Kenneth Lee wrote:
What has confused me is that there is no histogram which shows the dynamic range of the raw file itself: only the result of raw conversion.
I feel like this is somehow mucked up Adobe/Camera RAW settings, because I've never seen this type of behavior with Adobe Standard.
gdanmitchell wrote:
It isn’t exactly “discarded” by Adobe profiles. The data are all still there. It can be subject to a different curve that may open one end of the luminosity spectrum and narrow another, etc.. Choosing a different profile is one quick way to apply a curve that may be more appropriate to a given raw file. Another is to use the controls in your raw converter to customize the adjustments for your image.
Kenneth Lee wrote:
Now I am further baffled: Using the Adobe Color profile and lowering the Exposure setting, all kinds of details in the sky are revealed.
You have discovered the great thing about sticking with raw as long as you can — the files are very malleable! As long as the sky (or other bright subject) isn’t literally “blown out” (overexposed so that high luminosity areas end up with same maximum luminosity value), it is usually possible to recover detail/texture in the very bright areas.
There are some new tools in Adobe software that can help a lot with this, especially some powerful tools for selecting and masking. For example, you could select all areas with luminosity above some value, control how steeply the selection “rolls off,” combine it with other selection tools, and so on.
Imagine a scene with a blown out sky above people on a bright beach. While you could try to just apply something like a reduction in highlights or whites in the raw app, that would affect both areas, likely dulling the image of he beach. So you could simply mask the sky and adjust it separately. Or you could use a brush to select only highlight-filled areas and lower the highlights setting in just those areas. Combined with the masking capabilities, adjusting exposure, highlights, whites and more will let you recover a lot of the image from a raw file.
The tools are really powerful and flexible. Just be sure to avoid actually blowing the highlights in the original exposure and there will be a lot you can do.
To generalize, my own mantra is “expose for highlights and recover the shadows.)
Kenneth Lee wrote:
What has confused me is that there is no histogram which shows the dynamic range of the raw file itself: only the result of raw conversion.
You need https://www.rawdigger.com/
to see the raw histograms. There are four channels in raw, and the clipping may affect these differently. RawDigger will also give you the percentages of clipped pixels.
Note two significantly different reasons for clipping. In images acquired at the base ISO, the clipping is on the sensor - the amount of light projected onto the sensor exceeds the capacity of the sensor. To deal with this, one must increase the shutter speed or close the aperture.
At ISO > base ISO, the distinct different reason for clipping is ADC overflow. Then, reducing the ISO gain can help. This may require shooting in full manual, because modern cameras that I know change the shutter speed or the aperture in the Aperture priority or Shutter priority modes, respectively, when one changes the ISO.
When the raw isn't clipped in RawDigger, the highlights are fully retained and can be manipulated in post to one's taste. When only one channel in raw is clipped, the highlights can be mostly recovered if the raw converter has a good demosaicing algorithm. If you are interested in getting a better command of post-processing, you should be looking for a processing application that has the best demosaicing. The color profiles can be different as well, but the profiles are secondary to demosaicing. If you are sufficiently motivated and would like to test and compare the existing alternatives to ACR, you may well find the experience useful, ot at least interesting.
...If you are sufficiently motivated and would like to test and compare the existing alternatives to ACR, you may well find the experience useful, ot at least interesting.
I thought I had a fairly good grasp of the topic. I was wrong !
Here's what ChatGPT said about raw development tools and the use of "linear" profiles:
When people talk about “linear profiles” in Adobe Camera Raw, they’re essentially trying to make ACR behave a bit more like RawDigger or FastRawViewer—i.e., closer to the actual sensor data.
But there’s an important nuance:
👉 A “linear profile” does not make ACR truly raw—it just removes one major layer of interpretation.
1. What a “linear profile” actually is
In ACR, a camera profile normally includes:
A tone curve (nonlinear brightness shaping)
A color transform (camera → working color space)
Sometimes additional “look” adjustments
A linear profile means:
The tone curve is removed (or flattened), so brightness stays proportional to the original sensor data.
So instead of:
shadows lifted
highlights compressed
you get:
straight, proportional light values
2. The key effect: removing the “S-curve”
Default ACR profiles (like Adobe Color) apply an S-shaped curve:
Dark tones → lifted
Midtones → contrast boosted
Highlights → gently rolled off
A linear profile removes that.
👉 Result:
Image looks darker
Lower contrast
Highlights look harsher (closer to clipping)
Much more like FastRawViewer’s RAW view
3. What it does not remove
Even with a linear profile, ACR is still doing quite a lot:
Still applied:
Demosaicing
White balance
Color matrix (camera → RGB space)
Gamma encoding for display
Some internal highlight handling
Question: Using an adobe profile, if we perform adjustments to the brightness histogram to counteract the adobe tone curve, is that the same as using a linear profile, or will we introduce multiple corrections which introduce banding and other artifacts ?
ChatGPT: Good question—and the short answer is: no, it’s not the same as using a true linear profile, and yes, trying to “undo” Adobe’s tone curve can introduce subtle issues (though not always catastrophic ones).
Let’s unpack why.
1) Why “undoing” the tone curve isn’t equivalent
In Adobe Camera Raw, the tone curve in a profile is applied very early in the rendering pipeline—before most of your slider adjustments.
That has two important consequences:
a) The transformation is not perfectly invertible
The Adobe tone curve:
Compresses highlights (rolloff)
Expands some midtones
Lifts shadows selectively
This is not a simple gamma curve—it’s often piecewise and perceptual
When you try to reverse it using sliders (Exposure, Tone Curve, etc.), you’re:
Applying a different curve later in the pipeline
Trying to approximate the inverse, not actually remove the original
So mathematically:
You get “curve A + inverse-ish curve B,” not “no curve.”
b) Information has already been redistributed
This is the more subtle issue:
Highlight compression may have already squeezed tonal differences together
Shadow lifting may have already spread noise apart
Even if you “reverse” the shape visually:
You don’t recover the original distribution of values
You’re working on post-shaped data, not raw-linear data
2) Does this cause banding or artifacts?
Usually: not severe—but possible
Modern raw pipelines (including Adobe’s) use:
High bit depth (often 16-bit or floating point internally)
Smooth curve math
So:
You won’t usually see obvious banding just from compensating curves
However, you can introduce:
a) Micro-contrast distortions
Slightly unnatural tonal transitions
Especially in skies or smooth gradients
b) Highlight reconstruction differences
Once highlights are rolled off, reversing the curve:
Does not restore clipped or compressed relationships perfectly
c) Noise reshaping
Shadows lifted → then pushed back down ≠ untouched shadows
Noise can look different than with a linear profile
3) The deeper issue: pipeline order
Think of it like this:
With Adobe profile:
RAW → tone curve → your adjustments
With linear profile:
RAW → your adjustments
If you try to “cancel out” the curve:
RAW → tone curve → inverse curve
That is not the same as skipping it entirely, because:
Operations are not commutative
Some steps are lossy (especially highlight compression)
4) When the difference actually matters
You’ll notice it more when:
Doing technical evaluation (like with RawDigger)
Matching renders across software (e.g., RawTherapee)
Performing precise tone mapping or scientific imaging
You’ll notice it less when:
Making normal photographic edits
Working within moderate dynamic range
Outputting to JPEG/web
5) Practical takeaway
Can you approximate linear behavior using Adobe profiles + adjustments?
→ Yes, visually.
Is it technically the same as a true linear profile?
→ No.
Will it break your images?
→ No—but it can subtly alter tonal relationships and noise.
A useful mental model
Think of Adobe’s tone curve as:
“baking contrast into the negative”
Trying to undo it later is like:
“editing a baked image to look unbaked”
You can get close—but you’re not working with the same underlying material.
Question: Using an adobe profile, if we perform adjustments to the brightness histogram to counteract the adobe tone curve, is that the same as using a linear profile, or will we introduce multiple corrections which introduce banding and other artifacts ?
Hi. Actual human (AH in stead of AI) here...
The "tone curve" you refer to is a fixed thing — it applies a specific curve to the image. It is essentially a sort of pre-set. The linear profile is just a different tone curve, and also a preset.
If those work, fine.
But the you "perform adjustments to the brightness histogram" you can customize that curve in far more powerful and tailored ways that go beyond what is possible using those preset profiles.
Note also that those profiles are affecting more than simply the luminosity curve. They also alter how color is handled.
When working with raw files you are unlikely to introduce "banding and other artifacts" unless you push those adjustments way beyond what these profiles do.
A demosaiced raw file is much darker in reality than it appears to be when viewed in a raw converter. The light balance is corrected with a so-called gamma curve for the image to look properly exposed. This is normal.
When more than one color profile is available in the raw converter, one may be preferred other the others for different reasons, the simplest being a personal preference.
I prefer reasonably flat color profiles, because of a preference for adding contrast when needed, rather than processing from a contrasty image by reducing contrast. However, this depends on the image, and some that have been taken in flat light and look desaturated may benefit from more contrast and saturation to begin with.
Different raw converters have different color profiles. The colors are also affected by White Balance that is not the same in different converters. This means that a raw file can look differently in different converters. You can look into which converters are praised for their colors, for example, and try these converters on your files. Although working with different apps can be a challenge, the basic tools of different raw converters are more or less the same; thus, after a day or two one can get a reasonably good sense of what a raw converter is capable of. In fact, different raw converters may have distinct strengths and weaknesses. I use several photography apps, for different reasons, and I believe this is not uncommon.
