Lens Field Test [F-mount, Infinity, Vis/IR/UV]

[Andrea B.]

Purpose of Test: To verify that each lens can shoot "at infinity" in Vis/IR/UV and can record in those wavelengths.

 

Subject: Landscape of Trees and Vines

Target: Each lens was focused on the distant white sign on the tree trunk.

Location: Huber Woods Park, Middletown, New Jersey, USA

Date: 10 September 2014

Camera: Nikon D600-broadband

Raw Converter: Photo Ninja

TIFF to JPG Converter: Photo Mechanic

 

Visible White Balance: D600 Profile made from Color Checker Passport in Photo Ninja.

IR White Balance: Photo Ninja Color Correction dropper dragged over photo diagonal.

UV White Balance: Photo Ninja Color Correction dropper dragged over central tree trunk with sign. I do have UV profiles, but wanted to apply the same white balance to each photo and then exaggerate the resulting false colour in order to check for center/edge differences.

 

Editing: Typical edits were made as needed in Photo Ninja to adjust exposure, open shadows, control highlights and bring out detail. Saturation was purposely boosted in the UV photos to show differences in how each lens rendered the scene and to exaggerate center/edge differences due to vignetting or lens construction.

 

UV-capability: The sunlight did vary somewhat during the shoot because of some light clouds, so please do not read too much into the shooting times.

At f/11 and ISO 400 the UV speeds are:

 

Omegar 75/4.5 = 1/4"

EL-Nikkor 80/5.6 = 1/2"

UV-Nikkor 105/4.5 = 1/5" (fastest, but not by much)

Lentar 135/3.5 = 1.6" (very much the slowest)

 

Of the F-mount 35s, the Noflexars are faster in UV. No surprise as they have been long known to be good for UV since Klaus Schmitt's early tests. Their comparative speeds in the Vis/IR/UV shots vary slightly only due to changing light conditions.

 

Noflexar-1 35/3.5 = 1/4"

Noflexar-2 35/3.5 = 1/3"

Soligor 35/3.5 = 1/2.5"

Vivitar 35/3.5 = 1/2"

 

Lens: Rodenstock Omegar 75mm f/4.5

This is an enlarging lens.

Mount: M39.

Adapter: M39/M42.

Adapter/Helicoid: M42/F, 13.5-20.5mm.

FFD: approximately 60nm.

Front Filter Size: The front of the lens has no filter threads, so filters were held in place while shooting.

 

Visible Light [f/11 for 1/100" @ ISO 100 in Sunlight with Baader UVIR-Block Filter]

 

Infrared Light [f/11 for 1/50" @ ISO 100 in Sunlight with B+W 093 IR-Pass Filter]

 

Ultraviolet Light [f/11 for 1/4" @ ISO 400 in Sunlight with Baader UV-Pass Filter]

 

 

Lens: EL-Nikkor 80mm f/5.6 (newer version)

This is the newer style EL-Nikkor enlarging lens which has an aperture window. I have not made a permanent modification to the lens to prevent light leaks. Covering the aperture window with a folded microfiber cloth did not fully prevent light leak, so all photos have a central hot spot. The hot spot is most noticeable in the UV photo.

Mount: M39.

Adapter: M39/M42.

Adapter/Helicoid: M42/F, 20-41mm.

FFD: 70mm.

Front Filter Size: 40.5mm

 

Visible Light [f/11 for 1/125" @ ISO 100 in Sunlight with Baader UVIR-Block Filter]

 

Infrared Light [f/11 for 1/50" @ ISO 100 in Sunlight with B+W 093 IR-Pass Filter]

 

Ultraviolet Light [f/11 for 1/2" @ ISO 400 in Sunlight with Baader UV-Pass Filter]

 

 

Lens: Nikon UV-Nikkor 105mm f/4.5

Mount: Nikon F.

FFD: 46.50mm.

Front Filter Size: 52mm.

 

Visible Light [f/11 for 1/125" @ ISO 100 in Sunlight with Baader UVIR-Block Filter]

 

Infrared Light [f/11 for 1/80" @ ISO 100 in Sunlight with B+W 093 IR-Pass Filter]

 

Ultraviolet Light [f/11 for 1/5" @ ISO 400 in Sunlight with Baader UV-Pass Filter]

 

 

Lens: Lentar 135mm f/3.5

Mount: T2.

Adapter: Fotodiox T2/F.

FFD: 55mm.

Front Filter Size: 49mm.

 

Visible Light [f/11 for 1/60" @ ISO 100 in Sunlight with Baader UVIR-Block Filter]

 

Infrared Light [f/11 for 1/80" @ ISO 100 in Sunlight with B+W 093 IR-Pass Filter]

 

Ultraviolet Light [f/11 for 1.3" @ ISO 400 in Sunlight with Baader UV-Pass Filter]

[Andrea B.]

Lens: Novoflex Noflexar 35mm f/3.5 #1

Mount: Permanently adapted to Nikon F-mount. Original mount unknown.

FFD: 46.50mm.

Front Filter Size: 49mm.

 

The heavy vignetting in the Visible and UV photographs is caused by the step-rings used to mount the Baader filters.

 

Visible Light [f/11 for 1/160" @ ISO 100 in Sunlight with Baader UVIR-Block Filter]

 

Infrared Light [f/11 for 1/80" @ ISO 100 in Sunlight with B+W 093 IR-Pass Filter]

 

Ultraviolet Light [f/11 for 1/4" @ ISO 400 in Sunlight with Baader UV-Pass Filter]

 

 

Lens: Novoflex Noflexar 35mm f/3.5 #2

Mount: Nikon F.

FFD: 46.50mm.

Front Filter Size: 49mm

 

The heavy vignetting in the Visible and UV photographs is caused by the step-rings used to mount the Baader filters.

 

Visible Light [f/11 for 1/125" @ ISO 100 in Sunlight with Baader UVIR-Block Filter]

 

Infrared Light [f/11 for 1/80" @ ISO 100 in Sunlight with B+W 093 IR-Pass Filter]

 

Ultraviolet Light [f/11 for 1/3" @ ISO 400 in Sunlight with Baader UV-Pass Filter]

 

 

Lens: Soligor Wide 35mm f/3.5

Mount: T2.

Adapter: Fotodiox T2/F.

FFD: 55mm.

Front Filter Size: 43mm.

 

Visible Light [f/11 for 1/160" @ ISO 100 in Sunlight with Baader UVIR-Block Filter]

 

Infrared Light [f/11 for 1/100" @ ISO 100 in Sunlight with B+W 093 IR-Pass Filter]

 

Ultraviolet Light [f/11 for 1/2.5" @ ISO 400 in Sunlight with Baader UV-Pass Filter]

 

 

Lens: Vivitar 35mm f/3.5

Mount: T2.

Adapter: Fotodiox T2/F.

FFD: 55mm.

Front Filter Size: 43mm.

 

Visible Light [f/11 for 1/160" @ ISO 100 in Sunlight with Baader UVIR-Block Filter]

 

Infrared Light [f/11 for 1/80" @ ISO 100 in Sunlight with B+W 093 IR-Pass Filter]

 

Ultraviolet Light [f/11 for 1/2" @ ISO 400 in Sunlight with Baader UV-Pass Filter]

[Andrea B.]

Observation about the UV photos: You will notice that in all the UV photos there is quite a lot of variation in the false colour between center and corners/edges. I suspect that some of this is due to vignetting caused by step-rings. But I also wonder if any of the color changes can be caused by the dichroic construction of the Baader-U OR because I did not use any lens shades during the shoot? In close work I do not see this kind of false colour variation between center and corners/edges.

[enricosavazzi]

Some general possibilities to examine and quite possibly discard (in science, you need to eliminate the obvious before venturing into new territory):

 

- Dichroic filters, as you mention, are sensitive to the incidence angle. However, the Baader U uses UV-pass glass to filter out the VIS and some of the NIR, and the dielectric coatings to reflect out the rest of the VIS and NIR. Phenomena related to incident angle should be mostly detectable in VIS and NIR (perhaps as slight leaks of UV and NIR through the coatings?). It is possible that some angle-related phenomena also take place in the UV, but I would expect mainly at the upper and lower extremities of the UV-transmission peak (which might affect the false-color balance).

 

- Comparable phenomena may take place at every surface of optical elements inside the lens. If the lens has coatings not optimized for UV transmission, or perhaps designed to reflect UV, this may add up to a significant angle-related and wavelength-related effect. This depends also on the lens design, and may be more likely in wide-angle lenses. It should also be apparent in macro photography, but to a lesser extent (the lens moves away from the sensor and its cone of view becomes narrower, or in other words, the image circle grows but only its center is used by the sensor).

 

- As we know, atmospheric haze affects UV more than VIS and NIR. With a mostly flat subject perpendicular to the lens axis, the distance between camera and subject at the edges of the field of view is higher than at its center, therefore more haze near the edges and corners. This should be visible with wideangle lenses, not telephoto ones. It may also affect shorter UV wavelength more than long ones, thereby altering the false-color balance at the edges.

 

- Light from points located at the edge of the field of view may cross the optical elements at a lower incident angle than light from the center, and in this way pass through a higher thickness of glass. If the glass is not optimized for UV transmission, this might add up to a selective absorption. This also happens in macro photography, but to a slightly lesser extent as mentioned above. The same is true of ionic glass UV-pass filters, like the base material of the Baader U.

[Andrea B.]

Enrico, thank you so much for this excellent analysis. I am learning about optics "as I go". While I have a math background, I certainly don't have much background by way of practical physics - certainly not optics, in particular. I'm so happy now that I went out and performed this test because I've really learned from it.

 

To summarize: Center/edge differences in UV photos...

• are probably[added later] not due to the use of a Baader-U.
  
• are angle-related and/or wavelength-dependent phenomena.
  
• may be more apparent in wide angle lenses.
  
• may be caused by the interplay between these possible factors:
  
  • lens coatings affecting UV transmission.
    
  • lens design (flat field vs curved, for example).
    
  • distance from subject varying between center/edge affects shorter UV rays more.
    
  • UV light ray entry angle in non-UV-optimized glass.
    

Please let me know if that summary is adequate. I might like to put it into the Lens Sticky (with attribution of course).

 

In reviewing the UV photographs above, it is readily apparent that the UV-Nikkor produces the landscape photo with the least amount of center/edge differences.

 

I also note that some of the other UV photos could be quite useable if judiciously cropped.

[JCDowdy]

Andrea,

I am enjoying this posting, thank you!

 

I find the images taken with the Nikon UV-Nikkor 105mm f/4.5 most interesting and informative. This is primarily because the UV-Nikkor, long a flagship UV lens, eliminates most of the variables, to the extent likely possible, that Enrico lists. A control lens as you doubtless intended. Interesting then that the UV image with this lens seems to still show some variation in false UV-color towards the center.

 

Comparison with your corresponding Vis image shows this appears to correlate with areas where the same foliage (Kudzu)is in direct sunlight versus shadow and is not limited to the center of the lens. This is consistent with a wavelength related phenomenon independent of angle or center/edge effects. This may be an observation of wavelength dependent Rayleigh scattering, where shorter wavelengths are scattered more. This is an environmental effect independent of any lens artifact.

 

UV reflecting from the sunlit foliage is blue/gray where that from shadow has a yellow/green tint. A rule of thumb is that up to ~50% of the ambient UV in an unshadowed environment is scatter. UV reflected out of a shadowed area is essentially all scatter with a higher proportion of shorter wavelengths. There is a considerable amount of shadow in the sunlit Kudzu especially as compared to the horizontal brighter bands of lush grass appearing bluer in the foreground and across the base of the Kudzu.

 

If this is indeed an environmental scattering observation I am a impressed that it is observable within the transmission bandwidth of the filter.

 

​Fascinating

[Andrea B.]

Interesting then that the UV image with this lens seems to still show some variation in false UV-color towards the center.

 

Yes, it does. "-)

I'm thinking that I recall that the UV-Nikkor is meant more for close work as a "specialist" lens. As was noted by Enrico above, there likely would be less of the center/edge differences seen in close work, particularly so for the UV-Nikkor.

 

UV reflecting from the sunlit foliage is blue/gray where that from shadow has a yellow/green tint.

 

John, your observations about the scattering effects also serve to make the point that we need to be darn careful about how we interpret false colours.

 

******

 

Andrea must remember these facts:

• Wavelength dependent Rayleigh scattering => shorter wavelengths are scattered more.
  
• Up to ~50% of the ambient UV in an unshadowed environment is scatter.
  
• UV reflected out of a shadowed area is essentially all scatter with a higher proportion of shorter wavelengths.
  

******

 

Kudzu? No it is something else. Although we do have some kudzu in NJ, our winters kill off most of it. One vine pest here is called Mile-a-Minute. But I'll have to check what is in my scene. It might be porcelainberry.

[JCDowdy]

Mile-a-Minute and not Kudzu? Well, down south Kudzu has little competition!

[Alaun]

Some write faster than I can do:

 

my thoughts:

 

With reference to the UV-picture taken with the UV-Nikkor 105,

 

I wonder, if the white balance is fully correctly done in PN (or whether it is possible at all):

 

And this is a question, I got with many of my own pictures as well.

 

Taking a closer loock at the center of the picture and there at the areas in bright sun light and the areas within the shadows, it seems to me, that they show a slightly different color cast, a little bit into the yellow green in the darker areas and a little bit into the blue within the brighter areas.

 

I have seen similar differences in the color e.g. on a ball with a uniform surface, which is lit from one side and which reflects different amount of that light due to the curved surface, thus showing brighter and darker areas.

 

A correct WB would have to take this non-linearity into account somehow (I think, with some software/raw-developers, you can run the WB for differently lit areas).

 

　

 

Though I tried to avoid this color change with amount of light ("different grey scale") in post #28 here

 

http://www.ultravioletphotography.com/content/index.php?/topic/1009-cameras-for-uv-pros-cons-of-various-choices/page__st__20

 

 

I think it is still visible.

[JCDowdy]

Werner,

I had forgotten about your sphere, I must now find one in Virgin PTFE!

The sphere is a fine example, with the brighter aspect reflecting a higher proportion of the direct source.

Like the dark side of a crescent moon visible in earthshine, the shaded aspect of the sphere is reflecting scattered wavelengths.

http://upload.wikimedia.org/wikipedia/commons/3/30/Leonardo-Earthshine.png

Leonardo da Vinci, Codex Leicester, 1510

[enricosavazzi]

[...]

To summarize: Center/edge differences in UV photos...

• are not due to the use of a Baader-U.
  
• are angle-related and/or wavelength-dependent phenomena.
  
• may be more apparent in wide angle lenses.
  
• may be caused by the interplay between these possible factors:
  
  • lens coatings affecting UV transmission.
    
  • lens design (flat field vs curved, for example).
    
  • distance from subject varying between center/edge affects shorter UV rays more.
    
  • UV light ray entry angle in non-UV-optimized glass.
    

Please let me know if that summary is adequate. I might like to put it into the Lens Sticky (with attribution of course).

 

In reviewing the UV photographs above, it is readily apparent that the UV-Nikkor produces the landscape photo with the least amount of center/edge differences.

 

I also note that some of the other UV photos could be quite useable if judiciously cropped.

There is nothing certain at this point, but I would say the following:

About the first two points in your summary, we don't know yet. Making a similar test with a different broadband UV filter could be an idea. There is no need to re-test all lenses, only one or two that give the false-color shifts between center and periphery.

The remaining points in your summary should be OK. They remain possibilities, not proven causes.

[JCDowdy]

With as much as half the outdoors ambient UV being scattered, I also think your mention of no lens shade should be considered.

The EL-Nikkor is clearly UV hot-spotting and there is maybe some smokey central veiling in the Vivitar, Soligor 35 and possibly Novoflex#2. Correct me if I am wrong, but don't all of your bullet points apply equally well to simple UV lens flare?

 

I have recently been fighting flare on one of my new lenses and consequently am trying lens shades/hoods on all my lenses. Given that even the premier Coastal 60 needs a UV hood, I am coming to the belief that in the UV all lenses benefit from as much lens shade as they can tolerate.

 

It would be an easy enough hypothesis to test. It would be very interesting to see how much of the color variation remained with your control UV-Nikkor ± a deep lens shade.

[Andrea B.]

Enrico, I amended the summary to reflect possibility rather than certainty.

 

Werner, I use both white balance and camera profiling in Photo Ninja so results are as accurate as I am able to get.

However, remember that I strongly pushed the saturation just to emphasize the center/edge differences. These might not be quite as noticeable without that sat push.

 

JD, lens hoods are always a good idea. It would be an interesting adventure to try to collect lens hoods for these lenses and to figure out how to mate hoods with step-rings and filters. I may have the lens hood for the UV-Nikkor somewhere.

[Shane]

"Dichroic filters" are typically comprised of an interference coating over an absorption substrate. The absorption glass broadly defines the bandpass range and the interference (dichroic) coating fine tunes the cut-off through reflection.

 

Variation of incidence angles will result in varying pathlengths through the absorption substrate but this typically has a minimal effect compared to the effect when interacting with the interference coating.

 

The result of increasing incidence angle with any dichroic type filter is a cut-off shift to shorter wavelengths likely resulting in an increase in IR leakage for the Baader U2.

[enricosavazzi]

[...]

It would be an interesting adventure to try to collect lens hoods for these lenses and to figure out how to mate hoods with step-rings and filters. I may have the lens hood for the UV-Nikkor somewhere.

This is how I solved the problem of matching lenses, filters and lens shades in a modular way:

http://savazzi.net/photography/mounting_filters.html

[nfoto]

Andrea: the native lens hood for the UV-Nikkor is HS-14. HS-4, HN-7, or HN-8 are alternatives.

[nfoto]

My experience is that rear-mounted filters, or even better, internal filters over the sensor, may reduce the unevenness in UV rendition exhibited by Andrea's test shots. A further advantage of course is that almost any lens can be tried out for its UV response.

 

Currently I have two such systems, a Panasonic GH-2 using a rear-mounted Baader U2" (in the lens adapter), and a Nikon D3200 (internal Baader 2" filter; formerly placed in a D40X). Both arrangements allow for occasional hand-hold UV captures.

 

Here is a superwide capture of Caltha palustris (16 mm f/3.5 Nikkor Fisheye on D3200).

 

 

A hand-held capture with the GH-2 and Coastal 60 APO.

 

 

Again using the GH-2, now with a Noflexar 35 mm f/3.5 set to f/5.6. Exposure time 1/5 sec so I used a monopod to get additional camera support.

 

 

I was, and still am, a little puzzled by the apparent sharpness of the distant background in these captures. For the Caltha shot, the Fisheye was set to its nearest focusing distance (0.3 m) and I stopped down to f/22. Under these conditions, you would not expect much in terms of background sharpness even for such small apertures as f/22. The GH-2/Coastal 60 was shot at f/5.6 and the way the mountains appeared came as a surprise. Our expectations are of course determined by experiences in the visible range.

[enricosavazzi]

[...]

Under these conditions, you would not expect much in terms of background sharpness even for such small apertures as f/22. The GH-2/Coastal 60 was shot at f/5.6 and the way the mountains appeared came as a surprise. Our expectations are of course determined by experiences in the visible range.

Because of the shorter wavelength in the UV, the effects of diffraction are proportionally reduced. As a result, you get sharp images in the UV at small apertures that already give you a fuzzy image in the VIS. Atmospheric haze reduces contrast, but does not substantially affect image resolution. You could probably shoot at f/32 with 365 nm radiation and still get a very sharp image. The behavior of UV radiation in terms of geometric optics is instead the same as for VIS, so to get a sharp image the subject must still be within the parfocal range of distances.

 

This is why UV microscopes produce a substantially higher resolution than visible light ones at the same NA and magnification. This is also why vacuum-UV is currently used to produce LSI circuits.

[Andrea B.]

The UV-Nikkor manual seems to offer no insights as to recommended shade hood:

http://www.macrolenses.de/bilder/test/UVNikkorInstructions.pdf

 

I found that the hood I had stored with the UV-Nikkor was an HN-3.

But now I'm wondering what lens that hood really belongs to. :D

I will look for the above recommended HS-14 to see if I have one.

 

*****

 

QUESTION: Does rear-mounting of filters affect the flange focal distance? Seems like it would. So how do you still get an infinity focus from the UV-Nikkor on the Lumix GH-2? Or maybe you don't ??

[nfoto]

Rear-mounting alters the register distance by approx. 1/3 of the filter thickness. No problem getting infinity focus either with the UV 105, Coastal 60, or even the various 35/3.5 lenses in F mount, though.

 

HN-3 belongs to the shorter Nikkors such as 35/1.4, 35/2, 55/3.5, or 43-86/3.5.

[enricosavazzi]

Rear-mounting alters the register distance by approx. 1/3 of the filter thickness. No problem getting infinity focus either with the UV 105, Coastal 60, or even the various 35/3.5 lenses in F mount, though.

[...]

A filter mounted at the rear of the lens slightly decreases the optical distance between lens and sensor, so it is easy to compensate for this by focusing the lens to a distance slightly closer than infinity (i.e., moving the lens out). The problem with lack of infinity focus occurs only when the optical distance is increased, like when removing the filter pack without replacing it.

[Andrea B.]

I've added a few minor updates to the report.

[Andrea B.]

From Enrico Savazzi.

 

Center/edge differences in UV photos...

• are probably not due to the use of a Baader-U (because of its substrate).
  
• are probably angle-related and/or wavelength-dependent phenomena.
  
• may be more apparent in wide angle lenses.
  
• may be caused by the interplay between these possible factors:
  
  • lens coatings affecting UV transmission.
    
  • lens design (flat field vs curved, for example).
    
  • distance from subject varying between center/edge affects shorter UV rays more.
    
  • UV light ray entry angle in non-UV-optimized glass.
    

Additional factors from JC Dowdy.

• Wavelength dependent Rayleigh scattering => shorter wavelengths are scattered more.
  
• Up to ~50% of the ambient UV in an UNshadowed environment is scatter.
  
• UV reflected out of a shadowed area is essentially all scatter with a higher proportion of shorter wavelengths.
  

From Werner.

• possible linear white balance failure in sunlight vs shadowed areas. Need non-linear WB approach.
  

 

---

Collected all summary points. 10 June 2021.

---

 

 

Note to self. I have seen WB failures against shadowed Spectralon vs. unshadowed Spectralon.

Flower on front part of tilted Spectralon rectangle. Onboard or SB-14 flash. Back portion of Spectralon in slight shadow gives incorrect WB. You think "oh, it's Spectralon. I can white-click anywhere." Well, no, duh!! Remember the cyan fleabane you got from just that? And it took you about half a day to figure out why you were getting cyan fleabane! That Spectralon didn't look shadowed. But it was. La!