Quest for a UV-capable wide-angle lens

[Alex H]

My current best wide-angle lens that I use for UV is the 20mm F/3.5 lens for half-frame Olympus Pen-FT camera. Since I do wish one day to use full-frame camera for UV photography, I started accumulating other lenses in similar focal range, and today was a day I tested them. Here is my current little collection. All lenses were compared against 35mm F/3.5 Kyoie. The test results do not include 20mm Vivitar because I forgot to close the aperture, and 17mm Vivitar because I cold not focus it properly. They will be re-tested later on.

 

 

Since the weather is windy, and the sky is partly covered with moving clouds, I found one poor dandelion in shaded by a tall building and shot it as a test subject. PTFE was included in the scene. ISO and aperture was set the same to all lenses. WB was set in-camera with 35mm Kyoei and PTFE disk. Metering was done by the camera using a multi-zone setting. I set the camera at a fixed distance. The images were cropped to fit the same field of view. Two versions of the same picture were produced for each lens: 1 - as shot with "Kyoei" in-camera WB; and 2 - with custom WB based on PTFE reflectivity.

 

The triptychs include (from left to right): "target lens" shot with 35mm Kyoei, "test lens" with in-camera WB, "test lens" with custom WB.

 

1) 28mm F/2.8 Auto Tamron (adapt-a-matic) first version

 

 

2) 24mm F/4 MC Macro Revuenon (made by Enna)

 

 

3) 21mm F/3.8 Auto Super-Lentar (made by Tokina)

 

 

4) 20mm F/3.5 Olympus G.Zuiko Auto-W (for Pen FT)

 

 

5) 16mm F/2.8 MC Zenitar Fisheye (without rear UV-filter)

 

 

Among the lenses shorter than 25mm focal length, little Zuiko still wins it for me!

 

To be continued...

[nfoto]

Why isn't the Tamron 28/2.8 ranked higher on your list? It seems to deliver a decent performance in UV at least judged from these shots.

 

I reckon exposure times are tricky to compare directly as you had little control of the ambient light levels. But the lenses with nearly one order of magnitude longer exposures also are those which seems to perform poor against the better alternatives.

[Alex H]

Why isn't the Tamron 28/2.8 ranked higher on your list? It seems to deliver a decent performance in UV at least judged from these shots.

 

Of course Tamron is best of all the ones tested, but I was particularly looking at lenses in 16-24mm focal range, hence my statement about Olympus - see I mentioned "lenses shorter than 25mm focal length".

 

Exposure times are there just for reference as it is indeed impossible to control ambient light.

[nfoto]

OK, got it. You really are climbing out on a limb in this quest. Wide angle and UV don't mix very well.

 

I have had great results with my 18 mm f/4 AI Nikkor, but admittedly this lens doesn't extend deep into UV. Still suffices to show UV floral patterns though.

[Andrea B.]

Cool stuff, Alex. Thankee, mon!!

I'll be looking more closely later after todays other duties are taken care of.

 

One question pops to mind. Some lenses have more contrast than others. That can be either microcontrast or simple contrast or both. (I'm not sure I know the correct terms there.) So I am wondering if the Enna or Zenitar might work if one simply increased the contrast in the camera or in the editor??

[Alex H]

Andrea, the contrast can be increased, but the write-balanced pictures are almost monochromatic, which might indicate very narrow transmission band. Low contrast, also can be present if the AR coating of the lenses does not extends its functionality to UV, causing numerous internal reflections in this range, but that is just my speculation.

[Alex H]

Bjørn, there used to be a post comparing transmission of 18mm F/4 AI Nikkor and 17mm Tokina here: http://www.fotozones.com/live/index.php/topic/33674-uv-nikkor-f418mm-uv-transmission/

But the picture is no longer available it seems. My 17mm Vivitar (claimed to be made by Tokina) is similar in its transmission to 24mm Enna, but as I could not focus it properly this time, I have no results to show. I will retest it later on.

[nfoto]

An increase in chromatic aberrations would also lead to lower contrast. This effect adds to the others mentioned by Alex.

 

The simple fact is that nearly all lenses are *not* designed to work in the UV range. That a surprising number of them do is wonderful, but one has to balance the inherent lowered performance against the envisioned usefulness of that particular lens. Thus, as wide lenses are a scarece commodity for UV shooting, ordinary performance criteria probably need to be relaxed. Judicious post processing can remedy low contrast issues and even in some cases make the image sharpness appear distinctly improved.

[nfoto]

The transmission curve of the 18 mm probably wouldn't stand up to say a 35 mm Noflexar, but in my opinion this is moot as it serves its indended purpose well enough. Demonstrated recently inb this thread http://www.ultravioletphotography.com/content/index.php/topic/1356-the-kings-oak-and-other-uv-impressions/.

[Alex H]

I just remembered that IR contamination can also be a reason for lower contrast with lenses that do not transmit much UV when used under sunlight.

[nfoto]

Very true. It's all about the actual exposure value in different bands and when UV response is too weak, any UV signature will easily be overpowered from borderline IR or NIR contamination.

[OlDoinyo]

I don't know if this is completely on-topic, but I thought I would mention this:

 

http://techlinkcenter.org/summaries/all-reflective-wide-field-view-imaging-system

 

I have also seen a few attempts at designing wide-angle reflective optics by mounting a convex mirror in front of the camera (which bears a normal lens and is thus pointed directly away from the subject matter.) The camera is inevitably in the center of the field of view with this strategy.

 

A kind of reverse-Cassegrain geometry might also be possible by pointing a sharply convex primary at the subject matter and surrounding it with an annular concave secondary to focus at the rear plane--that would be truly glassless. Wide-angle lenses tend to have thick glass by their nature, and a maximally UV-capable refractive wide angle would have to be pure quartz or some similar material. In the absence of anything this exotic, look for slower lenses--they tend to have thinner glass.

[Alex H]

The transmission curve of the 18 mm probably wouldn't stand up to say a 35 mm Noflexar, but in my opinion this is moot as it serves its indended purpose well enough. Demonstrated recently inb this thread http://www.ultraviol...v-impressions/.

 

I put 20mm Zuiko and 35mm Kyoei in this test as a reference points, since we know transmission properties for both, and there are example pictures. 18mm Nikkor or 17mm Tokina could serve as another reference point just to show what can be done with lenses that have relatively limited UV transmission.

[Alex H]

In the absence of anything this exotic, look for slower lenses--they tend to have thinner glass.

 

Thinner glass and lower UV transmission does not necessarily correlate. Macro Revuenon has seven small elements in seven groups and it performs worse than Super Lentar that has much larger and thicker nine elements (in either eight or nine groups).

 

I am looking for a lens that I can use to take pictures. Exotic catoptric and catadioptric solutions are not what I am interested in.

[Timber]

This one seems to support my theory that with the same White Balance a lens producing more blue-ish is less capable of UV and a lens producing more yellowish is more capable of UV. With controlled UV lightsource (2 UV bulbs) I had the same results / findings. Basically that's how I rank my lenses now for UV capability. Taking one that's average and then I do a Yellow / Blue scaling.

[JCDowdy]

This one seems to support my theory that with the same White Balance a lens producing more blue-ish is less capable of UV and a lens producing more yellowish is more capable of UV. With controlled UV lightsource (2 UV bulbs) I had the same results / findings. Basically that's how I rank my lenses now for UV capability. Taking one that's average and then I do a Yellow / Blue scaling.

 

This did remind me of your interesting Yellow/Blue comparison matrix. By that method of WB comparison against a known deeper transmitting lens this suggests the 28mm F/2.8 Auto Tamron may transmit deeper than the others. At least is appears to WB closer to the Kyoei.

[JCDowdy]

Alex,

 

Did you see the Soligor Wide-Auto 28mm f/2.8 for UV mentioned on another forum? Made in so many variations who can guess which one, perhaps something optically similar to your 28mm F/2.8 Auto Tamron?

[Alex H]

Yes, John, I have seen that thread long ago. Besides the fact that his UV shots are not very sharp, he does not specify which lens it is. As you correctly mentioned, there are many 28mm F/2.8 Soligor lenses made in different years by many manufacturers. Finding their optical data is too complicated, so I was never willing to spend my time on it. Besides, I am more interested in wider lenses, in 17-21mm range. 28mm on full-frame is not much different from 20mm on APS-C.

 

There are in fact two optical versions of 28mm F/2.8 Auto-Tamron in adapt-a-matic mount, and each of these two had two design varieties. I am not sure if both optical versions are 7e/7g, but I had the later version in the past and it did not perform as well as the early version that I have now. They definitely used different types of glass with different refractive index, judging by the size of the front element (much bigger in the early version).

[Alex H]

Another day, another test. Now only the widest four lenses that I have, including two missed from yesterdays comparison.

 

First, a set with "Kyoei" in-camera WB.

 

 

And now a set with custom PTFE WB and adjusted contrast and saturation.

 

 

Olympus is still the best out of four, followed by Lentar (Tokina).

Vivitar (Kino) does not seem to be sharp at all, but I should re-test it under artificial light.

[nfoto]

I put in a word for the 18 mm f/4 AI Nikkor. If you can come across sample do test it. Should our paths cross at any future time, I'll bring it with me.

[Alex H]

I would have liked to test the Nikkor and compare it with other lenses I have tested, but, unfortunately, it sells for too much, and I do not know anyone who can lend me one. There are few more ultra-wides that I want to test, but the main point is to get the full-frame camera for UV. It may take some time.

 

I think for now I may just try the 21mm Lentar for "real" UV-photography and see how it does. 17mm Vivitar will be tested for visible photography. These old ultra wide angle lenses are known for decentering problems and sample variability.

[Andrea B.]

After your edits, the fotos in Post #19 are more difficult to decide between. But, like you, I do not think we should be making strong contrast edits to try to bring out the UV-absorbing areas. :D The before photo does obviously support the Pen and the Lentar.

 

I need to try to get links to this post and other lens posts added to the Lens Sticky.

[nfoto]

Alex: Agree that collectors influence the pricing too much on certain lenses amongst which the 18 Nikkor f/4 is but one. Elevated pricing is even more pitiful when one consider the likely future fate of these lenses is to be stood in a cabinet or cupboard as a part of a lens collection.

 

Unfortunately you are located inside the EU and I (fortunately) am on the outside. That means any shipping between our neighbouring countries will involve VAT and import taxation. Otherwise I would gladly send you the lens for testing. You wouldn't happen to have connections in Oslo who could take the lens with them to Sweden?

[Alex H]

That would be interesting, Bjørn. There might be a solution. But before we do it, I would like to collect two more lenses in this range, and re-test them all at the same time under more controlled artificial light.

[JCDowdy]

Alex,

Has anyone ever looked at the Canon FL lenses for UV? The FL 19mm f3.5 & f3.5R are fascinating but perhaps to complex to promise much in UV. The 28mm f3.5 (7 elements in 7 groups) and 35mm f3.5 (6 elements in 6 groups) with no cemented groups might be better UV candidates

Flange depth of 42mm is shorter than F-mount but should be adaptable to mirrorless cameras, including the full frame Sonys.