Nikon Z6/Z7 as a UV/IR Conversion: the Bad News

[Andy Perrin]

I don’t think LP care. I will not be using them again.

[dabateman]

With my old rangefinder lenses, getting a narrow enough helicoid would be next to impossible.

 

I didn't mean a 1mm helicoid, but the usual twice the mount ones. On M43rds there is quite a lot of space that a normal M42 or canon EF adapter can have a built in helicoid. You you collapse it down all the way to get normal infinity focus on a normal camera, then rotate it just open for you adjustment or crank it all the way for full close macro.

The Z mount should be able to offer the same with its 16 mm registration distance. I haven't seen an electronic one yet though. So may be only for manual lenses.

[nfoto]

The very wide mount presents challenges as to the light tightness of the adapter. In particular for UV, where even the factory FTZ adapter can cause grief. To be precise, the FTZ is not the culprit per se, rather the AI/AIS design of the F-mount lenses that might allow a tiny leak though the aperture ring. Using higher ISO settings exacerbates the issue. A piece of rubber band or tape solves that issue. With some helicoid I have had bigger leakage problems, though

[nfoto]

From my younger days onwards, I was fascinated by the idea of Man arriving at Mars. Would have signed up for the recent Mars Adventure had I still been young enough (in body; mindset is as it was before). I vividly remember a beautiful picture of a Martian sunset taken by one of NASA's rovers with the strange blue sky and cool distant sun.

 

Yesterday, nightfall had eerie colours and I got "my Martian experience" straight out of the Z6 with the UV-Nikkor 105/4.5 and Baader U. Hand-held.

 

 

UV photography-- fortunately -- is not just about the quest for ultimate sharpness and deep UV-band penetration.

[nfoto]

I updated firmware of the modified Z6 to 2.2 and it is better in respect to "striping" now. Not gone in UV, but the situation is markedly improved.

 

For false-colour and IR work, the camera now is quite enjoyable. Here is a recent capture done with the veteran Zoom-Nikkor 28-45mm f/4.5, a break-through as far as wide-angle zoom designs go back in the mid '70s. The lens still does serve me well for false-colour work imparting a smooth delicate transition of sharpness in front of and beyond the focused plane. No discernable hot spot either.

 

 

The camera was set to ISO 1600 to allow hand-held shooting.

[colinbm]

Interesting Birna.

[dabateman]

Beautiful image Birna.

So the Z6 is like a fine wine, starting to work with age. I guess you are glad to have kept it.

When I looked at the banding using 4channels for Birna, I saw it in 3 out of the 4. So maybe Nikon can use that to improve it.

Fir my Em1 I do see banding in deep dark 300nm UV. But its only in one of the two green channels. So you can rescue it but cutting out that green channel. I haven't figured out how to remerge the 3 color channels to get back full resolution. But you can pixel shift them for a 4 megapixel image. As in 16 megapixels down to 4, once the channels are split out using 4channels. You don't need to debayer it as the color is per pixel.

[nfoto]

With the short register distance of the Z system, many exotic lenses can be adapted. During my last trip to the Netherlands I purchased a beautiful sample of the old Panagor 21mm f/4 lens in F mount from my dealer in Delft. It is of the same vintage as my Tamron 21mm f/4.5 and the overall construction seems quite similar. However, it does focus much closer, and apparently hardly ever was used, so I had no scruples in acquiring it for 100 $.

 

These ancient wide-angle designs have a huge front diameter so 82mm filters are required, thus using them with rear-mounted filters is much more efficient and desirable. I temporarily recycled my rear-mount Baader U filter unit for the Panasonic/m43 system by adding an m43-Z adapter to my modified Z6.

 

The UV response, compared to the UV-Nikkor 105mm, is down approx. 3 stops and the extension into deeper UV is of course limited. Yet it suffices to make interesting UV photography. In the deep and darkness of a Nordic winter, there is not much UV outdoors, so actual testing at present has to rely on UV flash. A lot of flash output is of course required when a wide field is to be covered. In the example below, I used 2*Broncolor studio flash units (800Ws) at a distance of approx. 1.5m, inclined 45 degrees to the subject. This was just a quick and dirty test so I refrained from setting up proper reflectors around the subject (my bookshelf).

 

Panagor 21mm f/4 at f/11. ISO 800. Entire frame and 100% crop. This NEF was run through Capture NX-D (full frame) and Photo Ninja (100% crop), respectively.

 

 

 

Not bad for such an old lady

 

The infamous striping issue of the Z6 is barely discernable. The Panagor has some barrel distortion, but the amount is manageable and thus of little concern. I detected insignificant decentering and most probably is due to wedge effects from the rear-mounted Baader U (the filter is taken out of its cell and held in place by springs).

[nfoto]

Another interesting aspect is how various RAW converters can deal with the NEFs from the modified Z6. My usual choice, Photo Ninja, does show some of the striping, but otherwise colours and image contrast have the usual "PN" look to them and setting UV w/b is a breeze. Nikon Capture NX-D masks the vestiges of striping better, however getting the optimal UV w/b is a little less easy as there tends to be some residual cyan to be dealt with later. Or perhaps my capability in using NX-D is limited. Raw Therapee (RT 5.7 used here) can easily remove all vestiges of striping, but is really troubled in the w/b department for these UV files. Thus there is additional tweaking required to get a semblance of the w/b rendition so easily obtained by PN.

 

I availed myself of my various UV-Nikkors to shoot comparison shoots with front- and rear-mounted Baader U filters. The basic impression is that one gets indistinguishable outcome no matter what filter position is in use. I think for a wide-angle lens this would be different as front-mounted dichroic filters will show peripheral colour shift, however for the 105mm UV-Nikkor that was a non-issue. Front-mounting the Baader U implies I can use the FTZ adapter - I have one dedicated to the modified Z6 as it is shimmed to give proper focus - and as my UV-Nikkor is CPU-enabled ("chipped", a phrase grating on my ears for any field relying on filter use), full EXIF information is present. With the rear-mounted filters this at present is not possible.

 

Any way, here is the frame view seen by the UV-Nikkor. Camera position was moved an additional 2 m backwards (compared to the 21mm test series earlier), and flash output reduced by 3 EV (1/8 output).

 

UV-Nikkor 105mm f/4.5 lens, rear-mounted Baader U, 800 ISO, flash output -3EV. This is the output from processing through NX-D

 

 

The 100% crops show better the differences, if any, in output from the various RAW converters. NB: this is the same NEF file for all examples.

 

Photo Ninja 1.38b

 

 

Nikon Capture NX-D

 

 

Raw Therapee 5.7

 

 

I leave the final decision to the readers.

[colinbm]

You seem to have got yourself a nice wide UV lens with the Panagor 21mm f/4 lens....:-)

[nfoto]

You seem to have got yourself a nice wide UV lens with the Panagor 21mm f/4 lens....:-)

 

Indeed. The Panagor focuses to approx. 14 cm leaving just less than 5 cm free working distance in front. And it does infinity even in UV. I already look forward to the spring flowering season

[colinbm]

Indeed. The Panagor focuses to approx. 14 cm leaving just less than 5 cm free working distance in front. And it does infinity even in UV. I already look forward to the spring flowering season

Excellent Birna

[dabateman]

Wow,

There is a m43rds to Z mount adapter.

https://www.ebay.com/itm/M43-Z-Lens-Mount-Adapter-Ring-for-Micro-4-3-M43-MFT-Lens-to-Nikon-Z-Z7-Z6-Camera-/254356726734?_trksid=p2385738.m4383.l4275.c10

 

That is interesting. I could mount my 3D Panasonic lens onto a Z mount camera.

 

I think we were all hoping your experiments with the Z6 were more positive. As it truly is the most adaptable system. All of the stuff I have made for m43rds can be adapted to either Sony E-mount or Nikon Z-mount.

But I seem to like the Panasonic S1 more. Go figure. I almost got a S1R with lens for $2100 from B&H, but that is my new computer money. They were having some insane open box deal in Nov/Dec. Many people that did order got brand new never touched cameras like Kirk Tuck.

[nfoto]

That is the very m43-Z adapter I am using ...

 

Despite what I hear about the Panasonic S1, I am reluctant to start yet another mount/camera system. I do own a few Panasonics of the m43 kind and am not too impressed by their sensors or even more importantly, about their build quality. Perhaps the newer system is an improvement? If that be the case, good on those opting into the system.

 

As of now, my modified Z6 is -- at last -- usable. Whilst there are vestiges of the striping issue present in UV, for most cases I can do my initial processing in Photo Ninja, or NX-D, and only resort to either Raw Therapee or applying the Topaz debanding plugin for the few cases where large prints are required. That way I can utilise all the auxillaries I already have for the Z system, including excellent GPS facilities.

[nfoto]

Been archiving some recent footage made by the modified Z6 lately, and came across this capture. Again, the old Zoom-Nikkor 28-45mm f/4.5 has been used for false-colour IR, a task for which it is eminently suited. Hand-held camera, ISO 1600.

 

 

Processed by Photo Ninja 1.38b.

[nfoto]

Today I did a sequence of tests to learn the behaviour on various (moderately) shorter focal length lenses on the modified Z6. This test used he Baader U throughout, and due to inclement weather, I conducted shooting indoors, with the added benefits of having strong black coffee, loud music, and A/C mains for camera and studio flashes.

 

Some lenses could be used with rear filtrations, others required a filter attached to their front. All shots at f/11 and ISO 800. NEFs are run through Photo Ninja 1.38b and this some residual striping can be observed, which are immaterial to the test procedure.

 

F-mount lenses, rear filter: Noflexar 35mm f/3.5, Soligor 35mm f/3.5, EL-Nikkor 63mm f/3,5

M42 mount lenses, front filter: Kyoei 35mm f/3.5, Cassar 50mm f/2.8

Petri mount, front filter: Petr 35mm f/3.5 (similar to Kuribayashi apparently), Westrex 35mm f/3.5

Nikkor-S(RF) mount: W-Nikkor 3.5cm f/3.5, W-Nikkor 3.5cm f/1.8

 

I ran the 35mm lenses at a distance of 1m, which eliminated the Westrex lens as it couldn't focus this "close". Probably needs a realignment or extra shims. As its optical quality at best was average, I didn't go into that lens in detail.

 

The Kyoei showed the need for a non-leaking adapter ....

 

Kyoei 35/3.5

 

I thus found out -- the hard way -- that my DIY narrow M42-Z adapter had a light leak !! After locating the leak and fixing the issue, much better results were obtained.

 

Kyoei 35/3.5, light leak fixed

 

Despite the front-attached filter, little colour fringing due to the dichroic filter could be detected. However, the Kyoei (like the Petri) is quite susceptible to flare. Except that issue, which likely could be cured by using a lens hood, sharpness was good across the entire frame.

[nfoto]

The old workhorse Noflexar 35/3.5 had slightly softer corners and quite sharp centre. It had low propensity for flare so image contrast was good. Overall its image quality qualifies to "very good". It does render the colours a little cooler than most others.

 

 

 

The Soligor 35/3.5 lets a bit more UV in the most of the others (approx 0.4EV), but its overall quality could not match the better lenses in the 35mm class. Images were noticeably softer in the centre. It can only be regarded as having "average" image quality.

 

 

I have several Petri 35/3.5 lenses. They are a bit differently inside yet the optics seem identical across my samples. One of them flares really badly in UV, the other is much better. I used the latter for my tests. Image quality is very good. but a propensity to flare is exacerbated by the front-mounted filter.

 

[dabateman]

Today I did a sequence of tests to learn the behaviour on various (moderately) shorter focal length lenses on the modified Z6. This test used he Baader U throughout, and due to inclement weather, I conducted shooting indoors, with the added benefits of having strong black coffee, loud music, and A/C mains for camera and studio flashes.

 

Some lenses could be used with rear filtrations, others required a filter attached to their front. All shots at f/11 and ISO 800. NEFs are run through Photo Ninja 1.38b and this some residual striping can be observed, which are immaterial to the test procedure.

 

F-mount lenses, rear filter: Noflexar 35mm f/3.5, Soligor 35mm f/3.5, EL-Nikkor 63mm f/3,5

M42 mount lenses, front filter: Kyoei 35mm f/3.5, Cassar 50mm f/2.8

Petri mount, front filter: Petr 35mm f/3.5 (similar to Kuribayashi apparently), Westrex 35mm f/3.5

Nikkor-S(RF) mount: W-Nikkor 3.5cm f/3.5, W-Nikkor 3.5cm f/1.8

 

I ran the 35mm lenses at a distance of 1m, which eliminated the Westrex lens as it couldn't focus this "close". Probably needs a realignment or extra shims. As its optical quality at best was average, I didn't go into that lens in detail.

 

The Kyoei showed the need for a non-leaking adapter ....

 

Kyoei 35/3.5

 

I thus found out -- the hard way -- that my DIY narrow M42-Z adapter had a light leak !! After locating the leak and fixing the issue, much better results were obtained.

 

Kyoei 35/3.5, light leak fixed

 

Despite the front-attached filter, little colour fringing due to the dichroic filter could be detected. However, the Kyoei (like the Petri) is quite susceptible to flare. Except that issue, which likely could be cured by using a lens hood, sharpness was good across the entire frame.

 

Doesn't the first image clearly show that you have localized aurora borealis?

 

I would keep it and call it the Aurora filter. Maybe if you patent it you could sell the design off to lensbaby. Would be better then some of those glass modified filters that you could just do with Vaseline.

[nfoto]

A nice idea -- however, the leak was in the adapter not the filter.

[nfoto]

A few more follow.

 

The mid '50s W-Nikkor 3.5cm f/3.5 (35/3.5 in today's parlance) easily wins the prize for the smallest UV-capable lens in its class. A front filter was used. The colour rendition is on the slightly cooler side, like the Noflexar, but image sharpness is somewhat better and more evenly distributed across the frame. It does not flare much despite the front filtration and its vintage coating.

 

 

Overall the W-Nikkor f/3.5 version is a very good lens for UV on the Z cameras.

 

The 50mm f/2.8 Cassar is claimed to be good for UV and my test did confirm that impression. sharpness was surprisingly good despite the rather simple triplet optics and extended into the corners as well. There is a modest propensity for flare in my setup, perhaps not surprisingly as a front-mounted filter was used.

 

 

The "winner" in this informal test round was the 63mm f/3.5 EL-Nikkor, which exhibited excellent and even sharpness across the entire frame. Very low levels of flare added to the overall impression. As tested here filtration was of the rear-mounted variety, however I probably would need to put it on a longer helicoid to make it more versatile in the field and then front filtration will be required.

 

 

Perhaps I should make a direct A/B comparison to the Coastal 60/4 APO? a lens which I expect would be a notch or two higher up the quality ladder.

[Andy Perrin]

Perhaps I should make a direct A/B comparison to the Coastal 60/4 APO? a lens which I expect would be a notch or two higher up the quality ladder.

Yes please!

[nfoto]

OK, by popular demand here is the Coastal 60mm f/4 lens. Front filtration used. I had to turn down my flashes by 2 and 3 EV, respectively, in order not to blow the image exposure. Thus the lens in itself is at least 2 EV more responsive to UV than the other lenses tested so far. Do keep that in mind when comparing results.

 

 

It should be noted that the Coastal lens has its optimum performance already at f/5.6, but to ease comparison I kept the basic parameters of f/11, 800 ISO and adjusted the flash output instead.

 

Image is crisp and clear across the entire frame. All detail is rendered using every pixel the Z6 could throw at it. One simply could not ask for more from the 24 MPix sensor. Were the lens used at its optimal f/5.6 setting, it would outresolve the sensor anyway.

 

To re-iterate, flash output was -2 to -3EV compared to all previous tests. Thus the CO 60/4 behaves and responds like the UV-Nikkor 105. They both shine in a league of their own for UV applications.

 

Now, to the opposite end of the scale as it were. I have often said that any lens can do UV if sufficient UV is forced through it, but that doesn't entail results are worth the bother. An example to illustrate my point is the W-Nikkor 3,5cm f/1.8, which for ordinary visible (and IR) work is a very fine performer indeed. However, for UV it simply is abysmal. I had to turn flash output to maximum (+2 EV re the other 35mm lenses), then open the lens two additional stops to f/5.6 in order to get anything at all. Thus its UV response is down 4 EV at least, or 6-7 EV below that of the Coastal 60/UV-Nikkor. The famous razor-like sharpness of the 3.5cm/1.8 goes down the drain concomitantly.

 

 

This version of the 3.5cm (35mm in modern terminology) Nikkors is useless for UV. Period. Its slow f/3.5 sibling is very good.

[Andy Perrin]

Thanks! One day I need to buy a GOOD quartz lens. I have a bad one (the one that came with the KSS imager which is clearly uncorrected for chromatic aberration) and I have the famous EL-Nikkor 80mm/5.6 metal and a Noflexar 3.5/35.

[nfoto]

Two lenses come to mind that are really sharp in UV and have little chromatic aberrations for that use. They are Nikkor 85mm f/1.4 AIS and the AF-Nikkor 80mm f/2.8 (for F3AF originally). They don't go *far* down into UV, the first of them the least, but image quality for UV is excellent.

 

 

The image above was taken earlier today with the 80 AF and the Baader U. Shown is a 2000 yr burial mound set in a school yard. The planners hadn't considered the mound was protected by law, thus had to find a practical solution when the school was built. My guess they initially had planned on just razing it with a bulldozer.

 

The 85/1.4 AIS is the more common of the two, but even the 80 isn't that elusive if you look for it on eBay and similar places.

[Andy Perrin]

Thanks, I will keep an eye out for them.

 

I’m glad the mound was protected.

 

You should perhaps consider amending the title of this thread? It is not all bad news, clearly.