Chapter 03: Adobe Camera Raw and the RAW Processing Stage

Learning Objectives

After completing this chapter, you will be able to:

Explain what a RAW file actually contains at the data level, and how it differs from a rendered image (JPEG, TIFF, PSD).
Describe the demosaicing process conceptually and explain why it is necessary.
Trace the RAW development pipeline from sensor data to rendered output: demosaic, white balance, tone mapping, color space assignment.
Distinguish between Adobe Camera Raw (the standalone RAW processor) and the Camera Raw Filter (which operates on already-rendered pixel data inside Photoshop).
Explain the relationship between ACR and Lightroom Classic — same processing engine, different workflow model.
Articulate the difference between parametric editing (the ACR/Lightroom model) and pixel editing (the Photoshop model), and explain why each is suited to different tasks.
Make an informed decision about when to process in ACR and when to move to Photoshop, based on what each tool can and cannot do.
Open a RAW file as a Smart Object in Photoshop and explain the practical benefits of doing so.
Explain what "highlight recovery" actually does — and what it cannot do.

Conceptual Foundation

Before a photograph reaches Photoshop, it passes through a critical processing stage that most photographers interact with daily but few understand at the data level. When you open a RAW file, Adobe Camera Raw (or Lightroom, or any other RAW converter) performs a series of irreversible mathematical transformations: demosaicing, white balancing, tone mapping, and color space conversion. The output of these transformations is the rendered image you work with in Photoshop.

This chapter is about that processing stage. Understanding it matters for two practical reasons.

First, decisions made during RAW development determine the quality and character of the data you bring into Photoshop. White balance choices, highlight recovery, and tone mapping are all more effective when performed on RAW sensor data than on a rendered image. Knowing what RAW development can do — and what it cannot — lets you place each adjustment where it will be most effective.

Second, Photoshop offers two interfaces that look nearly identical but operate on fundamentally different data: Adobe Camera Raw (which processes actual RAW sensor data) and the Camera Raw Filter (which applies the same interface to already-rendered pixel data). Confusing these two leads to incorrect assumptions about editing headroom and recovery capability.

The goal of this chapter is to give you a clear mental model of the RAW-to-Photoshop pipeline, so you understand what has already happened to your image data before you make your first Curves adjustment.

Terminology

RAW File

Definition: A RAW file contains minimally processed sensor data from a digital camera. It stores the output of the analog-to-digital converter (ADC) for each photosite on the sensor, along with metadata (camera settings, lens information, white balance as shot, etc.). A RAW file is not an image in the conventional sense — it is a record of sensor measurements that must be processed (developed) to become a viewable image.

Meaning in photographic practice: The RAW file preserves the maximum amount of data captured by the sensor. Because the data has not yet been rendered — no demosaicing, no tone curve, no white balance baked in, no lossy compression — a RAW file gives you the most flexibility in post-processing. White balance can be changed without penalty. Highlights and shadows have more recovery headroom than in a JPEG. Tonal adjustments operate on higher-precision data.

Photoshop implementation: Photoshop does not open RAW files directly into the pixel editor. When you open a RAW file (CR3, NEF, ARW, ORF, RAF, etc.) in Photoshop, Adobe Camera Raw launches automatically as an intermediary. You develop the RAW file in ACR, then hand the rendered result to Photoshop. Alternatively, you can open the RAW file as a Smart Object, which embeds the RAW data inside the Photoshop document for later re-development.

Related concepts: DNG, demosaicing, ACR, JPEG (as the rendered counterpart), Smart Object.

Common misconception: "A RAW file is like a digital negative — it captures everything the sensor saw." A RAW file captures everything the sensor recorded, but the sensor itself has limits. If a photosite is saturated (fully charged), the data is clipped at the hardware level. No RAW converter can recover detail that the sensor did not capture. The analogy to a film negative is imperfect: film negatives have a gradual shoulder in the highlights, while digital sensors clip abruptly.

Demosaicing / Debayering

Definition: Demosaicing (also called debayering) is the process of interpolating a full-color image from the incomplete color data captured by a sensor with a color filter array (CFA). Most digital cameras use a Bayer pattern CFA, in which each photosite captures only one color (red, green, or blue). Demosaicing algorithms estimate the missing two color values at each photosite based on the values of neighboring photosites.

Meaning in photographic practice: Your camera sensor does not capture color the way your eye perceives it. Each photosite records only one color channel. A "24-megapixel" sensor captures approximately 6 million red values, 12 million green values (green is doubled because human vision is most sensitive to green), and 6 million blue values, arranged in a repeating 2x2 pattern. The demosaicing algorithm reconstructs the missing information to produce an image where every pixel has all three color values.

Demosaicing quality affects fine detail rendering and can introduce artifacts at high-contrast edges (such as color fringing or moire). Different RAW converters use different demosaicing algorithms, which is one reason the same RAW file can look subtly different when developed in ACR vs Capture One vs DxO.

Photoshop implementation: Demosaicing is performed by Adobe Camera Raw (or Lightroom) during RAW development. It happens automatically and is not directly user-controllable — you do not choose a demosaicing algorithm in ACR. By the time the image reaches Photoshop's pixel editor, demosaicing is complete. Every pixel in the rendered image has full RGB values.

Related concepts: Bayer pattern, color filter array, RAW file, moire.

Common misconception: "Each pixel on my sensor captures full color." It does not. Each photosite captures one color. The full-color image is a computed reconstruction. This is why true optical resolution of a Bayer sensor is lower than the raw photosite count would suggest, and why some sensors (Foveon, for example) use alternative color capture methods.

White Balance (in RAW Context)

Definition: White balance is the adjustment of the relative amplitudes of the red, green, and blue channels to neutralize the color cast of the illuminant under which the scene was photographed. In a RAW file, white balance is stored as metadata (the "as shot" setting) but is not yet applied to the sensor data. The RAW converter applies white balance during development by scaling the channel values.

Meaning in photographic practice: Because RAW data has not been white-balanced at the data level, you can change white balance in a RAW converter without any quality loss — the scaling is applied to the original sensor values. This is fundamentally different from adjusting white balance on a rendered JPEG or TIFF, where the channel values have already been baked through demosaicing, tone mapping, and gamma encoding. Adjusting white balance on rendered data is essentially a color cast correction — it works, but it operates on lower-precision, already-processed data.

This is one of the strongest practical arguments for shooting RAW: the ability to set white balance with full precision in post-processing.

Photoshop implementation: White balance is set in Adobe Camera Raw using the Temperature (blue-yellow axis) and Tint (green-magenta axis) sliders in the Basic panel. ACR also provides presets (Daylight, Cloudy, Tungsten, etc.) and an eyedropper tool for clicking on a known neutral area. When working with a RAW file, these adjustments operate on the raw sensor data. When using the Camera Raw Filter on a rendered image inside Photoshop, the white balance sliders adjust already-rendered pixel values — the result is visually similar but mathematically different and has less headroom.

Related concepts: Color temperature, tint, color cast, illuminant, RAW file, Camera Raw Filter.

Common misconception: "White balance is just a color filter applied on top of the image." In a RAW workflow, white balance is applied before tone mapping, as a channel gain operation on the linear sensor data. This is structurally different from — and more precise than — applying a color shift to an already-rendered image. The visual result may look similar for small corrections, but for large white balance shifts (such as correcting a tungsten shot exposed under daylight settings), the RAW-level adjustment produces significantly better results.

Tone Mapping (RAW to Rendered)

Definition: In the RAW development context, tone mapping is the process of converting the linear sensor data (which may span 12-14+ stops of dynamic range) into a displayable image with a defined tonal range and gamma encoding. This involves applying a tone curve that compresses highlights, expands shadows, and establishes midtone contrast, producing an image that looks natural on a display.

Meaning in photographic practice: The linear sensor data in a RAW file looks extremely dark and flat when viewed directly, because most of the data values are concentrated in the highlights (in linear encoding, half of the available code values represent the brightest stop of exposure). Tone mapping redistributes these values to produce a visually natural rendering. The specific shape of the tone curve determines the character of the image — its contrast, shadow depth, and highlight rolloff.

This is why the same RAW file looks different when processed with different tone curves: Adobe's default rendering has a different character than Capture One's or the camera manufacturer's own processing. None of these is "correct" — they are different interpretations of the same sensor data.

Photoshop implementation: In Adobe Camera Raw, tone mapping is controlled primarily by the sliders in the Basic panel: Exposure, Contrast, Highlights, Shadows, Whites, and Blacks. The Tone Curve panel provides additional control over the shape of the mapping. These controls operate on the linear data during RAW development. Once the image reaches Photoshop, the tone mapping has been applied and the data is gamma-encoded — further tonal adjustments in Photoshop (Curves, Levels) operate on the rendered result, not on the linear sensor data.

Related concepts: Linear data, gamma encoding, tone curve, dynamic range, HDR.

Common misconception: "ACR's default rendering shows the 'true' image captured by the sensor." ACR's default rendering is one interpretation of the sensor data, using Adobe's default tone curve and color profile. The "true" sensor data is the linear, unrendered data — which looks nothing like a finished photograph. Every rendering involves creative and technical choices. ACR's defaults are designed to produce a generally pleasing result, but they are a starting point, not a ground truth.

Adobe Camera Raw (ACR)

Definition: Adobe Camera Raw is Adobe's RAW processing engine. It reads RAW sensor data from camera files, performs demosaicing, white balancing, tone mapping, noise reduction, lens corrections, and other processing, and outputs a rendered image. ACR can operate as a standalone processor (launching when you open a RAW file in Photoshop) or as the underlying engine within Lightroom Classic and Lightroom (cloud).

Meaning in photographic practice: ACR is the first stage of the Adobe photographic workflow. Every tonal, color, and detail adjustment you make in ACR operates on the RAW sensor data (or, more precisely, on the data at the point in the processing pipeline where that adjustment applies). This gives ACR adjustments more headroom and precision than the equivalent adjustments applied later in Photoshop to rendered pixel data.

ACR is designed for global and regional adjustments — the kind of processing that applies to the entire image or to broad areas defined by gradients, radial masks, or subject-aware selections. It excels at exposure correction, white balance, highlight and shadow recovery, noise reduction, lens correction, and overall color grading. It is not designed for pixel-level retouching, compositing, or precise local masking.

Photoshop implementation: ACR launches automatically when you open a RAW file in Photoshop (File > Open, then select a RAW file). It presents its own workspace with adjustment panels. When you click "Open" (or "Open Object" for Smart Object), ACR hands the developed image to Photoshop. ACR is also accessible through the Camera Raw Filter (Filter > Camera Raw Filter) for applying ACR-style adjustments to already-rendered layers within Photoshop.

Related concepts: Camera Raw Filter, Lightroom Classic, RAW file, Smart Object.

Common misconception: "ACR is a simple RAW previewer that just opens files into Photoshop." ACR is a full-featured image processor. For many photographs, the processing done in ACR may be all the editing the image needs. The decision to move into Photoshop should be deliberate — because Photoshop offers capabilities ACR does not (pixel-level retouching, compositing, precise masking) — not automatic.

Camera Raw Filter (in Photoshop)

Definition: The Camera Raw Filter is a Photoshop filter (Filter > Camera Raw Filter) that provides the ACR interface for editing pixel layers or Smart Objects within a Photoshop document. It offers the same sliders, panels, and tools as ACR, but it operates on already-rendered pixel data rather than on RAW sensor data.

Meaning in photographic practice: The Camera Raw Filter is useful for applying ACR-style adjustments to non-RAW data: scanned film, JPEGs, composited layers, or stamped visible layers. However, it does not provide the same editing headroom as ACR processing actual RAW data. Specifically, highlight and shadow recovery is limited to what is present in the rendered pixel values, not what exists in the original sensor data. White balance adjustments operate on gamma-encoded RGB values, not on linear sensor data.

Photoshop implementation: Accessible via Filter > Camera Raw Filter, or Shift+Ctrl+A (Shift+Cmd+A on macOS). It can be applied to a pixel layer (destructive) or to a Smart Object (as a re-editable Smart Filter). The interface is visually identical to the standalone ACR workspace, which can mislead users into thinking they have the same processing headroom.

Related concepts: ACR (the standalone processor), Smart Filter, rendered image.

Common misconception: "The Camera Raw Filter gives me the same RAW processing power as opening the file in ACR." It does not. The Camera Raw Filter operates on whatever pixel data the layer contains — which is already demosaiced, white-balanced, tone-mapped, and gamma-encoded. The sliders look the same and have the same names, but they are operating on fundamentally different data. Highlight recovery on a RAW file in ACR can pull back genuinely clipped-looking highlights because the linear sensor data may contain values above the rendered white point. The same slider on a Camera Raw Filter applied to a JPEG cannot recover data that was clipped during JPEG encoding — that data does not exist.

Lightroom Classic (Relationship to ACR)

Definition: Adobe Lightroom Classic is a photograph management and editing application that uses the same RAW processing engine as Adobe Camera Raw. The processing algorithms, slider behaviors, and rendering pipeline are shared between the two applications. They produce identical results from identical settings.

Meaning in photographic practice: Lightroom Classic and ACR are two interfaces to the same engine. A photograph processed in Lightroom Classic and then opened in Photoshop passes through the same rendering pipeline as one opened directly in ACR. The practical differences are in workflow, not processing:

Aspect	Adobe Camera Raw	Lightroom Classic
Primary role	RAW development for Photoshop handoff	Complete catalog, development, and output workflow
Image management	None (processes individual files)	Full catalog with keywords, collections, metadata
Batch processing	Limited (sync settings across selected files)	Full batch development, presets, virtual copies
Interface	Single-image workspace, modal dialog	Library + Develop modules, persistent workspace
Output to Photoshop	Direct — "Open Image" or "Open as Smart Object"	Photo > Edit In > Photoshop (round-trip workflow)
Processing engine	Same as Lightroom	Same as ACR

Photoshop implementation: When you send a photograph from Lightroom Classic to Photoshop (Photo > Edit In > Adobe Photoshop), Lightroom renders the image with its current development settings and opens it in Photoshop. If you choose "Edit Original" for a RAW file, it opens via ACR. If you choose "Edit a Copy with Lightroom Classic Adjustments," Lightroom renders a TIFF or PSD with the adjustments applied. The processing is identical to ACR because the engine is the same.

Related concepts: ACR, parametric editing, catalog, virtual copies.

Common misconception: "Lightroom produces different results than Camera Raw." Given the same processing settings and the same ACR/Lightroom version, the rendering is identical. Perceived differences usually arise from different default settings, different versions of ACR/Lightroom being compared, or different profiles being applied. The processing engine is shared code.

DNG (Digital Negative)

Definition: DNG (Digital Negative) is an open, documented RAW file format created by Adobe. It is designed to serve as a universal RAW format, providing a standardized container for sensor data that would otherwise be stored in proprietary camera-specific formats (CR3, NEF, ARW, ORF, RAF, etc.).

Meaning in photographic practice: DNG addresses a long-term archival concern: proprietary RAW formats depend on the camera manufacturer's continued support. If a manufacturer stops updating their format or goes out of business, future software may not be able to read the files. DNG, being documented and open, mitigates this risk.

DNG files can embed the original proprietary RAW data (for maximum compatibility), or they can convert the data to the DNG structure (reducing file size but discarding the proprietary format). DNG can also embed development settings directly in the file, eliminating the need for separate XMP sidecar files.

Photoshop implementation: Photoshop and ACR read DNG files natively. You can convert proprietary RAW files to DNG using the free Adobe DNG Converter application, or within Lightroom Classic (on import or via Library > Convert Photo to DNG). ACR and Lightroom can also save development settings directly into DNG files.

Related concepts: RAW file, XMP sidecar, archival storage, ACR.

Common misconception: "Converting to DNG improves image quality." DNG conversion does not alter the sensor data or improve quality. It changes the container format. If you convert without embedding the original RAW file, you lose the proprietary format (which may contain manufacturer-specific metadata or processing options used by the manufacturer's own software). The quality of the image data inside the DNG is identical to the original RAW file.

XMP Sidecar

Definition: An XMP sidecar file is a small XML-based file (with .xmp extension) that stores development settings, metadata edits, and keywords for a RAW file. It is saved alongside the RAW file on disk, with the same base filename (e.g., DSC_1234.NEF accompanied by DSC_1234.xmp).

Meaning in photographic practice: Because RAW files are read-only data records, development adjustments cannot be written into the RAW file itself (with the exception of DNG, which can embed settings). Instead, ACR and Lightroom store your processing instructions — slider positions, crop, masks, local adjustments — in a sidecar file. The RAW data is never modified. When you reopen the RAW file, ACR reads the sidecar and applies the stored settings.

This means your development work is dependent on the sidecar file. If you copy a RAW file without its sidecar, the development settings are lost (the RAW file opens with default settings). If you use Lightroom Classic, the settings are stored in the Lightroom catalog database, and can optionally be written to XMP sidecars (Catalog Settings > Metadata > Automatically write changes into XMP sidecar files).

Photoshop implementation: When you adjust a RAW file in ACR and click "Done" (without opening into Photoshop), ACR writes an XMP sidecar file. When you reopen the RAW file, ACR reads the sidecar and restores your settings. For DNG files, settings can be written directly into the file rather than to a sidecar.

Related concepts: DNG, RAW file, ACR, Lightroom catalog, parametric editing.

Common misconception: "XMP files are just metadata — they are not important." XMP sidecar files contain your entire development recipe — every slider position, every mask, every local adjustment. Losing the sidecar means losing all your processing work. Always include XMP files when backing up, moving, or archiving RAW files.

Smart Object from RAW

Definition: A Smart Object that contains embedded RAW data along with the ACR development settings used to render it. Double-clicking the Smart Object thumbnail in Photoshop reopens ACR with the full RAW data and all processing controls, allowing you to re-develop the image without losing any work done in Photoshop above the Smart Object layer.

Meaning in photographic practice: Opening a RAW file as a Smart Object is the most flexible way to bring RAW data into Photoshop. It gives you the full power of ACR's RAW processing at any point during your Photoshop editing session — you can change white balance, re-do highlight recovery, modify the tone curve, or adjust noise reduction, and the changes propagate to the Smart Object while all Photoshop layers above it remain intact.

The trade-off is file size: the PSD stores the full RAW file data inside the Smart Object, in addition to all Photoshop layer data. For a 50 MB RAW file, this adds approximately 50 MB to the PSD.

Photoshop implementation: In Adobe Camera Raw, hold Shift to change the "Open" button to "Open Object," then click. Alternatively, configure ACR's workflow settings to always open as Smart Object: click the workflow link at the bottom of the ACR window (the line showing color space, bit depth, and resolution) and check "Open in Photoshop as Smart Objects."

VERSION NOTE: The exact location of the "Open as Smart Object" option and the workflow settings link has changed across ACR versions. In recent versions, the workflow options may be accessible via a gear icon or a settings link at the bottom of the ACR workspace. Verify the location in your version.

Related concepts: Smart Object (Chapter 04), non-destructive editing (Chapter 06), ACR.

Common misconception: "Opening as Smart Object and opening normally give the same result — I can always re-edit later." If you open with "Open Image" (not as Smart Object), ACR renders the image and hands flat pixel data to Photoshop. The RAW data is not embedded. You cannot reopen ACR with the RAW data from within Photoshop — the connection to the RAW file is severed. You would need to close the Photoshop document, reopen the original RAW file in ACR, re-develop it, and re-do your Photoshop work. Smart Objects avoid this entirely.

RAW Data vs Rendered Image

Definition: RAW data is the unprocessed (or minimally processed) sensor output stored in a RAW file. It is linear, single-channel-per-pixel (before demosaicing), and represents physical light intensity measurements. A rendered image is the result of processing that RAW data through demosaicing, white balancing, tone mapping, and color space conversion — producing a gamma-encoded, full-color, displayable image (JPEG, TIFF, PSD, or the pixel data inside a Photoshop layer).

Meaning in photographic practice: This distinction is the most fundamental concept in this chapter. Every adjustment you make in ACR on a RAW file operates on (or near) the RAW data level, with access to the full precision and dynamic range captured by the sensor. Every adjustment you make in Photoshop on a pixel layer operates on rendered data — data that has already passed through the RAW development pipeline and lost some of its original flexibility.

This does not mean Photoshop adjustments are inferior for all tasks. It means they are operating on different data, with different characteristics and limitations. The skill is knowing where in the pipeline each adjustment belongs.

Photoshop implementation: When you see pixel values in the Info panel in Photoshop, you are seeing rendered values — gamma-encoded, in the document's color space, after demosaicing and tone mapping. When you see values in ACR's interface, you are interacting with a processed representation of the underlying RAW data. The distinction is invisible in the interface but fundamental to understanding editing headroom.

Related concepts: RAW file, demosaicing, tone mapping, gamma encoding (Chapter 01), ACR, Camera Raw Filter.

Common misconception: "Once I open a RAW file in ACR, I'm working with the RAW data, so everything I do in ACR is 'RAW-quality' editing." This is partially true but imprecise. ACR's adjustments operate at different stages of the processing pipeline. White balance operates on linear data (early in the pipeline). Sharpening operates on demosaiced, tone-mapped data (late in the pipeline). The headroom advantage of RAW processing varies depending on which adjustment you are making and where in the pipeline it applies.

Parametric Editing (ACR / Lightroom Model)

Definition: Parametric editing is a non-destructive editing model in which all adjustments are stored as a list of parameters (slider values, mask definitions, crop coordinates) rather than being applied directly to the image data. The image is re-rendered from the original data each time parameters change. The original data is never modified.

Meaning in photographic practice: In ACR and Lightroom, you never edit pixels. You edit parameters. When you move the Exposure slider to +0.5, ACR stores "Exposure: +0.5" in the settings and re-renders the image from the RAW data with that parameter applied. If you later change it to +0.7, the original rendering at +0.5 is not modified — the parameter is simply updated and the image is re-rendered from scratch.

This is fundamentally different from Photoshop's pixel editing model, where each adjustment reads pixel values, transforms them, and writes new values. In pixel editing, sequential adjustments compound — each one operates on the output of the previous one. In parametric editing, all parameters are applied together during a single rendering pass from the original data.

Photoshop implementation: Photoshop is primarily a pixel editor, not a parametric editor. However, Adjustment Layers approximate parametric behavior within Photoshop's compositing model: they store parameters and apply them non-destructively at render time. The key difference is that Adjustment Layers operate on already-rendered data (from the layers below), while ACR's parametric editing operates on the original RAW sensor data.

Related concepts: Non-destructive editing (Chapter 06), Adjustment Layers (Chapter 04), RAW file, XMP sidecar.

Common misconception: "Parametric editing and non-destructive editing are the same thing." Parametric editing is a specific form of non-destructive editing. Photoshop's Adjustment Layers are also non-destructive, but they are not parametric in the same sense — they do not re-render from the original capture data. They store transformation parameters that are applied to the composited result of the layers below them. The distinction matters because parametric rendering from RAW data provides more headroom than non-destructive adjustments applied to rendered data.

Theory

What a RAW File Actually Contains

FACT: A RAW file stores the digitized output of each photosite on the camera sensor, after analog-to-digital conversion but before any image-processing algorithms (demosaicing, white balance, tone mapping, noise reduction, sharpening). The data is linear — proportional to the number of photons captured — and each photosite records only one color channel, determined by the color filter array above it.

The most common color filter array is the Bayer pattern, a repeating 2x2 grid:

G	R
B	G

In this pattern, 50% of photosites are green-filtered, 25% are red-filtered, and 25% are blue-filtered. Green is overrepresented because human vision is most sensitive to green wavelengths, so allocating more samples to green improves the perceived quality of the demosaiced image.

FACT: Typical camera ADCs produce 12-bit or 14-bit values per photosite. A 14-bit ADC outputs values in the range 0 to 16,383. The actual usable range is smaller due to sensor noise at the low end and the analog saturation point at the high end, but the full bit depth of the ADC is stored in the RAW file.

FACT: A RAW file also contains metadata: the "as shot" white balance, exposure settings (shutter speed, aperture, ISO), lens data, camera model, and a JPEG preview (the image you see on the camera's LCD). The JPEG preview is generated by the camera's internal processor using the camera's picture style settings — it is not the RAW data. This preview is what makes the image appear "correct" on the camera screen, even though the underlying data has not been processed.

The Demosaicing Process

Because each photosite captures only one color, the raw sensor output is a mosaic of single-color values, not a full-color image. The demosaicing algorithm must estimate the missing two color values at each photosite.

FACT: At its simplest level, demosaicing works by examining the values of neighboring photosites. At a green photosite, the algorithm has a measured green value and must estimate red and blue by looking at nearby red and blue photosites. More sophisticated algorithms also consider edge direction (to avoid smearing detail across edges) and use larger neighborhoods of photosites for better accuracy.

INTERPRETATION: The quality of the demosaicing algorithm affects the rendering of fine detail. Simple bilinear interpolation produces soft results with color artifacts at edges. Advanced algorithms (such as those used in ACR, which Adobe does not fully document) produce sharper detail with fewer artifacts. This is one of the technical reasons why different RAW converters produce visibly different results from the same RAW file — they use different demosaicing algorithms.

FACT: Demosaicing is irreversible in the sense that it is a reconstruction, not a lossless transformation. The full-color image produced by demosaicing is an estimate of what the scene looked like, based on incomplete samples. The quality of this estimate is generally excellent with modern algorithms, but it is inherently limited by the spatial sampling density of the sensor and the assumptions built into the algorithm.

The RAW Development Pipeline

When ACR (or Lightroom, or any RAW converter) processes a RAW file, it applies a series of transformations in a specific order. While Adobe does not publicly document the exact internal pipeline of ACR, the general stages are well-understood:

1. Linearization and black-level subtraction. The raw ADC values are linearized (correcting any nonlinearity in the ADC) and the black level (the signal recorded with zero light) is subtracted. The result is linear data proportional to actual light captured.

2. White balance. The red, green, and blue channels are scaled by different factors to compensate for the color temperature and tint of the illuminant. In linear data, this is a simple multiplication — the green channel might be left at 1.0x while the red channel is scaled to 1.3x and blue to 0.8x (the specific factors depend on the illuminant). This is why white balance adjustment on RAW data is essentially free — it is just changing the scaling factors before any other processing occurs.

3. Demosaicing. The single-channel-per-pixel data is interpolated into full RGB values at every pixel, using the color filter array pattern and a demosaicing algorithm.

4. Color space conversion. The camera-specific RGB values (which depend on the spectral sensitivity of the particular sensor's color filters) are converted to a standard color space for editing and display (such as ProPhoto RGB, Adobe RGB, or sRGB, depending on the output settings).

5. Tone mapping and gamma encoding. The linear data is mapped through a tone curve that establishes the overall contrast, brightness, and shadow/highlight rendering. The gamma encoding (transfer function) is applied, converting from linear light values to perceptually encoded values suitable for display and further editing.

6. Additional processing. Depending on user settings, this stage includes noise reduction, sharpening, lens corrections (distortion, chromatic aberration, vignetting), and any local adjustments (graduated filters, radial filters, brush adjustments).

7. Output. The processed image is delivered to Photoshop (or saved as JPEG, TIFF, DNG, etc.) in the specified color space and bit depth.

INTERPRETATION: The order of these stages matters. White balance operates on linear data before demosaicing (or during — the exact interleaving depends on the implementation). This is why it has essentially unlimited headroom in a RAW workflow. Sharpening and noise reduction operate on fully demosaiced, tone-mapped data — they are "late" in the pipeline and have the same limitations they would have on any rendered image. Highlight and shadow recovery operate on the linear data before tone mapping, which is why they can pull back values that appear clipped in the rendered preview.

Why RAW Has More Editing Headroom Than JPEG

The editing headroom advantage of RAW is not a vague quality improvement. It comes from specific, measurable differences in the data:

FACT: A RAW file from a 14-bit camera stores 16,384 values per channel per photosite, in a linear encoding. A JPEG stores 256 values per channel per pixel, in a gamma-encoded format.

FACT: In linear encoding, each additional stop of exposure doubles the number of photons and thus the signal level. This means the highest stop of exposure occupies half of the available code values (8,192 out of 16,384 for 14-bit), the second-highest stop occupies a quarter (4,096), and so on. The bottom stops have very few code values each. This is not a deficiency — it reflects the fact that the signal-to-noise ratio is highest in the brightest stops, so allocating more values there preserves useful information.

FACT: When ACR adjusts exposure on RAW data, it operates on these linear values before tone mapping. Pushing exposure by +1 stop on RAW data is equivalent to multiplying all values by 2 — a clean scaling operation that introduces no quantization error (though it amplifies noise in the shadows). Pushing exposure by +1 stop on a rendered JPEG requires remapping gamma-encoded 8-bit values, which stretches a limited number of code values across a larger range, creating gaps (posterization).

FACT: A JPEG from a camera has already been processed by the camera's internal image processor: demosaiced, white-balanced, tone-mapped with the camera's picture style, gamma-encoded, and then lossy-compressed. The data has been through multiple irreversible transformations, each of which discards some information. A RAW file has been through none of these.

What "Highlight Recovery" Actually Does

FACT: In a RAW file, the three color channels may clip at different points. The red channel might saturate before the green channel, for example. When the rendered preview shows a blown highlight (white, all channels clipped), it is possible that one or two channels still contain unclipped data.

FACT: ACR's highlight recovery (the Highlights slider pulled to the left, or the older Recovery slider in earlier versions) works primarily by using the data from the unclipped channel(s) to reconstruct plausible values for the clipped channel(s). If the red channel is clipped but the green and blue channels are not, ACR can estimate what the red channel value should have been, based on the known color relationships at that location.

INTERPRETATION: This is genuine recovery — the result is based on real sensor data, not fabrication. However, it has strict limits:

If all three channels are clipped (the photosite reached full saturation in all colors), there is no data to recover. The sensor recorded maximum value in all channels, and no algorithm can determine what the values would have been. The result is featureless white. This cannot be fixed.
Recovery quality degrades as more channels clip. When two channels are clipped and one remains, the reconstruction is based on minimal information and may show color inaccuracies.
Clipping in the RAW data is absolute. The ADC has a maximum output value. Values at or above the saturation point are all recorded as the same maximum. There is no "hidden" data above the clipping point.

RECOMMENDATION: Do not confuse highlight recovery with highlight creation. ACR can pull back highlights that appear clipped in the default rendering but have usable data in one or more channels. It cannot create detail that the sensor did not capture. The solution to genuinely clipped highlights is better exposure, not post-processing.

Photoshop Implementation

The ACR Workspace

When you open a RAW file in Photoshop, Adobe Camera Raw launches in its own workspace before Photoshop's main editor. The workspace is organized into panels covering different aspects of RAW development.

VERSION NOTE: ACR's interface has undergone significant redesign over the years. Panel names, locations, and the available tools change between major versions. The following describes the general structure common to recent ACR versions, but you should verify specific panel locations in your installed version.

The primary editing panels include:

Panel	Purpose
Basic	Exposure, contrast, highlights, shadows, whites, blacks, white balance, clarity, vibrance, saturation
Tone Curve	Fine control over the tone mapping curve — parametric and point curve
Detail	Sharpening and noise reduction
Color Mixer (HSL)	Hue, saturation, and luminance per color range
Color Grading	Split-toning — color tints applied to shadows, midtones, and highlights
Optics (Lens Corrections)	Distortion correction, chromatic aberration removal, vignette removal
Geometry (Transform)	Perspective correction, guided upright
Effects	Post-crop vignetting, grain
Calibration	Camera profile and per-channel hue/saturation fine-tuning

FACT: ACR provides local adjustment tools — graduated filter, radial filter, and adjustment brush — that allow you to apply different settings to different areas of the image. These are parametric: each local adjustment stores its own set of slider values and a mask defining where it applies.

VERSION NOTE: Recent versions of ACR and Lightroom have introduced AI-based masking capabilities, including "Select Subject" and "Select Sky" masks that automatically detect and select subjects or sky regions. These features are actively evolving. Verify the specific masking tools available in your version.

The Camera Raw Filter

The Camera Raw Filter (Filter > Camera Raw Filter, or Shift+Ctrl+A / Shift+Cmd+A) provides the same visual interface as ACR, but within a Photoshop document, operating on pixel data rather than RAW sensor data.

FACT: The Camera Raw Filter presents the same panels and sliders as the standalone ACR workspace. The critical difference is the data being processed: the Camera Raw Filter operates on the gamma-encoded, fully rendered pixel values of the target layer.

Practical implications of this difference:

Capability	ACR on RAW File	Camera Raw Filter on Pixel Layer
White balance precision	Full — operates on linear data	Limited — adjusts gamma-encoded values
Highlight recovery	Can recover from unclipped channels	Limited to data present in pixel values
Shadow recovery	Full dynamic range of sensor	Limited to data present in pixel values
Noise reduction	Access to pre-demosaic data characteristics	Operates on rendered data
Lens corrections	Access to lens metadata, raw distortion data	Can apply profile corrections to rendered data
Quality ceiling	Maximum (processing from sensor data)	Good, but constrained by input data quality

RECOMMENDATION: Use the Camera Raw Filter when you want ACR's slider interface for adjusting a non-RAW layer (a scanned image, a composited result, a JPEG). Do not mistake it for a substitute for actual RAW processing. If you have the original RAW file, always process it through ACR (or Lightroom) rather than converting to JPEG and then using the Camera Raw Filter.

When applied to a Smart Object, the Camera Raw Filter becomes a Smart Filter — re-editable and non-destructive within the Photoshop document.

Opening Options: "Open Image" vs "Open as Smart Object"

When you have finished developing a RAW file in ACR, you have two primary options for handing the result to Photoshop:

Open Image (the default button): ACR renders the image at the specified bit depth and color space and opens it in Photoshop as a regular pixel layer (the Background layer). The RAW data is not embedded in the Photoshop document. The connection to the original RAW file is not maintained within the document. The development is baked into the pixel values.

Open Object (hold Shift to toggle the button, or configure in workflow settings): ACR opens the image in Photoshop as a Smart Object. The RAW file data and the current ACR development settings are embedded inside the Smart Object. Double-clicking the Smart Object thumbnail later will reopen ACR with the full RAW data, allowing you to re-develop from scratch.

Property	Open Image	Open as Smart Object
RAW data in PSD	No	Yes (embedded in Smart Object)
Can re-develop RAW later	No (must reopen original file)	Yes (double-click Smart Object)
Can paint directly on layer	Yes	No (must use separate layers)
File size impact	Standard	Larger (RAW data + layer data)
Flexibility	Lower — development is final	Higher — development is re-editable
Workflow complexity	Simpler	Slightly more complex

RECOMMENDATION: For images that will receive significant Photoshop work, open as Smart Object. The ability to re-develop the RAW data at any point during your editing session — without losing your Photoshop layer stack — is a substantial workflow advantage that justifies the file size increase. For images that need only minor Photoshop work (a quick crop or a single retouching pass), opening normally may be sufficient.

Professional Workflow: RAW to ACR to Photoshop Pipeline

The Pipeline Model

A professional photographic editing workflow typically follows this pipeline:

RAW File  -->  ACR / Lightroom  -->  Photoshop  -->  Output
                (RAW processing)     (pixel editing)   (export)

Each stage handles the work it is best suited for:

Stage 1 — RAW Development (ACR or Lightroom):

White balance correction or creative white balance choice
Exposure correction (global)
Highlight and shadow recovery
Black and white point setting
Overall contrast (Tone Curve)
Noise reduction (operates on data closer to the sensor level)
Lens corrections (distortion, chromatic aberration, vignetting)
Color grading (Color Mixer/HSL, Color Grading panel)
Sharpening (capture sharpening — compensating for the inherent softening of the demosaicing and anti-aliasing filter)

Stage 2 — Photoshop:

Pixel-level retouching (clone stamp, healing brush, content-aware fill)
Compositing (combining multiple images or elements)
Precise local adjustments with complex masks (luminosity masks, channel-based selections)
Frequency separation for skin retouching
Advanced dodging and burning
Creative effects requiring pixel manipulation
Output sharpening (sized and targeted for the specific output medium)

Stage 3 — Output:

Resizing for the output medium
Output sharpening (if not done in Stage 2)
Color space conversion for the output medium (sRGB for web, printer profile for print)
File format conversion (JPEG, TIFF, PSD)

What Belongs in ACR vs What Belongs in Photoshop

The decision of where to apply a given adjustment is not arbitrary. It follows from understanding the data each tool operates on:

Adjustment	ACR / Lightroom	Photoshop	Why
White balance	Preferred	Possible but inferior	ACR operates on linear data; Photoshop adjusts gamma-encoded values
Exposure correction	Preferred	Possible	ACR operates on full dynamic range of sensor data
Highlight recovery	ACR only (for RAW headroom)	Limited (data already clipped in rendered image)	Recovery requires access to unclipped channel data
Shadow recovery	Preferred	Possible	ACR has more headroom in linear shadows
Noise reduction	Preferred for global NR	Use for local/targeted NR	ACR has access to data characteristics closer to sensor
Lens corrections	Preferred	Possible via Adaptive Wide Angle or manual	ACR accesses lens metadata directly
Retouching (healing, cloning)	Basic (ACR has limited tools)	Preferred	Photoshop's retouching tools are far more precise and flexible
Complex masking	Limited (gradient, radial, brush, subject/sky)	Full capability	Photoshop provides channels, paths, luminosity masks, Select and Mask
Compositing	Not possible	Photoshop only	ACR processes single images
Layer-based editing	Not possible	Photoshop only	ACR has no layer system
Precise local tonal control	Limited	Full (Curves + masks, dodge/burn layers)	Photoshop's masking precision is far greater

RECOMMENDATION: Do as much as you can in ACR, then move to Photoshop for what ACR cannot do. This is not a rule of purity — it is a data quality strategy. Processing that benefits from RAW headroom (white balance, exposure, recovery) should happen on the RAW data. Processing that requires pixel-level control (retouching, compositing, complex masking) should happen in Photoshop.

Practical Workflow Example

The following sequence illustrates a typical professional workflow from capture to output:

Import and cull in Lightroom Classic (or your file management tool).
Develop in ACR/Lightroom: Set white balance. Correct exposure. Recover highlights and shadows. Set black and white points. Adjust overall contrast with the Tone Curve. Apply lens corrections. Apply global noise reduction. Perform capture sharpening. Make any global color adjustments (HSL, Color Grading).
Open in Photoshop as Smart Object (for images requiring Photoshop work).
Build layer stack in Photoshop (Chapter 04 architecture): retouching layers, tonal adjustment layers, color adjustment layers, local adjustments, output preparation.
If RAW development needs adjustment: Double-click the Smart Object to reopen ACR, modify settings, click OK. The Photoshop layer stack remains intact.
Output: Save the layered PSD as the master file. Export flattened copies (JPEG, TIFF) for specific output targets (web, print, client delivery).

The Round-Trip Workflow (Lightroom to Photoshop and Back)

If you use Lightroom Classic as your primary editing environment, the round-trip workflow adds an additional consideration:

Develop the RAW file in Lightroom Classic.
Send to Photoshop: Photo > Edit In > Adobe Photoshop. Lightroom renders the image and opens it in Photoshop.
Perform Photoshop work.
Save the file (Ctrl+S / Cmd+S). Photoshop saves a TIFF or PSD.
The saved file automatically appears in Lightroom's catalog alongside the original RAW file.

FACT: In the Lightroom round-trip, the file sent to Photoshop is a rendered image (not a RAW Smart Object), unless you specifically choose "Open as Smart Object in Photoshop" from the Photo menu or use other workflows to maintain the RAW connection. The standard "Edit In" command renders the image with Lightroom's current settings.

RECOMMENDATION: If you need to re-develop the RAW data during Photoshop editing, open the RAW file as a Smart Object (either from ACR directly, or from Lightroom via Photo > Edit In > Open as Smart Object in Photoshop, if available in your version). The standard round-trip renders the image before Photoshop receives it.

Common Mistakes

Confusing the Camera Raw Filter with RAW Processing

The mistake: Applying the Camera Raw Filter to a JPEG or TIFF layer in Photoshop and expecting the same highlight/shadow recovery headroom as processing the original RAW file in ACR.

Why it matters: The Camera Raw Filter operates on rendered pixel values. If highlights are clipped in the layer data, the Highlights slider cannot recover detail that does not exist. Users who see the familiar ACR interface assume they have full RAW processing power, but they do not.

The fix: When you have the original RAW file, always process it through ACR (or Lightroom) rather than through the Camera Raw Filter on a rendered derivative. Reserve the Camera Raw Filter for situations where you do not have access to the original RAW data.

Opening RAW Files as Regular Pixel Layers Instead of Smart Objects

The mistake: Clicking "Open Image" in ACR without considering whether "Open Object" (Smart Object) would be more appropriate.

Why it matters: Once the RAW file is rendered as a pixel layer, the connection to the RAW data is severed. If you later discover that the white balance is wrong, the exposure needs adjustment, or the noise reduction was too aggressive, you must close the Photoshop document, reopen the RAW file, re-develop it, and redo your Photoshop work.

The fix: Default to opening as Smart Object for images that will receive significant Photoshop editing. Configure ACR's workflow settings to open as Smart Object by default if this matches your workflow.

Performing All Processing in Either ACR or Photoshop, but Not Both

The mistake: Either (a) over-processing in ACR — trying to do precise local retouching with ACR's limited tools — or (b) rushing through ACR and trying to do all tonal work in Photoshop on rendered data.

Why it matters: Each tool has strengths that complement the other. ACR excels at global tonal and color work on RAW data. Photoshop excels at precise local control and pixel-level editing. Using only one means sacrificing either RAW headroom or editing precision.

The fix: Follow the pipeline model. Use ACR for what it does best (global RAW processing), then move to Photoshop for what it does best (local control, retouching, compositing).

Ignoring XMP Sidecar Files in Backup and File Management

The mistake: Backing up or moving RAW files without their accompanying XMP sidecar files. When the RAW files are reopened, all development settings are gone and ACR shows the default rendering.

Why it matters: XMP sidecar files contain the complete record of your development work. Losing them means losing hours of processing effort.

The fix: Always include XMP sidecar files when copying, moving, or archiving RAW files. If you use Lightroom Classic, enable "Automatically write changes into XMP sidecar files" in Catalog Settings > Metadata, or perform an explicit "Save Metadata to Files" operation before moving files.

Assuming ACR Can Recover Any Highlight

The mistake: Overexposing in the field because "I can always pull it back in Camera Raw."

Why it matters: ACR can recover highlights where one or more channels have unclipped data. If all three channels are clipped at the sensor level, the data does not exist and cannot be recovered. The extent of recoverable range varies by camera and depends on how the specific sensor handles saturation.

The fix: Expose conservatively for highlights. Use the camera's histogram and highlight clipping indicators during capture. Treat ACR's highlight recovery as a safety margin, not a substitute for correct exposure.

Converting RAW to JPEG Before Editing

The mistake: Converting RAW files to JPEG in the camera, in a batch export, or by saving from ACR as JPEG before performing editing work.

Why it matters: JPEG conversion applies tone mapping, gamma encoding, chroma subsampling, and lossy compression. All of the RAW processing advantages — white balance flexibility, highlight/shadow recovery headroom, full bit depth — are lost. Editing a JPEG is editing rendered, compressed, 8-bit data.

The fix: Keep RAW files as your source throughout the editing process. Convert to JPEG only as the final export step for specific output targets (web delivery, client preview).

Summary

A RAW file is not an image — it is a record of sensor measurements. Each photosite in the sensor captures a single-color intensity value through a color filter array. The RAW file stores these values in a linear encoding at 12-bit or 14-bit precision, along with metadata. To produce a viewable image, the RAW data must pass through a development pipeline: linearization, white balance, demosaicing, color space conversion, tone mapping, and gamma encoding.

Adobe Camera Raw (ACR) and Lightroom Classic share the same processing engine and perform this development. ACR operates as the gateway between RAW files and Photoshop; Lightroom provides the same processing within a catalog-based workflow. The development settings are stored parametrically — as slider values, not as pixel modifications — and the original RAW data is never altered.

The Camera Raw Filter in Photoshop provides the ACR interface for editing pixel layers, but it operates on already-rendered data. It does not provide the same processing headroom as ACR operating on RAW sensor data. This distinction is invisible in the interface but fundamental to understanding what each tool can and cannot recover.

Opening a RAW file as a Smart Object embeds the RAW data in the Photoshop document, allowing you to re-develop the image at any point without losing your Photoshop work. Opening as a regular pixel layer severs the connection to the RAW data.

The professional workflow places each adjustment where it is most effective: global tonal and color corrections in ACR (where RAW headroom is available), and pixel-level retouching, compositing, and precise local control in Photoshop (where these capabilities exist). This is not a rigid boundary but a data-quality strategy — processing on RAW data where precision matters most, and moving to pixel editing where control matters most.

Exercises

Exercise 1: RAW vs JPEG Editing Headroom

Open the same photograph as both a RAW file (through ACR) and as the camera-generated JPEG. In ACR, push the Exposure slider to +2.0 and then to -2.0 on the RAW file. Then open the JPEG in Photoshop, apply the Camera Raw Filter, and push the same Exposure slider to +2.0 and then to -2.0.

Compare the results at +2.0. Where does the JPEG show banding, noise, or color breakdown that the RAW version does not?
Compare the results at -2.0. What is the quality difference in the recovered shadow detail?
Reset both to 0. Now change the white balance dramatically on both (e.g., from Daylight to Tungsten). Compare the color accuracy and smoothness of the result.

The goal is to see — not just understand conceptually — the difference in editing headroom between RAW and rendered data.

Exercise 2: ACR vs Camera Raw Filter Comparison

Open a RAW file in ACR with intentionally clipped highlights (a bright sky, a light source in frame). Use the Highlights slider to recover as much detail as possible. Note the result.

Now click "Open Image" to bring the developed image into Photoshop as a pixel layer. Apply the Camera Raw Filter (Filter > Camera Raw Filter) and use the Highlights slider to attempt the same recovery.

How do the results compare? Can the Camera Raw Filter recover the same detail?
Why does this difference exist?

Exercise 3: Open Image vs Open as Smart Object

Open a RAW file in ACR. Process it with a specific white balance (e.g., Daylight) and moderate exposure correction.

Open as Smart Object. In Photoshop, add a Curves Adjustment Layer and a retouching layer with some clone stamp work.
Now double-click the Smart Object thumbnail. ACR reopens with the full RAW data. Change the white balance to Tungsten. Click OK.
Observe: the Photoshop layers above (Curves, retouching) are intact. The base image has changed.
Undo the ACR change to return to the original development.

This exercise demonstrates why Smart Objects from RAW are valuable: you can revisit RAW development decisions without losing Photoshop work.

Exercise 4: XMP Sidecar Investigation

Open a RAW file in ACR. Make several adjustments: change white balance, adjust exposure, apply a graduated filter. Click "Done" (not "Open").

Navigate to the RAW file's location on disk. Find the XMP sidecar file with the same base name and .xmp extension.
Open the XMP file in a text editor. Identify the stored settings — you should be able to find your white balance, exposure, and other adjustments stored as named parameters.
Rename or move the XMP file. Reopen the RAW file in ACR. Observe that all your adjustments are gone — ACR shows the default rendering.
Move the XMP file back to its original location and name. Reopen the RAW file. Your adjustments are restored.

This exercise demonstrates that your RAW development work lives in the XMP sidecar file, not in the RAW file itself.

Advanced Exercises

Advanced Exercise 1: Highlight Recovery Analysis

Find or create a photograph with partially clipped highlights — bright enough that some channels clip but not all three. A bright sky with some cloud detail, or a window in an interior shot, works well.

Open the RAW file in ACR. Before any adjustments, examine the histogram. Identify whether clipping is occurring and in which channels.
Pull the Highlights slider to -100. Observe the recovered detail.
Now examine the recovered area closely. Is the color accurate, or does it show a color cast? Which channels were clipped, and how does that relate to the color of the recovered area?
Find an area where all three channels are clipped (if one exists). Can ACR recover any detail there? Why or why not?
Reset to defaults. Now deliberately overexpose the image in ACR by pushing Exposure to +3.0. How much of this can you recover with the Highlights slider? At what point does the recovery become ineffective?

Advanced Exercise 2: Pipeline Stage Comparison

Open a RAW file in ACR. Make two versions:

Version A: Apply all tonal corrections in ACR — set exposure, contrast, highlights, shadows, whites, blacks, and tone curve to produce a finished-looking image. Open into Photoshop with minimal further adjustment needed.

Version B: Open the RAW file with default (zeroed) ACR settings into Photoshop. Apply equivalent tonal corrections using Curves Adjustment Layers on the rendered data.

Compare the two results at 100% zoom. Are there visible quality differences? Where?
Examine the histograms of both results. Which shows smoother distribution? Which shows more gaps?
If you were working with a 14-bit RAW from a modern camera, is the quality difference practically significant for your output medium (web? print?)?

This exercise tests whether the theoretical headroom advantage of RAW processing translates to a visible difference in your specific workflow.

Blackbelt Challenge

A client provides you with a set of photographs from an architectural interior shoot. The set includes:

RAW files from a camera you support in your version of ACR
Camera-generated JPEGs of the same shots
Some shots have mixed lighting: daylight from windows and tungsten from interior fixtures

Your task (analyze without opening Photoshop):

For the mixed-lighting shots, explain the white balance problem at the data level. Why can you not simply "correct" white balance to make both light sources appear neutral? What does the RAW development pipeline allow you to do about this, and what would require Photoshop?
The client wants one shot where the window view is properly exposed and the interior is properly exposed, but the dynamic range exceeds the sensor's capture in a single frame. The client shot brackets. Describe the complete pipeline: how would you process and combine the brackets, and at which stage of the pipeline (ACR, Photoshop, or both) does each operation belong? Which operations benefit from RAW headroom, and which require Photoshop's compositing capabilities?
Several shots have clipped highlights in the window areas (all channels clipped). The client asks: "Can you recover the view through the windows?" Provide an honest, technically precise answer, distinguishing between what is recoverable and what is not, and what alternatives exist.
The client also provides some older shots as JPEGs only (RAW files were deleted). These need the same processing. What are the specific limitations you will face compared to the RAW files, and how should you adjust your workflow to get the best possible result from the JPEGs?

This challenge requires you to synthesize knowledge from this chapter and Chapters 01-02 to make practical workflow decisions based on an understanding of the data, not just the tools.