PrimaLuce Nikon Z5 Full-spectrum camera review

[nfoto]

The Italian astronomy specialists, PrimaLuce Labs https://www.primalucelab.com/, recently advertised their full-spectrum modification of the new Nikon Z5 camera. UVP (Birna) has received a review sample and my findings will be reported in this thread.

 

 

A mirrorless model in the Nikon Z line-up, the Z5 has a "full-frame" FX 24 MPix sensor, calibrated ISO range 100-51.200, and dual card slots for SD cards. It is powered by the ubiquitous EN-15-class batteries common to most Nikon mirrorless models. The Z5 comes with the EN-15C version, however the older types can be used too, but maximum number of frames might be reduced. There is an option of using real-time charging via USB-C to avoid any downtime from a battery swap, or one can use the A/C mains option available.

 

Full specifications of the Z5 can be found at Nikon's web site: https://imaging.niko...mirrorless/z_5/

 

The PrimaLuce modification entails replacing the stock AA filter with their own "full-spectrum" version, plus recalibrating the camera to enable reliable AF with the native Z lenses.

 

 

The graph (from the PrimaLuce web site) shown above indicates the modified camera should be ideal for IR photography nd by extension, also for various kinds of multispectral photography. The replacement internal filter rolls off more rapidly towards 300nm than the UV aficionado would prefer. Whether this is a serious drawback remains to be seen when conditions allow in-depth UV testing, using dedicated UV lenses such as hte Coastal Optics 60mm f/4 APO and the UV-Nikkor 105mm f/4.5. At present, winter is approaching and good subjects for UV evaluation are harder to find.

 

First impressions are quite positive as I reveived a very clean camera with no sensor dirt (not always the case from better known conversion services). PrimaLuce throw in a 64GB SD-card for good measure so one can be up and shooting in a short while. They will give a 4-year camera warranty once the camera is registered on the Nital.it web site, however I could not complete this process as the site refused (?) non-Italian citizenship. I guess Nikon Nordic would honour a world-wide warranty though.

 

The camera feels and handles just like an ordinary Z5/Z6/Z7, thus is easy to set up exactly as the end user wants. Accessories such as L-brackets, GPS devices, and remote controls from other Z models all function as expected.

[nfoto]

A basic issue with modified cameras is locating lenses that do work to satisfaction and yielding sharp, contrasty images within the narrowed spectral band of interest.

 

For IR this boils down to avoiding the dreaded "hot spot" issue, which - unfortunately - is very commonly seen in modern lenses. For UV the hot spot aspect is quite uncommonly observed, however selecting a lens that renders UV well into the deeper UV range and has adequate image quality is a task far from trivial, unless the dedicated UV specialist lenses can be deployed on the camera.

 

I did a quick test to check for IR hot spot issues with a number of candidate lenses, and details can be found in the compilation below. It should be noted that more extensive use of a given lens might indicate problems not listed here, as obviously I could not test every comination of filter, light, aperture, and focal setting (for zoom optics). However treated as a first approximation the compilation should help.

 

The list will be updated periodically. last update 16 Nov 2020

 

Z native lenses:

 

24mm f/1.8 S-Line: very weak IR hot spot at f/16, gone by f/11 or wider

35mm f/1.8 S-Line: weak IR hot spot at f/16, almost gone f/11, not visible from f/8 and wider

50mm f/1.8 S-Line: virtually no IR hot spot

85mm f/1.8 S-Line: weak IR hot spot from f/8, small but very visible at f/16

 

24-70mm f/4 S-Line: hot spots at all focal settings

 

 

F-mount lenses (native Nikkor or third-party, via FTZ or other adapters):

 

16mm f/3.5 Fisheye-Nikkor: no problem

20mm f/3.5 Nikkor UD: no problem

AFS 20mm f/1.8 G: a faint spot from f/8, visible but rather weak at f/11

Panagor 21mm f/4: no hot spot, however vignetting is pronounced

24mm f/2.8 Nikkor AI: no problem

24mm f/2.8 Nikkor AIS: no problem

24mm f/2 AIS: very weak hot spot at f/11, getting more distinct on further stopping down

28mm f/3.5 AI/AIS: very weak spot at f/22, otherwise OK; no obvious vignetting

28mm f/2.8 AIS (0.2m near limit): no problem

35mm f/2.8 Nikkor AI (6/6 design small front): no problem

AFS Micro-Nikkor 40mm f/2.8 DX: weak spot around f/8, growing more visible thereafter upon stopping down

50mm f/1.2 Nikkor AI: surprisingly, no IR hot spot was seen, further testing/alternate filtration might show otherwise(?)

Micro-Nikkor 55mm f/3.5 AI: no problem

Micro-Nikkor 55mm f/2.8: small intense hot spot at smaller apertures

Coastal Optics 60mm f/4 APO: perfect

AFD Micro-Nikkor 60mm f/2.8: very weak IR hot spot at f/22-f/32, otherwise OK

AFS Micro-Nikkor 85mm f/3.5 DX: hot spot from f/8 onwards

85mm f/2 AI: no problem

85mm f/1.4 AIS: no problem

AFD 85mm f/1.4: no problem

100mm f/2.8 Laowa Ultra-Macro: visible hot spots at all apertures

UV-Nikkor 105mm f/4.5: perfect

Micro-Nikkor 105mm f/4 AIS: OK up to f/8, signs of hot spot at f/11, thereafter more visible hot spot to f/22-f/32

Micro-Nikkor 105mm f/2.8 AIS: OK up to f/5-5.6, thereafter increasingly visible hot spot

AFD Micro-Nikkor 105mm f/2.8: very weak hot spot from f/16-f/22

105mm f/2.5 Nikkor AI: no problem

AFD 105mm f/2 Nikkor DC: weak spot starting at f/4, very bad at f/16

Voigtländer 125mm f/2.5 APO-Lanthar: visaible hot spots at all aperture settings

AFD Zoom-Micro-Nikkor 70-180mm f/4.5-5.6: no problems

Micro-Nikkor 200mm f/4 AI: no problem

AFD Micro-Nikkor 200mm f/4 ED: very weak hot spot from f/11-f/16

AFS 200mm f/2 ED VR: no problem

 

AFS 24-70mm f/2.8 Nikkor: varies across zoom range and aprtures, some hot spots are seen in particular when the lens is stopped down

Zoom-Nikkor 28-45mm f/4.5 AI: very visible hot spot shaped as septagon (aperture shape)

Zoom-Nikkor 35-70mm f/3.5 AIS (62mm filter): no problem

Zoom-Nikkor 43-86mm f/3.5 AI: no problem

Nikon 75-150mm f/3.5 SE: no problem

Zoom-Nikkor 80-200mm f/4 AIS: no problem

 

 

S-mount lenses (Nikon rangefinder):

 

Voigtländer 25mm f/4 Skopar: no problem

Voigtländer 85mm f/3.5: no problem

[nfoto]

For UV I plan on using the usual work horses;

 

21mm f/4 Panagor (rear filter box)

35mm f/3.5 Kyoei

35mm f/3.5 Noflexar

Coastal Optics 60mm f/4 APO

UV-Nikkor 105mm f/4.5

 

Further in-depth testing for UV will be postponed until I can find suitable test subjects. Now we are heading deep into the dark Nordic winter.

 

For now, this nice bunch of Lilies have to suffice. As they are cultivated, I see little merit in trying to attach a species name to them.

 

Here is the overall view, using a BG40 filter to tame the excessive red the modified camera will show in visible light due to its enhanced near-infrared response. UV-Nikkor 105mm f/4.5 set to f/16, daylight, ISO 100.

 

 

In UV, the corollas tend to respond more or less equal whether the visible colour is red, yellow, or white. This is a behaviour seen by other ornamental flowers, by the way.

 

 

UV-Nikkor 105mm f/4.5 set to f/16, ISO 100, Broncolor studio flash (uncoated xenon tube), 13 frames stacked in Zerene Stacker. I had planned for more frames, but these flowers are phototropic as and thus there are just so many times one can pop the flash and flower details stay put.

 

All frames processed in Photo Ninja and at least for this set up, the dreaded "UV-induced" striping seen with Z6 is kept well at bay (PDAF structure is the underlying reason for the striping issue, but with the Z6, it mainly manifested itself with UV photography and was difficult to get rid of even with dedicated software filtration).

 

 

A 100% view of the file shows the excellent details pulled out by the UV-Nikkor lens. The nectaria at the base of the style are mildly UV reflective (and probably UV-fluorescent as well, not tested), whilst the anthers and the pollen grains absorb UV and hence then to show up dark. The hairy staminodes are highly UV reflective. The presence of conical cells on the corolla is indicated by shimmering areas in UV.

[nfoto]

Here are some examples of IR photography with the PrimaLuce Z5.

 

Panagor 21mm f/4 with B+W 093 filter (rear filter box). ISO 1600 in heavy frost fog.

 

 

 

 

This lens renders IR images silky smooth, but there is corner light fall-off that is made more pronounced under the conditions of very low scene contrast. Landscapes show this effect clearly. For shots in the near range, vignetting is less of an issue with this lens.

[nfoto]

False-colour emulated IR ('EIR', 'infrared Ektachrome') is another favourite field of photography with full-spectrum camera. The idea is to filter out UV and much or nearly all of visible light range, leaving only some green, red and infrared to build the image.

 

For this to be successful, the lens used should not have too much chromatic aberration either of the lateral or the longitudinal kind. Many zoom lenses fail this requirement when they are pushed into service for IR as well as visible light.

 

One early favourite with other full-spectrum cameras for 'EIR' is the Zoom-Nikkor 28-45mm f/4.5. It does quite well on the PrimaLuce Z5.

 

 

 

Another lens potentially of great value for 'EIR' and other multiband purposes is the AFS 200mm f/2 Nikkor VR.

 

 

 

Some lenses, such as the non-retrofocus Fisheye-Nikkors, tend to produce too much chromatic errors for 'EIR' applications. Sometimes they can work this issue notwithstanding, however, this is a slipperly slope to enter.

 

 

(10mm f/5.6 OP-Fisheye-Nikkor on DIY focusing F-Z adapter)

[nfoto]

Internal sources of "light" pollution can ruin long exposures. Such problem areas are shutter-control LEDs, amp glow, and/or light leaks through the camera body (ports?), lens mount, or the lens casing.

 

My LifePixel-modified Nikon Z6 clearly show these issues in abundance. One has to keep exposures short, preferably shorter than 5 secs, avoid high ISO (1600 max.), and perhaps more unexpectedly, avoid certain lenses, such as the Z 35mm f/1.8 S-Line.

 

Whilst the Z 35/1.8 does show a (weak) almost thread-like ghost not seen with other lenses, the PrimaLuce Z5 appears little troubled with the other issues listed above.

 

Test shots were made using a capped lens and a camera set to ISO 5000, 30 sec, no NR. The gear was completely covered with a thick blanket in a dark room. Images are run straight through Photo Ninja with all settings neutral, and thus displayed white balance is entirely different between the camera models.

 

Here are the Z6 results with the Z 35/1.8 and Z 50/1.8, respectively.

 

Z6, Z 35/1.8 S-Line

 

Z6, Z 50/1.8 S-Line

 

Both images show strong amplifier glow and in case of the Z 35, a thread-like extra ghost on the left-hand side of the frame.

 

The images from the PrimaLuce Z5 appeared much cleaner.

 

Z5, Z 35/1.8 S-Line

 

Z5, Z 50/1.8 S-Line

 

Now, if we pull up the shadows by a staggering 10 EV, we do see the same difference between the Z 35 and Z 50 native lenses as demonstrated earlier on the Z6. Thus the issue of the left-hand side ghosting is unique to the Z 35.

 

Z5, Z 35/1.8 S-Line, +10EV

 

Z5, Z 50/1.8 S-Line, +10EV

 

It is pretty obvious that the noise pattern of the Z5 is much more subdued than on the Z6, and that amplifier glow of the latter is virtually non-existing.

[nfoto]

The review is now open for user comments and suggestions

[dabateman]

Wow excellent!

Even though it cuts off the bottom end of the UV spectrum, I wouldn't change it. The IR lens issue is none existent.

 

That 13 stack flower image is incredible.

This looks to be the Z-mount camera we were hoping for.

[nfoto]

I repeated the Lily UV stacking experiment, now a bit closer, and used the PMax algorithm of Zerene Stacker instead of the previously shown DMax output. PMax can sometimes delineate detail better, but tends to make the final image more grainy. Both characteristics are seen in the image below, which is a crop at 100%.

 

 

UV-Nikkor 105mm f/4.5 with Baader U, f/16, ISO 100, Broncolor studio flash (uncoated xenon tube), 13 frames in Zerene (PMax). Again processed through Photo Ninja, and no undue striping detected.

[nfoto]

Later that night, I tried the "Nemo" UV-LED torch on these flowers. However, they turned out to be very phototropic thus movement during 30 sec exposures robbed images of critical sharpness (long exposures where required as I need to stop down the lens to f/16-f/22 to obtain adequate depth-of-field for these 3-D shaped flowers. Due to phototropism, DOF requirements proved to be moot. Oh well. That's learning in practice).

 

On one occasion, I made the movement far worse by starting to lift camera off its tripod before exposure had completed. Apart from providing a frame filled with nice wash of delicate pastel colours, the image showed how remarkably noise-free the Z5 was for that 30 sec exposure. Silky smooth rendition, no hot or stuck pixels. No striping either, despite the NEF being run through Photo Ninja which has shown itself uncannily capable of picking up PDAF-generated striping.

 

 

I actually liked the image so much that I added it to my image archives. These colours resonate deeply in my altered mindset.

[Stefano]

As the review progresses, this seems a better and better camera. I think a monochrome version would be good for UVB/UVC photography (with a different filter than the filter installed by PrimaLuce).

[nfoto]

For now, I'll continue to use the camera with the Prima Luce cover glass inside. At present when we are entering the dark Nordic winter and most light will be gone ('true Nordic noir'), there is hardly any UV present. Thus it's only towards the spring I'll be able to put the camera through its UV paces in the field.

 

I have tested various RAW converters -- the outcome is not as given and clear-cut as I had imagined beforehand. Photo Ninja (PN, v.1.39) is able to extract all of the detail, handles blown highlights very well, and in most scenes, the dreaded striping issue is hardly seen or practically gone. Residual noise can sometimes be an issue with PN, though. Colours are off as well, since PN has no profiles for the Z5. However, using a full-spectrum camera means saying goodbye to utmost colour fidelity anyway. Thus this objection is practically moot.

 

Nikon's own Capture NX-D (v. 1.64) obviously is familiar with Z5 and does produce very smooth, clean images without undue striping. However, some IR images in particular can trip it up so highlights suffer badly and end up completely blocked. NX-D is also much slower than PN and lacks the (rudimentary but highly useful) file manager of the latter. Thus the work flow with NX-D tends to be of the "import folder - make previews - use lots of resources - getting lost" kind. RawTherapee (RT, v. 5.8) sits in an intermediate position. It does not choke on Z5 NEFs at all, even the UV stuff, unlike what it did with Z6. Colours are as good as it can get without true profiling. Details are good but not entirely up to the crispness found with PN. RT also handles highlight and blown areas somewhat better than NX-D, but noise suppression is not spectacular, or at least I never managed to match the excellence of NX-D in this regard (I am open to the notion this could be my own inexperience showing up). Striping for UV images sometimes show up faintly, but the Line Filter deals adequately with that issue. RT has no built-in file manager.

 

My current work flow with PrimaLuce Z5 uses Photo Ninja, with the occasional alternate converter for images that don't do well in PN.

[bvf]

What a lovely bit of kit - I expect it costs a little bit more than my full spectrum A6000. The UV reach is not too shabby by my standards - looks like it hits 0 at about 290nm, so you can get images at 300 or possibly just below - but not in Norway in November, as you say.

 

Where is the photo of the cows?

[nfoto]

Many old lenses are surprisingly good for IR work. The old Micro-Nikkor 105mm f/4 is among them.

 

 

This is a snapshot with the Hoya R72 filter that allows some additional false colour to persist in the final outcome. It was pretty dark that evening so I had to shoot at ISO 1600 and 3200 to record anything at all.

[nfoto]

Wouldn't it be nice to use native Z Nikkors on the Z5? They provide AF which is about as accurate as on the stock camera body, albeit a tad slower if a filter is used over the lens. However, as I quickly found out, most Z lenses are troubled with various issues such as IR hotspot or colour vignetting in IR.

 

The best of the lot so far seems to be the 50mm f/1.8 S-Line Nikkor Z. Sharpness is by and large kept up in IR, and there is no hot spot in the traditional sense. However, there are colour shifts across the frame if less dense filters such as deep orange, red or 720nm class IR filters are used. For denser filters such as B+W 093 or Hoya RG1000 these shifts show up as vignetting.

 

Here is an example with the 50Z Nikkor. The scene is fresh snow on my neighbourhood which does make for a greyish IR rendition (from water content of the snow pack I'd guess). I have greatly exaggerated the colour shift by a way-off-the-charts increase in image contrast.

 

 

As the IR rendition of this scene is pretty nice otherwise, I did run the NEF through a b/w process to make the following image, which shows the Z5 is capable of excellent IR, once one invests the extra work required.

 

 

(the identical image just with a slight rectification of my ever-present 1.5 degree tilt. Seems I cannot get rid of that even using a tripod-mounted camera)

 

By the way, the white stuff is snow, not IR-bright grass ,,,,

[Andrea B.]

Wonderful photos. I got caught up in them and had to come back again to pay attention to the review part.

 

I have a question about the "banding" and some vertical anomalies in the IR photos.

Are they due to resizing and JPG conversion?

 

[otoien]

There is also visible banding in the sky in addition to the vertical boundaries (crop at 2x magnification):

 

[Edit: I now see that Andrea also refers to the horizontal banding. The problem is not present in the first very nice captures above, so one could wonder what is going on.]

[nfoto]

Error analysis on compressed jpgs is difficult. Photo Ninja - or at least the manner in which I use it - tends to make "striping" more visible, though. Even if there hardly is any. The same behaviour is seen with my monochrome NEX-5N, for which images can be rather "stripey" if I don't watch my steps. Further downstream processing will of course exacerbate these issues.

 

For the Z5, using NX-D makes for smoother renditions as already observed. However, I don't like the program and avoid it as often as possible..

[Stefano]

Can stripes be captured in a dark frame image and then subtracted? I can't see any in your (very) pushed images above. Strange.

[nfoto]

Andrea: the post correction of the "personal tilt" could lead to rounding errors in the image file. That might explain the vertical stripes you pointed out.

[nfoto]

--

Where is the photo of the cows?

 

Posted early in this thread. The cows, like all the subjects shown, are local attractions. I did them with the Panagor 21mm f/4 and B+W 093 filter.

[bvf]

Posted early in this thread. The cows, like all the subjects shown, are local attractions. I did them with the Panagor 21mm f/4 and B+W 093 filter.

 

I meant the photos of the cows taken by the photographer who was included in your picture of the cows. Looks like they would have been interesting pictures!

[Stefano]

This? It is in post #4.

[nfoto]

Ah -- my friend wasn't using the Z5 ... thus his photos are of no relevance to this review.

[Andrea B.]

I was thinking that perhaps a rotation adjustment could cause some resampling artifacts which might show up after jpegging, but wasn't sure.

 

Anyway, the Z5 certainly does seem to be more promising for UV/IR work than that a Z6 or a Z7 with their banding problem.