Digital Sensitometry for Black and White ~ Zone System ~ RAW processing for film curves

[Updated 7 October, 2024 ~ Important correction: Zone 5 at 0EV is 76hex/118decimal, not 7F/127 as I previously wrote.]

Moving from synthesized digital Zone System step wedges into the real world, I gathered the photos necessary to see how Sony A7 100ISO RAW files behave.

For this blog entry I would like to do three things.  

First, I would like to look at characterizing the sensor by seeing how the Open Source Software distributes tones as a result of the DeMosaic stage and before applying any further processing. 

Second, using the data collected from the synthesized digital Zone System step wedges I created earlier, I would like to use RawTherapee Curves to match tones from Zone 3 through Zone 8 using 0EV as the Zone 5 reference.

Third, I would like to see what effect using -1EV as the Zone 5 reference does to the step wedge.

Importantly I will show my work by including screen shots of the input and output luminosity curves.  Hopefully the data will be more easily interpretable than reading hexidecimal Zone values.

Please note: The following is not really Rocket Science and I'm sure many other interested parties have already crossed this very same bridge years ahead of me. So if you find any errors, please feel free to help me better understand.

Materials -

• Sony A7
• 100ISO
• Manual mode (to accurately control exposure)
• Center spot meter (to avoid vignette impact on exposure)
  
• Light solid color surface
  

Data Collection -

• For each test pass, select the output file type, in this case...
  
• RAW - for processing on a Big Machine off-camera 
• Photograph a solid color surface making exposures...
• From -8EV...
• Through to +5EV
• On a Big Computer...
• Using Open Source Software RawTherapee
  
• Process each exposure to the demosaicing stage, avoiding any further image adjustments
  
• Then measure the tonal values using RawTherapee (or any appropriate image processing software) by
  
• Mousing over the center of the image
• Reading the tonal values (found under the histogram in the upper left)
• Make notes of the tonal values and corresponding EV the image were shot at
• Build a step wedge of EV, tonal values, and Zone indications
• Repeat this sequence after modifying curves
  

Results -

Here is a look at the tonal distribution after RAW file import and demosaicing, but before any further processing.

At this sensor characterization step it is easy to see the output curve is what I'll call hallowed out.  The shadow tone, toe region of the output curve is an interesting ramp, but the highlight shoulder region takes a rather sharp turn at the top.  It's easy to see that the 0EV value, specified as 18% middle gray, does not yet meet the Zone System definition.  Further image processing is required as this value needs to be raised to 76(hex)/118(dec) and Zones 3 thru 8 need also to match the specified tonal values.

Doing that now, here are the results.

0EV/Zone 5 now meets the center of the scale 18% middle gray as specified by the Zone System, and Zones 3 thru 8 tonal values align nicely with expectations.

In general, it should be obvious that the Sony A7 full frame sensor has good dynamic range, exceeding 11EV or 11 f-stops of potentially usable information.  This confirms Photon to Photos dynamic range measurement for this camera.

Looking at the symmetry of tones across the step wedge, the highlight roll-off shoulder region is much narrower than the dark tone roll-on, or toe, which is really quite long.

Stepping back a moment and thinking things through, there is nothing "magical" about using 0EV as Zone 5/18% middle gray.  It's what the industry does, so I've tended to _assume_ that is just how things are.  So what would happen if I liberalized my thinking slightly and select -1EV as the Zone 5 reference as well as correctly matching Zones 3 through 8 to a synthesized and idealized Zone System step wedge?

Let's have a look.

The overall step wedge has shifted to the left and the highlight "shoulder" region has expanded.  

This is rather exciting.

Remember the AMPEX photo reference manuals?  They published film tone curves which I find useful, even now.  They help with understanding how film behaved.  These digital exposure correction curves are well within the "ball park" what those old references shared.  

It's fun to realize that I now have a much truer "film-like" tonal curve for digital than I've ever had before.  Additionally, if for any reason I don't like the shape of the output curve, modifying the input correction curve is incredibly easy to do.

Regarding Zone System digital image processing, I think I have an adequate answer. 

---------- Where am I on my checklist of perceived "needs"? ---------------

Is what I'm doing here in trying to transfer old film-based Zone System knowledge into something practicable for digital...

• Measurable - Yes
• Accurate - Yes
  
• Repeatable - Yes
• Controllable - Yes
• Clearly understandable - I believe yes. If not, please let me know.
• Reveals what Sony's in-camera jpg processor is doing - Not yet 
• Helps me choose a set of in-camera jpg engine operations and/or settings that match my RAW output - Remains to be seen
• Bonus: I now have a digital match to the old film-based Zone System 

Note: What I'm sharing in this series is limited to my experience with Sony full frame and APS-C mirrorless cameras and the sensors they come with.  In truth I have zero idea how other digital systems map tones.  I hope the description of what I'm doing here helps interested parties sort this out for their situations.