for two lenses cannot be taken as equal UV transmission.

#7 Andrea B.

    Desert Dancer

  • Owner-Administrator
  • 9,103 posts
  • Location: UVP Western Division, Santa Fe, New Mexico

Posted 24 May 2017 - 16:06

This is an absolutely delightful and, of course, most useful survey of some fine lenses in use by our very own UV-meister. It is nice to see some long lenses included.
Thank you so much, Bjørn, for the great write-up.


With so many contributions from all our members, UVP has certainly built a good pool or UV-capable lens information.
Andrea G. Blum
Often found hanging out with flowers & bees.

#8 Andrea B.

    Desert Dancer

  • Owner-Administrator
  • 9,103 posts
  • Location: UVP Western Division, Santa Fe, New Mexico

Posted 24 May 2017 - 16:14

Bjørn, with reference to Post #4: would you judge the CO60 to be a bit more contrasty than the UV-Nikkor? Or is the increased CO60 contrast (and saturation) in that post I see simply a result from conversion?
Andrea G. Blum
Often found hanging out with flowers & bees.

#9 nfoto

    Former Fierce Bear of the North

  • Owner-Administrator
  • 3,254 posts
  • Location: Sørumsand, Norway

Posted 24 May 2017 - 16:40

They are more or less the same. However, the CO 60 scene included higher contrasts in itself compared to the one of the UV-Nikkor. Add small fluctuations in incoming light and real differences would vanish.

Processing was in principle the same for all these files, by the way. I copy/pasted settings from a master file.

#10 Andrea B.

    Desert Dancer

  • Owner-Administrator
  • 9,103 posts
  • Location: UVP Western Division, Santa Fe, New Mexico

Posted 24 May 2017 - 18:45

Thanks!
Andrea G. Blum
Often found hanging out with flowers & bees.

#11 OlDoinyo

    Member

  • Members(+)
  • 887 posts
  • Location: North Carolina

Posted 25 May 2017 - 04:25

How are you getting the Tamron to focus to infinity in UV? Mine uses a rear-mounted filter thicker than the original, and definitely does not do so, though it is sort-of-sharp at f/16 (but nothing like the Asahi 35/3.5, which is really sharp.) I would think even an in-camera filter would cause similar issues, since it interposes an extra thickness of material behind the lens.

Teleconverters are largely useless with modern digital sensors, which have much higher resolution than film did. A teleconverter cannot increase angular resolution versus the base lens alone; in fact, it will invariably decrease it somewhat. You are better off cropping the center of your frame and blowing it up than you are using a TC in most cases.

Edited by OlDoinyo, 25 May 2017 - 04:31.


#12 nfoto

    Former Fierce Bear of the North

  • Owner-Administrator
  • 3,254 posts
  • Location: Sørumsand, Norway

Posted 25 May 2017 - 07:21

The Tamron 21 was modified in the lens mount to be a native F lens by my close friend, Erik Lund (aka 'Dr.Lens'). Meaning the Adaptall part was removed and replaced by a factory replacement F mount after trimming off sundry parts inside. We did this to allow a CPU chip to be added, which otherwise would not be possible. The aperture linkage on the Tamron had to go, though, so it is stop-down only; but that is a small price to pay for other advantages of having the lens natively as F mount. It does indeed focus with ease to infinity and as already stated, delivers surprisingly sharp images in UV (and IR). For visible light, though, its performance is bland and undistinguished.

I don't like the 'impossible' and thus also have replaced the original mounts on the longer Petri lenses with F mounts. The 300 and 400 will focus to infinity after the surgery, the shorter lenses such as 135/3.5 and 200/4 stops before infinity is reached. I didn't modify the 35 mm variants as my Noflexars already fitted perfectly in F-mount, so the Petri-mount 35 lenses are kept for my Panasonics (I have m43-Petri adapters).

While it is true you can simulate 'reach' by cropping, other aspects of actually increasing the focal length of lens by adding a TC won't follow. The primary magnification will be increased, for example, which cropping will not accomplish. Increasing magnification helps overcome reduced optical performance. This is true for ordinary photography as well and one of the insider tips for getting more out of a Micro-Nikkor up close. It's been many years since I first came across that tip in a textbook on optics and have verified its value many times since then.

The home-made 2X "UV TC" was simply build by moving the optics etc. from a Petri TC into a short extension ring for Nikon F. I mounted the optics in a threaded inner mount so as to be able to fine-tune its infinity focus. The combination has occasionally proved itself to be useful and being neat and compact, no harm is done by putting it in the bag with one of the UV-Nikkors anyway.

The modified Nikon D3200 has a custom-cut Baader U (2nd gen.) over the sensor to replace the stock filter pack. An arrangement with which I'm very pleased as it allows a plethora of lenses to be used for UV and many to reach infinity focus. Thanks go to my friend Vivek Iyer for the filter (and the D40X originally hosting the Baader).

#13 nfoto

    Former Fierce Bear of the North

  • Owner-Administrator
  • 3,254 posts
  • Location: Sørumsand, Norway

Posted 25 May 2017 - 08:53

A final conclusion: I will hereafter use old Nikkor-Q 250 mm f/4 as my preferred long lens for UV. It certainly beats my previous "long" lens (UV-Nikkor + 2X TC) by a comfortable margin. This comes at a price of course in terms of weight as the 250 is a massive 1.3 kg (with NF tube and hood). Its tripod mount while commendable robust has a very small footprint, typical of its vintage, thus some ingenuity is called upon to make it behave properly on a tripod in particular if extension is added to its rear end.

The 300/5.5 Petri will be a backup alternative if I need to travel light.