Metering for Portraits and the EVF

I'm going to point the car in a different direction...
In the EVF, everything looked perfect i.e. perfectly exposed and when I took the photos into LR, they were all underexposed by at least 1/2 to 1 stop.
How did you make this determination? Visually? And while you're at it, please state what a correct exposure is in your mind.

And then we get to "what were your Lightroom settings?" Adobe's raw conversions are well known to apply unseen "exposure" adjustments (really moving the DNs in a tonal curve). Your camera's viewfinder is using JPEG streams out of the EXPEED chip, which use Nikon's Picture Controls.

Everyone expects that every manufacturer or software developer uses the same interpretations of data. This is absolutely not the case. Which is why I've been preaching since the beginning of digital that you have to understand what each component is actually doing.

An exposure (correct use of the term) presented itself to the image sensor. What happens from that point onward is a series of manipulations, assumptions, and adjustments by a variety of components/algorithms. Who/what is doing that, and what it is that they're doing is the topic we all have to come to grips with.

Which returns me to my original question: how did you make the determinations (plural) that one was a proper exposure and the other wasn't?
 
Kodak correctly say 18% is NOT a mid grey and to add half a stop if reading directly off a Kodak gray card or angle the card so it reflects less than 100% of the light source output - by how much angle depends on when the card was printed.
Depends...Nikon often uses a different calibration for midtone depending on metering mode. As an example...many Nikon DSLR's when set to Spot and CW metering solve for a 12% grey midtone while Matrix metering solves for 18% gray. I have not checked the OP's model
Basics first.

18% reflectance plus 1 stop is 36% reflectance.

Add a second stop is 72% reflectance.

Add another half stop is just over 100% reflectance - impossible without a mirror like subject reflecting the light source.

18% cannot be a mid tone - unless you restrict dynamic range to 5 stops.

In the era of film 8 or 9 stops negative film dynamic range was a common maximum. A mid tone for 8 or 9 stops is a lot less reflectance than 18% ;-)

Nikon reflectance levels next

While I have "only" extensively used Nikon for 24 years I have never owned a Nikon body calibrated to 18% reflectance for spot or centre weighted - mine have always been set at the factory to less than 18%.

A note on spot metering in DSLR's - the spot size in the viewfinder indicated by an AF point is dramatically larger than the AF point rectangle - as a guide about a third of the size of a 12% centre weighted metering circle.

A note on matrix - it applies its own compensation without saying how much :-( based on matrix scene analysis - except it sets on upper exposure limit based on bright sunlight which I think is why it gets sunlit snow scenes near perfect - and whites in lower light somewhat under-exposed.

21st Century.

When I was aged about 5 many houses had what in UK was called a larder with a small north facing room with a large stone slab to keep food cool and fresh.

In the 21st-century most own a fridge freezer.

Similarly the histogram although slightly limited by using jpeg rather than RAW information provides far more exposure information than one tone reflectance systems.

Except in a few not everyday subjects such as highlight weighted spot metering for a spotlit performer on a stage; as many are saying the histogram is usually the best exposure aid available,
Interesting. Can you explain this?
We're getting into an entirely different subject that probably needs a thread of its own. Decades ago I wrote an article about "metering" that attempted to bring forward the original Ansel Adams versus Kodak debate about where "middle gray" was. Included in that article was an examination of what the camera makers were doing and what the standard they claimed to be following said.

I was constantly pummeled with counterarguments to that article, most just repeating mythical things they learned along the way from random sources. I eventually got tired of defending it, so took it down.

What Leonard wrote above is similar to what I had been writing back in the 90's (with film, but at that point I also had considerable digital experience, and you get into the same discussion, just not chemically ;~). Basically: middle gray isn't necessarily what you think it is.
I use the greyscale target with the 3 steps patches measuring the 40 IRE mid-tone gray patch of the ColorChecker Passport Video: https://calibrite.com/us/product/colorchecker-passport-video/
Well, IRE brings up a different set of standards ;~). IRE is a broadcast television thing, and originally had to do with where you put analog values in the NTSC space. If that weren't enough, we now have LOG files, and you'll note that they're named N-LOG, S-LOG, C-LOG, etc. That's because while they're all logarithmic in nature, where each manufacturer places skin tones and how they distribute the other tonal values. And then we use a LUT to make that what we want ;~). More recently, we've gotten HDR video, which also informs HEIF, which is yet another interpretation of where tonal values should be recorded and how they should be displayed.

So:

EXPOSURE = LIGHT filtered by APERTURE filtered by SHUTTER SPEED

Exposure is what the image sensor sees. Everything that happens after that is going to provoke debates, discussions, arguments, reverse engineering, and more.
 
I measured 8.8% on the Z9. I suspect all the Z's are likely the same.
Thank you for that! Is it the same across all metering modes?
Yep. The only difference you'll see across metering modes is the relative weighting given to different parts of the frame. The notion of middle gray doesn't change between them though.
 
I made the mistake of shooting in crap 2pm harsh light today
never do outdoor portraits at 2pm, lol Unless you have shady spots or assistants holding up diffuser/shades etc
but noticed something interesting in post.

When shooting on the Z7ii > in A > using Face detection (AF Area Wide People) and Matrix metering…

In the EVF, everything looked perfect i.e. perfectly exposed and when I took the photos into LR, they were all underexposed by at least 1/2 to 1 stop.

Generally, I was wide open f/2.8 thru f/4 and shooting at 1/1600 - 1/3200.

I’m thinking that maybe Spot metering is a better idea for portraits on the Z7ii or at least in situations where you’re dealing with harsh light. Thoughts?

Will experiment more with Center weighted + spot for portraits, I generally don’t shoot in such crap light but sometimes you have no choice.

Just odd that the EVF preview was perfect and results came out underexposed…
 
You can get the in-camera generated histogram to be a closer representation of the distribution of tones in a raw file by doing two simple things: set the color space to Adobe RGB and the camera profile to flat.



Granted, , because of the Flat profile what you see in the EVF (if mirrorless) and on the camera’s monitor Weill look a little washed out and low contrast but remember, your only looking at a preview of the raw files content.
 
I made the mistake of shooting in crap 2pm harsh light today but noticed something interesting in post.

When shooting on the Z7ii > in A > using Face detection (AF Area Wide People) and Matrix metering…

In the EVF, everything looked perfect i.e. perfectly exposed and when I took the photos into LR, they were all underexposed by at least 1/2 to 1 stop.

Generally, I was wide open f/2.8 thru f/4 and shooting at 1/1600 - 1/3200.

I’m thinking that maybe Spot metering is a better idea for portraits on the Z7ii or at least in situations where you’re dealing with harsh light. Thoughts?

Will experiment more with Center weighted + spot for portraits, I generally don’t shoot in such crap light but sometimes you have no choice.

Just odd that the EVF preview was perfect and results came out underexposed…
If you are shooting portraits under controlled conditions (not run and gun), I would recommend trying an incident light meter.
Your camera meter works with reflected light and that can throw the readings off.

Incident light readings measure the light falling on your subject rather than reflected back from your subject.

On commercial shoots I always use a light meter and a color meter (spectrometer) to set the exposure and white balance.
It removes a lot of uncertainty and trial and error later in post.

If you don't want to use a light meter you can use a grey card and spot meter from that.
Just make sure the card is in the same light as the important part of your subject (such as the face).

You can do a similar thing for white balance - use a white balance card held in front of the subjects face and do a custom white balance in camera.

The Calibrite ColorChecker Passport Photo 2 is a great investment for setting exposure, white balance and color corrections.
It retails for around £110.
 
I made the mistake of shooting in crap 2pm harsh light today but noticed something interesting in post.

When shooting on the Z7ii > in A > using Face detection (AF Area Wide People) and Matrix metering…

In the EVF, everything looked perfect i.e. perfectly exposed and when I took the photos into LR, they were all underexposed by at least 1/2 to 1 stop.

Generally, I was wide open f/2.8 thru f/4 and shooting at 1/1600 - 1/3200.

I’m thinking that maybe Spot metering is a better idea for portraits on the Z7ii or at least in situations where you’re dealing with harsh light. Thoughts?

Will experiment more with Center weighted + spot for portraits, I generally don’t shoot in such crap light but sometimes you have no choice.

Just odd that the EVF preview was perfect and results came out underexposed…
If you are shooting portraits under controlled conditions (not run and gun), I would recommend trying an incident light meter.
Your camera meter works with reflected light and that can throw the readings off.

Incident light readings measure the light falling on your subject rather than reflected back from your subject.

On commercial shoots I always use a light meter and a color meter (spectrometer) to set the exposure and white balance.
It removes a lot of uncertainty and trial and error later in post.
How do you use a spectrometer to determine WB ?
If you don't want to use a light meter you can use a grey card and spot meter from that.
Just make sure the card is in the same light as the important part of your subject (such as the face).

You can do a similar thing for white balance - use a white balance card held in front of the subjects face and do a custom white balance in camera.
Or take a picture of the person holding the WB card in front pointing to the camera and determine WB during post processing with teh eye drop tool ?
The Calibrite ColorChecker Passport Photo 2 is a great investment for setting exposure, white balance and color corrections.
It retails for around £110.
 
How do you use a spectrometer to determine WB ?
I use the Sekonic C-800 spectrometer.
To use this with your camera you take a reading with the meter.
It will give you a value in Kelvin (K). Lets say it reads 5670K.

In your camera you set your white balance to Kelvin and adjust it to the closest value to the metered value (5670K in this example).

If your camera cannot exactly match the value, set the closest value available.

On my Nikon Z9 I can manually adjust the blue/yellow and green/magenta.

If the desired kelvin value (5670K) isn't available I will use the blue/yellow adjustment to shift the white balance in the right direction.

The closest kelvin value on the Z9 is 5560K so I add a bit of yellow to compensate.

There is a formula you can use to calculate the offset needed.

The Z9 will show the kelvin value of the photo taken so you can tweek the offset till you get as close to your target as possible.

Spectrometers are expensive at around £1500 so most won't use one.
But I find it very useful and in the greater scheme of things it is a very valuable tool.

Spectrometers also tell you what color corrections you need to apply when using mixed lighting.
It can tell you what gels to use on a light to balance it to another.
It can also tell you what filter to use on your camera to balance it to a light source.

If you shoot with flash in an ambient environment you can work out the correct gels for your flash to match it the ambient light. Then set your camera white balance to the ambient color temperature and all should look correct.

If you don't want to use a light meter you can use a grey card and spot meter from that.
Just make sure the card is in the same light as the important part of your subject (such as the face).

You can do a similar thing for white balance - use a white balance card held in front of the subjects face and do a custom white balance in camera.
Or take a picture of the person holding the WB card in front pointing to the camera and determine WB during post processing with teh eye drop tool ?
That works too as long as you shoot RAW.

The Calibrite ColorChecker Passport Photo 2 is a great investment for setting exposure, white balance and color corrections.
It retails for around £110.
 
How do you use a spectrometer to determine WB ?
I use the Sekonic C-800 spectrometer.
To use this with your camera you take a reading with the meter.
It will give you a value in Kelvin (K). Lets say it reads 5670K.

In your camera you set your white balance to Kelvin and adjust it to the closest value to the metered value (5670K in this example).

If your camera cannot exactly match the value, set the closest value available.

On my Nikon Z9 I can manually adjust the blue/yellow and green/magenta.

If the desired kelvin value (5670K) isn't available I will use the blue/yellow adjustment to shift the white balance in the right direction.

The closest kelvin value on the Z9 is 5560K so I add a bit of yellow to compensate.

There is a formula you can use to calculate the offset needed.

The Z9 will show the kelvin value of the photo taken so you can tweek the offset till you get as close to your target as possible.

Spectrometers are expensive at around £1500 so most won't use one.
But I find it very useful and in the greater scheme of things it is a very valuable tool.

Spectrometers also tell you what color corrections you need to apply when using mixed lighting.
It can tell you what gels to use on a light to balance it to another.
It can also tell you what filter to use on your camera to balance it to a light source.

If you shoot with flash in an ambient environment you can work out the correct gels for your flash to match it the ambient light. Then set your camera white balance to the ambient color temperature and all should look correct.
If you don't want to use a light meter you can use a grey card and spot meter from that.
Just make sure the card is in the same light as the important part of your subject (such as the face).

You can do a similar thing for white balance - use a white balance card held in front of the subjects face and do a custom white balance in camera.
Or take a picture of the person holding the WB card in front pointing to the camera and determine WB during post processing with teh eye drop tool ?
That works too as long as you shoot RAW.
The Calibrite ColorChecker Passport Photo 2 is a great investment for setting exposure, white balance and color corrections.
It retails for around £110.
 
I made the mistake of shooting in crap 2pm harsh light today but noticed something interesting in post.

When shooting on the Z7ii > in A > using Face detection (AF Area Wide People) and Matrix metering…

In the EVF, everything looked perfect i.e. perfectly exposed and when I took the photos into LR, they were all underexposed by at least 1/2 to 1 stop.

Generally, I was wide open f/2.8 thru f/4 and shooting at 1/1600 - 1/3200.

I’m thinking that maybe Spot metering is a better idea for portraits on the Z7ii or at least in situations where you’re dealing with harsh light. Thoughts?

Will experiment more with Center weighted + spot for portraits, I generally don’t shoot in such crap light but sometimes you have no choice.

Just odd that the EVF preview was perfect and results came out underexposed…
If you are shooting at 2.8 to F4 and only at a shutter speed of 1600 to 3200 you are not in that bright of outdoor light.

I shoot in much worse every week. Many times I can not shoot at a 2.8 as it over the 8000 of a shutter speed even at a F4 it can be over at times on some days and we start a hour before sunset.

So you are in Aperture priority this is what I use the most outdoor for natural light shooting and on my Z6 I have no problem seeing the real live exposure I rely on it and its very dead on. What I see is what I get.

When you do a image review in the camera after a shot its it not showing you the same as what you just shot?

Do you have the live view turned on?

I use exposure comp on everything outdoors in bright light to get a better exposure as the sun can make the camera underexpose peoples faces.

Did you have C focus turned on that can be a problem when using face tracking and I would never use it or wide area ,

If C was on what can happen is as you have the person framed up and you move the camera framing or they move the littlest bit and Is tracked the face the exposure also changes as well for every little movement . So you can get a lot of under or over exposed images.

I dont use the face tracking as for portraits its really not needed you are shooting a person most of the time not moving around much so just use single focus point and single shot never C.

The z6 and 7 face tracking is not very good in the first place.

I also have a newer Sony now and its tracking is fantastic in so many ways I am starting to use it. But even with it for portraits outdoors the tracking also tracks the exposure default settings so the exposure changes for any little movement of re framing the shot so fast I get exposure problems. I think I now have it set to lock the exposure on the first 1/2 shutter push even in C focus mode so the exposure stays the same as the person moves or I re-frame some but then I can not use exposure comp easy.

So in the end using just single focus point I put on the person and single shot works the best to keep the exposure I want and I can use exposure comp to get and set the exposure I want.

Also I can care less about the WB as I shoot raw and can change it to anything when editing. Shooting outdoors it changes all the time. A cloud moves over the WB changes it moves aways change again different angles shooting it changes.
At sunset you get a orange cast from the sun and other parts of the day different sun casts that change as you are shooting.

--
Started shooting digital back with the first 2MP cameras. Over 20 cameras later still going. I shoot for a living - family and people portraits, weddings, Sports and a little of everything.
 
Last edited:
I measured 8.8% on the Z9. I suspect all the Z's are likely the same. This was done by measuring the actual raw data of a middle gray target relative to raw saturation, using the Z9's 1/96 increments of shutter speed precision (the high-speed flicker mode).
This is about 1.2 stops darker than 18% reflectance - which makes sense to me.

Thom's comments are interesting - especially as probably over 90% of Kodak gray card actual owners have not taken note of the Kodak guidance to adjust the meter reading off the card to get to close to 12% - and then contradict Kodak guidance :-O

On metering accuracy - there is no exact science as individual photographers have personal preferences of how the image should both be produced in camera; and often how it should be post processed to achieve their photographers preferences.

On matrix metering - as far as I can find out Nikon has not said exactly how it works in Z cameras - only that it works similar (implying not identical) to previous DSLR's etc.

In my F100/F5 slide film days I found matrix tended to expose accurately for the foreground and slightly over expose for the sky.

With the sky in a scene I tended to use some minus exposure compensation to get the exposure that I preferred.

On spot metering and compensation for subject reflectance, invercones and similar alternatives - their widespread current use seem to have generally receded into history.

This may be because intelligent use of the histogram (including the 3 colour option) usually provides far more comprehensive and almost always more accurate information at the touch of a button in under 1 second - much faster than separately using a hand held meter and maybe an angled reading from a gray card.
 
I just think the Matrix Metering is spotty in certain conditions vs. a more accurate center-weighted or even spot metering for portraits, for example.
If by "spotty" you mean the matrix meter reading "seems to be" particularly influenced by the reflectance of what is under the active AF point - in the Z6 and 7 - I agree this is so.

For "action photography" over extended periods for subjects consisting of a wide range of reflectances taking test exposures and then using manual exposure is often widely practised.
 
I just think the Matrix Metering is spotty in certain conditions vs. a more accurate center-weighted or even spot metering for portraits, for example.

I
Matrix on Nikon Sucks

They say that it is programmed for 1,000's of photos (or even more) and if I recall should handle 99% or (95%) I forget. But I found out years ago that I need that 1% and hence do not use it very often.

Also 95% of my photos is Studio Strobe model photography and hence I use a Lightmeter and don't use Nikon's meter.

But the best part of FF sensor you can get away with 3 EV or even more under exposure.

Center weight average is what I used and on my D300 (4 settings 6mm-13mm) it also had different settings for the size of the center. On my Z7ii ir is only 12mm
 
I just think the Matrix Metering is spotty in certain conditions vs. a more accurate center-weighted or even spot metering for portraits, for example.

I
Matrix on Nikon Sucks
I don't agree.
They say that it is programmed for 1,000's of photos (or even more) and if I recall should handle 99% or (95%) I forget. But I found out years ago that I need that 1% and hence do not use it very often.
So it's a problem for you, and not in general. That clarification matters a lot.
Also 95% of my photos is Studio Strobe model photography and hence I use a Lightmeter and don't use Nikon's meter.
Or just use spot metering? Or center weighted?
But the best part of FF sensor you can get away with 3 EV or even more under exposure.
Well, you can...but it's best to expose optimally to capture the best data.
Center weight average is what I used and on my D300 (4 settings 6mm-13mm) it also had different settings for the size of the center. On my Z7ii ir is only 12mm
See above, spot metering might work for you.
 
Matrix on Nikon Sucks...
Center weight average is what I used and on my D300 (4 settings 6mm-13mm) it also had different settings for the size of the center. On my Z7ii ir is only 12mm
Which is really just a form of Matrix metering with a bias towards the center. With Nikon D300...CW metering is not spot metering of the center. It's Matrix metering with a center bias and the size of the bias can be changed to a degree.
 
I just think the Matrix Metering is spotty in certain conditions vs. a more accurate center-weighted or even spot metering for portraits, for example.

I
Matrix on Nikon Sucks
This statement of you is at least extreme exaggerated.
They say that it is programmed for 1,000's of photos (or even more) and if I recall should handle 99% or (95%) I forget. But I found out years ago that I need that 1% and hence do not use it very often.

Also 95% of my photos is Studio Strobe model photography and hence I use a Lightmeter and don't use Nikon's meter.

But the best part of FF sensor you can get away with 3 EV or even more under exposure.

Center weight average is what I used and on my D300 (4 settings 6mm-13mm) it also had different settings for the size of the center. On my Z7ii ir is only 12mm
For most people in most situation Matrix-Metering on Nikon gives good results, specially if there is not time for more sophisticated metering methods, like multi-spot-zone-metering + exposure calculation. It is not possible to create a metering method that will work perfect in every situation for every person. To make the statement that something 'sucks' because it does not match your expectations on special lighting conditions seems to me a little bit 'out-of-track'.
 
Matrix on Nikon Sucks...

Center weight average is what I used and on my D300 (4 settings 6mm-13mm) it also had different settings for the size of the center. On my Z7ii ir is only 12mm
Which is really just a form of Matrix metering with a bias towards the center. With Nikon D300...CW metering is not spot metering of the center. It's Matrix metering with a center bias and the size of the bias can be changed to a degree.
As far as i know, CW metering is very different from matrix metering, in a technical way and in results.

The classical metering modes (average, center-weight, spot) use one sensor element and the different characteristics are done with optical methods.

Matrix metering works with multiple sensor elements (i think 5 in early Nikons) for different areas of the picture, and try to find a matching exposure based on samples that are used to train/calibrate the system. With using multiple sensor elements, matrix metering gets a guess of the DR of the scene, what is not possible for the classic one-element metering methods.

When i used matrix metering the first time (i think with a Nikon F80), i was impressed how good it handled difficult scenes like snow in foreground, sky in background or backlight.
 
I just think the Matrix Metering is spotty in certain conditions vs. a more accurate center-weighted or even spot metering for portraits, for example.

I
Matrix on Nikon Sucks

They say that it is programmed for 1,000's of photos (or even more) and if I recall should handle 99% or (95%) I forget. But I found out years ago that I need that 1% and hence do not use it very often.

Also 95% of my photos is Studio Strobe model photography and hence I use a Lightmeter and don't use Nikon's meter.
All matrix Metering sucks when you have a subject you want to expose correctly and particularly when you are in a controlled environment. Yes are correct to use manual exposure, setting lighting for your subject and who cares what matrix or other in camera metering says - be this based on a spot, a circle or matrix -- all these do is work on various sized averages (yes look at how spot mode works and the various settings you can choose between).

It is the light on your subject that matters the most. Hopefully you are able to control highlights so these just do not blow out and control contrast - by using the appropriate light modifiers and suitable other lights and reflectors.

I had a chat with Peter Coulson, whose video "P for Professional" - where he says "I am an artist and I want full control over everything, I also find when I'm in any auto mode I lose my creative feel and only get happy snaps". He makes the point like you - it should not matter what the colour or lighting of the background is when you are shooting a model's face -- it is the lighting on the face and yes you can use a meter or just work it out like Peter does by experience and then confirming on Capture One -- shooting digital we can waste a few shots to get it right.
But the best part of FF sensor you can get away with 3 EV or even more under exposure.

Center weight average is what I used and on my D300 (4 settings 6mm-13mm) it also had different settings for the size of the center. On my Z7ii ir is only 12mm
OK -- but what happens when you move your subject away from the centre -- you can "meter" using centre weighted average to find OK starting point settings and then lock the settings and turn off meterings ability to change settings in the camera by going entirely manual. AND then check the exposure in test shots and adjust.

--
areallygrumpyoldsod
Nikon and Hasselblad shooter -- wildlife and and --
https://www.andymillerphoto.co.uk/
https://www.flickr.com/photos/ajm057/
I do not respond to PMs or messages via my website
 
Last edited:

Keyboard shortcuts

Back
Top