Correcting Overexposed RAW Files

Richard M Nikon

Well-known member
Messages
202
Reaction score
305
I strive to expose my shots accurately, but occasionally I have to do some corrective work afterwards.

Using Nikon NX Studio, I've been able to "reel in" various degrees of overexposure on some of my 12-bit lossless, compressed Z7 RAW files. Typically, the correction has been a fraction of a stop, but I've had occasional success with 3 stops. By success, I mean the post-correction histogram shows no highlight clipping.

Does anybody know what, if any, negative effects these corrections might have on the final image?

Many thanks.
 
You just have to judge highlights by how they look in the photo, not so much how the histogram looks. You can of course adjust it to make it not look overexposed on a histogram, but that often means that blown highlights are now bright gray instead of white. Lightroom does this. It recovers some highlights, but the rest it just pulls them into gray.

There are some tricks with luminosity masks you can use. You can add a tint to highlights, or grain, anything so that when you print it's not just a blank spot on the paper.

--
https://www.flickr.com/photos/jkrumm/
https://juneauphotographs.org/
 
Last edited:
I strive to expose my shots accurately, but occasionally I have to do some corrective work afterwards.

Using Nikon NX Studio, I've been able to "reel in" various degrees of overexposure on some of my 12-bit lossless, compressed Z7 RAW files. Typically, the correction has been a fraction of a stop, but I've had occasional success with 3 stops. By success, I mean the post-correction histogram shows no highlight clipping.

Does anybody know what, if any, negative effects these corrections might have on the final image?

Many thanks.
As long as you stay within the bounds of where the sensor has collected useful information, the drawbacks should be minimal. However, at some point you will get colour shifts, posterisation, or simply a grey blob (instead of a white blob). I would say that as long as it looks good, you're good. ;)
 
There is no inherent penalty to adjusting image lightness in post. We can't alter exposure after a photo has been made. The biggest drawback of adjusting lightness is revealing noise, clipped highlights, or crushed shadows that were already present in the image. There are pros and cons to consider when weighing which tool or technique to use to adjust image lightness. But that's a topic for another thread.

--
Bill Ferris Photography
https://billferrisphotography.pixieset.com/arizonaslittleserengeti/
 
Last edited:
I strive to expose my shots accurately, but occasionally I have to do some corrective work afterwards.

Using Nikon NX Studio, I've been able to "reel in" various degrees of overexposure on some of my 12-bit lossless, compressed Z7 RAW files. Typically, the correction has been a fraction of a stop, but I've had occasional success with 3 stops. By success, I mean the post-correction histogram shows no highlight clipping.

Does anybody know what, if any, negative effects these corrections might have on the final image?

Many thanks.
Overexposure is a lot harder to fix than underexposure. Personally I use the "Highlight Weighted" metering mode to protect highlights in the frame and I know I can pull the shadows back up even a few stops without a huge penalty. It depends on what ISO you're using of course, the higher-ISO images may reveal more noise when shadows are pulled.

Since highlight weighted metering tends to underexpose by a lot in some more extreme situations, I dial in +0.7EV to prevent it from going too far.
 
I strive to expose my shots accurately, but occasionally I have to do some corrective work afterwards.

Using Nikon NX Studio, I've been able to "reel in" various degrees of overexposure on some of my 12-bit lossless, compressed Z7 RAW files. Typically, the correction has been a fraction of a stop, but I've had occasional success with 3 stops. By success, I mean the post-correction histogram shows no highlight clipping.

Does anybody know what, if any, negative effects these corrections might have on the final image?

Many thanks.
I sometimes manage to do this too. When it happens for me I suspect that it might have something to do with which metering mode the camera may have been in. I often use Spot or Center metering.

I only use NX Studio, I've not learned other programs, so I will not guess what they do. I have Z7 and Z9 cameras. Files seem to be fairly flexible in NX Studio using a combination of of the tabs for "Exposure Comp." and "Adjust Brightness and Color". In the "Picture Control" tab the sub tab "Brightness" is less friendly to the overall picture and requires a light touch.
 
I also use NX Studio as my primary or at least initial image processor, and I always shoot raw. If I see a wacky under or over exposed image, I always look at what I call add-on settings like Active D-lighting and really contrasty or weak picture controls. I set everything to neutral as a first step and then take it from there. I usually don't leave Active D on, so if I'm under exposed I'll sometimes check it in normal and high settings to see what it gives me. If it's really overexposed then I'll maybe start with exposure comp and go from there. I'll also use the local edit (control points) to bring up a low portion. I find the control points are better when bringing up exposure than taking down an over exposure.

I find that with an EVF I rarely get an image that's more than a stop out, which is usually pretty easy to correct.
 
I'm with @TheSoaringSprite on this, highlight-weighted metering is king. Wasn't a part of my decision to get a Nikon Z 6, but I'm sure glad it's there. Every time I've switched off it to compromise-handle high DR situations I've regretted it. I spend more time futzing with lifting tone curves to yank shadows into view, but I find it's worth that effort.

Correcting it after the fact depends on how many channels you've blown. One or two, and a decent highlight recovery algorithm can paint that into the region without making a pink mess. Three, and something has to make up stuff to fill in the lack of information, yay AI?

Multi-capture HDR bears consideration. The phones do it without much fuss; I watched in horror as one of my selfie's blown background sky got replaced automatically with nicely-exposed data from one of the other captures surrounding the moment of interest. Geesh...
 
I strive to expose my shots accurately, but occasionally I have to do some corrective work afterwards.

Using Nikon NX Studio, I've been able to "reel in" various degrees of overexposure on some of my 12-bit lossless, compressed Z7 RAW files. Typically, the correction has been a fraction of a stop, but I've had occasional success with 3 stops. By success, I mean the post-correction histogram shows no highlight clipping.

Does anybody know what, if any, negative effects these corrections might have on the final image?

Many thanks.

Does NX Studio display the true RAW histogram?
 
No, the histogram in Nikon NX Studio is not a "true" raw histogram showing unprocessed sensor data. It is based on the
applied in-camera or in-software Picture Control settings and other adjustments.
No, Lightroom Classic does not show a "true" or completely unprocessed raw histogram
In Lightroom classic . The histogram it displays is based on its own default rendering and interpretation of the raw data, using a working color space (which is a variation of ProPhoto RGB with an sRGB tone curve applied by default).

RawDigger shows the true raw histogram, which differs significantly from the in-camera or standard raw converter histograms, because it displays the unprocessed data directly from the camera's sensor. This allows for a much more accurate analysis of true overexposure, underexposure, and noise, especially for photographers practicing exposure-to-the-right (ETTR) techniques.

In Lightroom classic or ACR you can select HDR and then the histogram shows 4+ stops above SDR, which you can bring down to SDR.
 

Keyboard shortcuts

Back
Top