Nikon Z5 full spectrum white balance preview

[maikm]

Hi there.

 

After using a modified D70 (720nm) for 13 years, I moved to the Z system and finally decided to also get a modified Z camera to join my regular Z9/Z6II pair. I squinted at the Primaluce Z5 for a while (and read the thread about it here), but they wouldn't come down with their price, so I took advantage of the current cashback rebates, got one new and had it modified for full spectrum by Makario. I'm aiming for IR landscape photography but still chose full spectrum to later experiment with other uses. Using an screw-on 720nm filter for now.

 

As it is widely known, the Z5 (and probably all the other Z cameras as well) won't measure custom WB from an infrared image, as it's apparently doing a "sanity check" on the resulting values and concludes there's something wrong with the measurement. Of course, I tried to trick it into accepting something anyway. Among the things I tried:

 

- putting a D70 jpg onto the card to measure from

- modifying a Z5 jpg with experimental WB values via exiftool and measuring from that

 

The D70 is not accepted, and it also seems that if the values in the image for PRE are too extreme (like >2), the camera will not reject them for PRE measurement use, but the PRE setting remains unchanged anyway. Even then I managed to get some wild results, but all are resulting in either too green or too red or too purple images.

 

I will of course set the white balance in RAW processing later, but I'd like to see a reasonable preview of the brightness values in the viewfinder. The closest I've got now is setting the value to 2500K and tint all the way to green, then set a monochrome picture profile.

 

I also contacted the Nikon support, which has been very helpful in other occasions including resulting in actual firmware changes for bug fixes, and asked for a way to set white balance off any source image for artistic freedom, infrared imaging, and colored lighting situations. I would encourage anyone "suffering" from the same problem to do this, too.

 

Did anyone have more success or try other ideas to trick a Z camera into better WB preview? Again, I'm aiming for 720nm in daylight currently.

 

Maik

[lukaszgryglicki]

Sorry I can't help, but I can confirm that your approach:

2500K and tint all the way to green, then set a monochrome picture profile

Seems to be the best I can get on all my modified Nikons.

I usually set manual WB, do max green possible and then move a bit into blue. After this is done, I also choose mono to see kinda-ok exposure in Lv.

 

[Andrea B.]

I've written a lot of this before and you probably already know most of it. But for the general viewership, I'll be complete.

First, you are on the right track with your 2500K + green tint and Monochrome setting for your Nikon Z5. That'll do you better than trying to shoot in Daylight or Tungsten with its oversaturated reds & oranges in UV. 

 

Second, to perhaps partially answer your question, I want to mention that there are ways to "push" the Nikon white balance measurement. This push will never give the same result in-camera that you can get in the converter, but you can get to a useful WB for UV that has some blue & yellowish tints along with some mild orange. I'm going to find a link for you about pushing WB and put it here:  Pushing WB to Shoot in Nikon Red. I'll try to include an example also.

 

NOTE:  The link explains pushing a WB towards Red or Blue or Green. That is not quite what you need for UV. So I need to add something to that topic!! But the point of the link explains the pushing steps which will be the same. Only the WB cursor setting will change.

 

*****

A bit of (repeated) info for those newbs to Nikons used for UVIR.

 

I've been using full spectrum Nikons now for 15+ years. They simply do not measure white balance under dark UV-pass or IR-pass filters. For what it is worth, the Sony, Pentax and Panasonic full spec conversions I have which are able to measure WB under dark filters are not always accurate. (Insofar as we can define what "accuracy" means for reflected UV or IR work.....but I digress.)

 

In recent years, I've been setting the Nikon Picture Control to Monochrome (as I note that you also do above) when shooting reflected UV, IR or UV+Blue+Green. That way I can use the brightness histogram to keep from blowing all 3 channels by keeping it off the right side. Nikons have excellent headroom and you can almost always pull back one blown channel or two blown channels with no problem in a good converter. You can even pull back a complete 3-channel blowout (non-specular) sometimes, but I try not to let the exposure get that far out. The histogram is for JPGs as we all know, but in Monochrome it is good enough to use for preventing serious blowout and preventing underexposure in UV photos. Using Monochrome to shoot & review a UV/IR frame also permits you to determine whether you have captured the details you want because they are not obscured by oversaturated red channels.

 

The Monochrome idea works for any full spectrum camera regardless of brand.

 

At the beginning of any photo shoot, I first photograph a piece of PTFE and Spectralon. That photo is white balanced in the converter and the resultant WB is saved for batch application to the photos just made. I'm so used to doing this now, that I can just fly through it. Please note that I recommend you do this with any brand of full spec conversion. You might have a brand such as Sony (or others) which can measure a WB through your UV-pass filter, but you will find that the lens, the UV-pass filter and the light can sometimes combine to produce a measurement which is not the same one you would get from white balancing (for the given lens+filter+lighting) against PTFE in the converter. White balancing false colours is tricky.

 

 

[Shane]

I have had a full spectrum Z5 for a while now but have not had time to use it yet (got married and temporarily living in Brazil). In the past I always used to set up a UWB using Jim Kassons method. For me, I have used nothing but UWB for over 15 years now for every camera I owned including factory stock cameras. This should be done prior to conversion and in my rush to convert my Z5 to full spectrum I forgot to do this and still need to set this up prior to using the camera. One possible method for a full spectrum conversion is to add a "correction" filter to the camera to make it perform similar to pre-conversion conditions and then run Jim Kassons method. I haven't had time to try this yet but I suspect it will probably work.

 

I originally purchased the Z5 full spectrum as a test camera to see if I could make it fulfill all my requirements, UV, IR, visible and night photography through the use of rear mounted drop-in filters. I have still yet to try this (a year later!) but hopefully will get to it soon. If it works, I will probably replace all my cameras with a single Z9 full spectrum camera. A single camera would help reduce the issues of traveling with multiple bodies.

[maikm]

I'm at the other end of the spectrum. Never would travel with only one body, always have at least one backup. A Z5 with filters including a UVIR-cut would both expand the shooting possibilities and provide a usable backup to my main Z9. But relying on a single body? Drop the drop-in cut filter and you're screwed. Puns intended.

 

The goal of UniWB is obviously to see actual captured data represented in RGB histogram and to ETTR, while my goal is to have a preview in the viewfinder that's closer to the edited end result, so I can better judge what I'm doing on the artistic side, and notice mistakes earlier. I'll probably have an UniWB stored in one of the PRE settings as well, but that's not what I'm aiming for here. I do have a cut filter but no access to it until next week for logistic reasons.

[photoni]

@maikm For IR over 720 nm I see no problems, is B&W
For UV I don't trust the vision on the monitor, even if it is set to BW; the best thing is to look at the histograms of the three RGB channels, and verify that they are within range, then take a second and a third underexposed shot.
With the ISO invariance of recent digital cameras at 100 iso you have a range of 5 / 6 apertures underexposed.

 

P.S. you must have a good software to view and develop RAW... what do you use?

.

[Shane]

"But relying on a single body?" ..... Actually, I always have backup body despite in 18 years of shooting digital, I've never had a camera fail and never broken one either.

 

 

 

 

[maikm]

UV is a field I have no experience with yet, as my previous mod was a D70 with fixed 720nm modification, and I have plenty of past knowledge to update with my new stuff on the IR side before I can think of UV.

 

Speaking of the D70, that had a histogram that effectively only showed a green channel. I developed the habit of always shooting 3-bracket exposures (with "motor" on continuous, which matched the D70 well because it's buffer could hold exactly 3 frames before throttling...). I'd have no problem doing the same with the Z5 at times, rather than fiddling with the histograms and switching back and forth between WB settings that allow me to either see an accurate histogram or judge my actual composition. But I'm speaking from almost zero experience with mirrorless IR at this point.

 

As a RAW converter, I use Nikon NX Studio. Capture NX-D, NX2, NX before that. Sometimes Affinity Photo (which has a problem with IR WB in V2 at the moment, but it's in the process of getting fixed). I also sometimes use Rawdigger (Profile Edition) or even dcraw if I'm investigating something. Raw Therapee, want to try 5.9 once the Mac build is done.

[Andrea B.]

IMPORTANT REMINDER:  Some Nikon Z cameras produce crazy banding in UV or IR because of PDAF. I do not have a definitive list.

 

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

 

I also do not have a nice summary of that 7+ page topic. I do recall that high ISO noise hides some of the banding. (Not a solution, IMHO.) And a denoise app of some kind might help reduce the banding. 

 

So look thru that entire article before you decide to convert a Z.

 

**********

 

But relying on a single body? Drop the drop-in cut filter and you're screwed.

 

This is why we carry back-up filters. There's a famous story floating around about Andrea B. who failed to carry a backup BaaderU out to the desert on one trip.

I've dropped every filter I've ever owned during field work. They're pretty tough. Unless.......

 

**********

 

my goal is to have a preview in the viewfinder that's closer to the edited end result, so I can better judge what I'm doing on the artistic side

 

As mentioned, this will never happen for a Nikon body whether DSLR or Mirrorless. But I am still working on the method to produce the interesting reasonably useable, pushed WB which I currently use (which is not uniWB). It all started with white balancing against a peach colored wall. But I need to reproduce that on a monitor screen.

[Shane]

I did review this article and several others before I decided to buy a new Z5 and convert it. Overall, the Z5 appeared to suffer less from issues than the Z6/Z7. However, I am very critical of images and will push the PP hard to see what might show up.

[Andrea B.]

my goal is to have a preview in the viewfinder that's closer to the edited end result, so I can better judge what I'm doing on the artistic side....

 

Here is the best I can do in the viewfinder with my Nikon D610 full spectrum conversion using a pushed white balance as shown by the second SOOC photo in each set. But to get this effect in the viewfinder you really, really, really need a LOT of UV light. Nevertheless, this white balance is useful also during review.

 

I pushed the white balance (in a certain way) and got these white balance multipliers:

• Red = .546875
• Green1 = Green2 = 1.0 (of course)
• Blue = .718750

 

A straight-out-of camera photo made with this white balance shows (in most cases) areas of yellow and blue which will become the false yellow and false blue after white balancing the raw file in a converter.

 

Here are two examples using a Nikon D610 + Zeiss UV-Planar 60/4.5 + BaaderU UV-Pass Filter + Sunlight.

 

Visible

This little pink flower is Houstonia rubra.

 

 

UV Straight-out-of-Camera (no edits)

The flower is moderately UV-absorbing and the leaves are very UV-absorbing.

Note the purple-blue petals here.

 

 

UV after White Balance in Photo Ninja

After white balancing, the false blue occurs in the same location (the petals)

where the purple-blue occurs in the preceding SOOC.

(The photo needs a bit more work than this quick conversion made just to show how my pushed WB works.)

 

 

 

Visible

This orange flower is Sphaeralcea fendleri. commonly called Globe Mallow.

 

 

UV Straight-out-of-Camera (no edits)

The petals are UV-reflecting and the center parts are UV-absorbing.

Note that the flower petals are yellowish.

 

 

UV after White Balance in Photo Ninja

The false yellow occurs in the same location (the petals) as the yellow in the preceding SOOC version.

Note that some tiny bits of false blue also popped up after the white balance.

It isn't so obvious in the SOOC version because of the tiny area.

 

 

A digression:  As much as I have loved the D610 conversion for its ease of use and excellent image quality, I note that with age the shutter is beginning to show the same spatter/splatter problem that the D600 had. Nikon probably would not even consider a repair given that the D610 has been converted. 

[Andrea B.]

One Example for Pushing the White Balance in a Full Spectrum Nikon Conversion

 

Learn this method first before worrying about ambient light, the camera's grey point,

how close you are to the screen, color shifts from a lens and other such details.

 

STEP ONE:   Neutral picture control > Magenta screen > Set exposure > Measure white balance.

• Set Nikon Picture Control to a desaturated Neutral (Sharp 0, Contrast 0, Bright 0, Saturation -2, Hue 0).
• Fill a monitor screen with a desaturated Magenta JPG (or TIFF or whatever).
  • (255, 153, 255) has 40% saturation.
• Open the filterless lens wide, aim at the Magenta screen, and set the proper exposure.
• Take a test photo and check exposure settings.
  • Don't blow any channels, but don't underexpose either.
  • Reset the exposure as needed.
• Select a manual White Balance preset slot.
• Measure White Balance against the Magenta screen.

 

STEP TWO:  Push fine tune setting > Measure white balance > Repeat.

 

Let D be your just measured WB slot.

 

for( i=1; i<=5; i++) {

 

Select slot D;

Set Fine Tune to B6, G6 (cyan in the upper left corner);

Select OK;

Measure White Balance against the Magenta screen;

}

• The point of Step Two is to measure a White Balance using an altered Fine Tune
  which has been "pushed" away from red or magenta.
• Note that after each White Balance, Fine Tune returns the cursor to the center AB/0 and GM/0.
  So Fine Tune must be reset to B6, G6 before each WB measurement.

.

---

 

For the Nikon D610 full spec and UV-Nikkor lens (set to f/4.5), 

I obtained these RGB White Balance multipliers after 5 iterations using the B6/G6 Fine Tune.

• R 0.546875
• B 0.648438
• G 1.0 1.0

As demonstrated above this WB lets you see most areas where the UV false blue and UV false yellow

will eventually appear after conversion and white balance of a file.

 

When shooting with this pushed WB in a Nikon, use the same Neutral Picture Control 

for reviewing your shots but add some sharpening (+3 or +4) to better see how focus

and details are being captured.

 

 

---

 

 

EXPERIMENTS

• Experiment with setting Fine Tune to B6/G0 (blue) or B0/G6 (green).
• Experiment with altering the Fine Tune push between blue/cyan/green.
• Play around with different color screens and various Fine Tune pushes to see what you get.
• I once managed to obtain R .546875 and B .718750 which was really nice
  for reflected UV in a Nikon, but I don't remember how I got that.

.

---

 

 

Observations about RGB Multipliers in other cameras

 

Some cameras are capable of measuring White Balance through a dark UV-Pass Filter.

I was curious what the RGB multipliers are for such in-camera WB.

For easier comparisons, I'm going to round off to 2 places.

 

Sony A7R Full-Spectrum 

Asahi Ultra-Achromatic-Takumar

SEU-Red UV-Pass Filter (records more in the blue channel than the BaaderU)

Sunlight

Red  .47 

Blue .68

 

Sony A7R Full-Spectrum

Carl Zeiss Jena UV-Objektiv

BaaderU UV-Pass Filter (records more in the red channel than the preceding SEU-Red)

Sunlight

Red   .42

Blue 1.45 

[photoni]

@Andrea B.  an off-topic question... but which soft do you use for converting raw? in these you used Photo Ninja in high posts I saw Dark Table , after 20 years of Photshop and 3 of Capture One I installed the latest Nikon Soft and Dark Table, but they seem all "stupid & diificult" to me.

[Andrea B.]

I usually use Photo Ninja for conversion, white balance and color profiles.

[maikm]

3 hours ago, Andrea B. said:

• Note that after each White Balance, Fine Tune returns the cursor to the center AB/0 and GM/0.
  So Fine Tune must be reset to B6, G6 before each WB measurement.

 

Unfortunately not true for the Z5. The fine tune values are preserved after new measurements for the same slot.

 

And I want to emphasize again that I'm aiming for IR (720nm) white balance here, not (yet) UV.

[Andrea B.]

For IR 720 nm, start by making a filterless white balance against a green lawn.

 

 

---

 

Unfortunately not true for the Z5. The fine tune values are preserved after new measurements for the same slot.uti

 

I'll go play with my Z for a sec and amend the routine above.

 

---

 

 

For the Nikon Z7, any WB fine-tuning (for example, B6,G6) is reset to center (0,0) after the white balance measurement is made.

 

The Z5 must be different?

[maikm]

17 hours ago, Andrea B. said:

For IR 720 nm, start by making a filterless white balance against a green lawn.

 

Well, that gives me a white preview of that green lawn, and a red image for any normal scene, which only gets even redder with a 720nm filter. I don't see the point?

[Andrea B.]

I give up!!

[otoien]

On 12/6/2022 at 7:54 AM, maikm said:

  On 12/5/2022 at 2:07 PM, Andrea B. said:

For IR 720 nm, start by making a filterless white balance against a green lawn.

Perhaps take out the word "filterless" since maikm is using a full spectrum body and the IR720 nm long pass property is based on having a filter on the lens?

[maikm]

You know what? I just made a successful WB on the Z5 with 720nm filter against green screen (RGB 153/255/153), the camera accepted the measurement, and to be honest this is working pretty well for regular daylight, even without much WB fine tuning added!

[Shane]

You do realize there is a difference in the IR content for green grass vs a green computer screen. The screen probably exhibits no IR, so effectively you have WB the visible leakage of the 720nm filter against visible Green of the screen.

[maikm]

Shane, that may be the theory, but actually it already works well against grass and foliage. Here's a quick example from right now outdoors, with the white balance achieved by this, and with a manual wb later on the leaves (in Nikon Studio NX):

 

 

 

And that's even without WB fine tuning. I think that works well enough to judge the scene in the viewfinder for now. Here's a quick channel swap image with no changes to the white balance before channel swapping (in Affinity Photo 2):

 

[Shane]

It's no theory, I've been doing this for 18 years.

[maikm]

That's not how I meant it. I believe your explanation, but nevertheless I'm surprised it still works in the sense that the Z5 didn't reject the measurement and it works well enough for the intended purpose.

 

And yes, the "filterless" totally threw me off. Still don't know how that was meant, as it's obviously not "without a filter screwed on".

[otoien]

The standard green grass WB is how I did the in-camera WB on my IR converted (720nm Lifepixel) D40x. It works so well that I generally do not repeat it, even on snowy scenes. It is Interesting that the ability to capture and use  this WB appers to be a model specific hit and miss even with these early camera models, so one could wonder about the reason it is limited in some bodies. Great news that you got it working with the Z5. Perhaps try the green grass next to see if you can get all the way? If I ever upgrade the D40x, it would have to be a mirrorless model like Z50/Z5 where accurate manual focus can be achieved directly  in the viewfinder without tweaking regardless of the lens.