High Dynamic Range Imaging: How HDR Photos, Video and Displays Work

Updated for 2026: This guide has been expanded to cover HDR photography, computational imaging, modern file formats, video standards, displays and perceptual tone mapping.

High dynamic range imaging, usually shortened to HDR or HDRI, is a family of capture, processing, storage and display techniques designed to preserve more detail across very bright and very dark parts of a scene. HDR can describe a merged photograph, a computational camera pipeline, a high-precision image file, an HDR video signal or a display capable of showing brighter highlights and deeper shadow detail.

That shared name causes much of the confusion. An HDR photo made from three bracketed exposures is not the same thing as HDR10 video, and neither is identical to an Ultra HDR image from a modern phone. They all address the same basic limitation: cameras, files and screens cannot reproduce the full range of light found in the real world without making choices about what to preserve.

Short answer: HDR captures or represents a wider range between the darkest useful shadow and the brightest useful highlight than standard dynamic range. The result can look more realistic when capture, processing and display are all handled correctly.

HDR at a glance

Term	What it means	Typical use
Dynamic range	The ratio between the darkest and brightest reproducible signal	Cameras, scenes, files and displays
HDR photography	Combining or processing exposures to preserve shadows and highlights	Landscapes, interiors, architecture and phones
HDR radiance image	Scene-referred data that can exceed the range of a normal display	VFX, lighting, research and archival workflows
Tone-mapped HDR	A wide-range image compressed for an SDR or limited display	Web images and conventional prints
HDR video	Video encoded with an HDR transfer function and wider color system	Streaming, discs and broadcasting
HDR display	A screen capable of reproducing greater luminance and contrast	TVs, monitors, phones and tablets
Ultra HDR	A compatible SDR image plus a gain map for brighter HDR rendering	Android photos and supported apps

What dynamic range actually measures

Dynamic range is the usable ratio between a system’s darkest and brightest values. Photographers usually express it in stops. One stop represents a doubling or halving of light, so a system with 14 stops can theoretically represent a luminance ratio of roughly 16,384:1.

Display engineers often use contrast ratios and luminance measured in candelas per square metre, commonly called nits. A screen’s peak brightness matters, but peak nits alone do not determine HDR quality. Black level, local contrast, color volume, tone mapping, panel uniformity and the size and duration of bright highlights all affect the result.

Dynamic range can refer to four different things:

• Scene dynamic range: the span of light present in the real environment.
• Camera dynamic range: the span a sensor can record before shadows become unusably noisy or highlights clip.
• File dynamic range: the tonal range an encoding can store and manipulate.
• Display dynamic range: the span a screen can reproduce under particular viewing conditions.

These values should not be treated as interchangeable. A high-bit-depth file does not automatically contain high dynamic range, and a display advertised with a large peak-brightness number may still produce weak HDR if its black level or tone mapping is poor.

HDR photography: capture more usable light

Traditional HDR photography combines several photographs of the same scene at different exposures. A dark frame protects bright highlights, a middle exposure records midtones, and a bright frame reveals shadow detail. Software aligns and merges the frames into a wider-range representation.

HDR is most useful when a single exposure cannot hold both ends of the scene. Common examples include an interior with bright windows, a sunset with a dark foreground, architecture beneath a bright sky and night scenes containing illuminated signs.

A practical exposure-bracketing workflow

1. Use RAW when possible. RAW preserves more sensor data and gives merge software more latitude than processed JPEG files.
2. Stabilize the camera. A tripod is ideal for architecture and landscapes, although modern alignment can handle small handheld movements.
3. Keep aperture, focus and white balance consistent. Change shutter speed between frames. Changing aperture alters depth of field, while automatic white balance can create color differences.
4. Bracket the exposure. Three frames at approximately −2 EV, 0 EV and +2 EV cover many scenes. Extremely contrasty scenes may need five or more frames with smaller spacing.
5. Avoid clipped highlights in the darkest frame. Check the histogram and highlight warning rather than relying only on the rear-screen preview.
6. Merge and deghost. Lightroom, Photoshop and specialist HDR applications can align exposures and reduce artifacts from moving people, foliage or water.
7. Tone map conservatively. Protect natural local contrast and color. Excessive microcontrast creates halos, gray shadows and the exaggerated “HDR look” that many viewers dislike.

When not to bracket exposures

Bracketing is a poor fit for fast action, unstable handheld framing or scenes with extensive movement. A modern RAW file may already contain enough recoverable range for a natural result, particularly at base ISO. In those cases, one properly exposed RAW capture avoids ghosting and simplifies the workflow.

Exposure fusion is another option. It combines well-exposed parts of several normal images without first creating a physically calibrated radiance map. Mertens and colleagues popularized a method based on contrast, saturation and well-exposedness weights. It often produces attractive results with less complexity, although it is not equivalent to a scene-referred HDR measurement.

How phones create computational HDR

Smartphones rarely depend on the old “three photos on a tripod” method. They capture a rapid burst before and after the shutter press, align the frames, reject motion, reduce noise and combine useful information into one result. Google’s HDR+ pipeline helped establish this approach; its research describes burst capture, alignment and merging for greater dynamic range and lower noise.

Modern phones may combine short exposures that protect highlights with longer frames that improve shadow signal. Multi-frame super-resolution, semantic segmentation and local tone mapping can also influence detail, faces, skies and color. This is why phone HDR is often active without a visible HDR button.

Computational HDR has limits. Moving subjects can create ghosting, hair and leaves may be reconstructed incorrectly, and strong local processing can make faces or skies look unnatural. A phone can also produce a bright, attractive image that is not a scene-referred HDR file. The capture method, stored file and final display should be considered separately.

Old HDR photos versus modern HDR display photos

For many years, photographers used “HDR image” to mean a wide-range capture that had been tone mapped into an ordinary SDR JPEG. The file itself could not make a compatible display produce brighter highlights; it only simulated wide range through local contrast and tonal compression.

Modern HDR photos can retain additional displayable luminance. Gain-map formats store a normal SDR base image plus auxiliary information that tells a compatible HDR display how much to brighten different regions. Unsupported software can show the SDR base, preserving backward compatibility.

Android’s Ultra HDR format uses this approach. Google’s documentation describes a primary image plus a gain map. Android 14 introduced platform support, and later APIs expanded editing and HEIC-based HDR options. Cropping, scaling or rotating must preserve the gain map correctly; filters and compositing may require the gain map to be updated as well.

Adobe has also expanded HDR capture, editing and output in Lightroom. A true HDR editing workflow requires an HDR-capable display, correct operating-system support and a compatible output format. Otherwise, the image must be tone mapped for SDR delivery.

HDR video and display standards

HDR television is built on defined transfer functions, color systems and metadata. The current ITU-R BT.2100 recommendation specifies parameters for HDR television and supports two transfer-function families: Perceptual Quantization (PQ) and Hybrid Log-Gamma (HLG).

• PQ, standardized as SMPTE ST 2084, maps code values to absolute display luminance. SMPTE describes a practical range up to 10,000 cd/m². Existing consumer displays do not need to reach that maximum; tone mapping adapts mastered content to the screen.
• HLG, developed for broadcasting, is a relative system designed to fit live and broadcast workflows and provide a degree of compatibility with conventional television transfer curves.

HDR10, HDR10+, Dolby Vision and HLG compared

Format	Transfer function	Metadata	Typical strength
HDR10	PQ	Static metadata for the program	Open baseline supported broadly by HDR TVs
HDR10+	PQ	Dynamic scene- or frame-level metadata	More adaptive tone mapping on compatible displays
Dolby Vision	PQ-based ecosystem	Dynamic metadata and managed authoring/playback	End-to-end adaptation across supported devices
HLG	HLG	Generally no display-target metadata required	Live broadcast and operational compatibility

HDR10, HDR10+ and Dolby Vision should not be described as separate transfer functions. They are delivery formats or ecosystems that use PQ. BT.2100 is a system recommendation, while ST 2084 defines PQ. Keeping those layers separate prevents one of the most common technical errors in HDR explanations.

HDR also often travels with wide color gamut, commonly using BT.2020 container primaries and 10-bit or higher signals. Wider color and higher bit depth reduce banding and expand color volume, but they are related features rather than the definition of dynamic range itself.

What makes an HDR display good?

A convincing HDR display needs more than a logo. Evaluate these factors together:

• Peak and sustained brightness: Small highlights may be much brighter than a full white screen. Measurements should state window size and duration.
• Black level: OLED controls light per pixel and can produce extremely dark blacks. LCD displays depend on backlight control.
• Local dimming: Mini-LED LCDs use many zones to brighten highlights while darkening nearby regions. Too few zones can create blooming around bright objects.
• Color volume: A display should maintain saturated colors as luminance rises rather than washing them toward white.
• Tone mapping: The display must adapt content mastered beyond its capability without clipping highlights or making the whole image too dark.
• Accuracy: Correct tracking of the intended transfer function matters more than an exaggerated showroom mode.
• Viewing environment: Ambient light changes perceived contrast. A bright living room and a dark grading suite require different priorities.

OLED, mini-LED LCD, dual-layer LCD and MicroLED can all support HDR, but each has trade-offs. OLED offers per-pixel black control. Mini-LED LCD can reach very high brightness but may bloom. Dual-layer professional displays can combine deep blacks with high luminance at high cost. MicroLED is emissive and promising, but remains expensive and limited in many markets.

HDR file formats and when to use them

Format	Representation	Best suited to
Radiance RGBE (.hdr)	RGB mantissas with a shared exponent	Image-based lighting and legacy HDR tools
OpenEXR (.exr)	16- or 32-bit floating point, multi-channel	VFX, compositing, rendering and archival masters
TIFF	Integer or floating-point variants	Photography and interchange, depending on software
HEIF/AVIF HDR	High-bit-depth PQ, HLG or gain-map workflows	Efficient modern delivery
JPEG XL	High precision and HDR-capable encodings	Advanced imaging workflows where supported
Ultra HDR JPEG	SDR JPEG plus gain map	Backward-compatible HDR phone photography

OpenEXR is designed to store scene-linear high-dynamic-range image data rather than a finished display image. Its official technical introduction explains support for 16- and 32-bit floating-point pixels, large dynamic range, fine color resolution and production-oriented channels. That makes it ideal for VFX and rendering, but excessive for most web photographs.

An ordinary 8-bit JPEG can show a tone-mapped interpretation of an HDR scene, but it cannot preserve the same editing latitude or signal range as a floating-point master. A compatible gain-map JPEG is different because it can carry additional luminance reconstruction information alongside its SDR base.

Tone mapping: fitting a large range onto a limited display

Tone mapping transforms HDR values for a display or output medium with a smaller usable range. A global operator applies a broadly consistent curve across the image. A local operator changes the mapping according to nearby luminance and can reveal more local detail, but it also creates a greater risk of halos or unnatural contrast.

Important approaches include:

• Reinhard photographic tone reproduction: a compact global method inspired by photographic practice.
• Durand and Dorsey bilateral decomposition: separates base illumination from detail to compress the large-scale range.
• Fattal gradient-domain compression: reduces strong gradients while preserving fine structure.
• Drago logarithmic mapping: adapts logarithmic compression to very bright scenes.
• Mantiuk contrast-domain methods: use perceptually motivated contrast processing.
• Mertens exposure fusion: blends exposures directly rather than constructing a calibrated radiance map.
• Filmic and ACES pipelines: shape highlight roll-off and map scene-referred content for cinema, games and production displays.

Major tone-mapping operators compared

The following comparison preserves the original article’s research-based overview of influential global, local, perceptual and cinematic tone-mapping approaches. The named studies are included in the selected references below.

Table: Major HDR tone-mapping operators, working spaces, core ideas and practical trade-offs. Sources include Reinhard et al. (2002), Durand and Dorsey (2002), Fattal et al. (2002), Mantiuk et al. (2006), Mertens et al. (2007), and Mehmood and colleagues (2020–2024).

The Debevec–Malik HDR merge

The modern photographic HDR pipeline was formalized by Paul Debevec and Jitendra Malik in 1997. Their method estimates a camera response function from differently exposed photographs, then combines the reliable parts of each exposure into a radiance map. In simplified form:

Eᵢ = [Σⱼ w(Zᵢⱼ) · g⁻¹(Zᵢⱼ) / tⱼ] / [Σⱼ w(Zᵢⱼ)]

Here, Zᵢⱼ is the recorded pixel value for location i in exposure j, tⱼ is exposure time, g⁻¹ converts the recorded value toward scene exposure, and w reduces the influence of values near noise or clipping. Modern software adds alignment, deghosting, denoising and camera-specific RAW processing.

Perceptual and color-appearance tone mapping

Tone mapping is not only a signal-compression problem. The human visual system adapts to local brightness, surround and viewing conditions, while simple RGB or luminance curves do not fully preserve perceived color and contrast.

Recent work by Imran Mehmood, M. Ronnier Luo and collaborators explores tone mapping in perceptually uniform and color-appearance spaces. The group’s Perceptual Tone Mapping Model incorporates contrast sensitivity and local adaptation. Related CIECAM16-based work performs compression in a color-appearance lightness dimension, while generic and adaptive chroma-correction methods aim to restore colorfulness lost during tone compression.

This research belongs in the larger history of perceptual tone reproduction rather than replacing the classical foundation. Reinhard, Durand, Fattal, Drago, Mantiuk, Mertens, ACES and newer perceptual models solve different parts of the pipeline and make different trade-offs among speed, local detail, appearance stability and artifacts.

Common HDR problems and how to fix them

Halos around buildings or trees

Reduce local-contrast strength, clarity or aggressive dehazing. Check masks near high-contrast edges and prefer a gentler global curve when the scene does not require strong local compression.

Ghosting from people, leaves or water

Use the software’s deghosting control and select a clean reference frame. For fast movement, use a single RAW exposure or a computational burst designed for motion rather than a slow bracket.

Gray or flat-looking shadows

Do not lift every shadow equally. Preserve a black point and allow some areas to remain dark. HDR means greater usable range, not equal brightness everywhere.

Neon colors and unnatural skin

Reduce saturation after tone mapping, inspect local adjustments and work in a color-managed workflow. Luminance compression can change perceived chroma, which is why perceptual and chroma-correction research remains important.

HDR video looks washed out

This usually indicates a transfer-function or color-management mismatch. Confirm that the operating system, player, cable, GPU output and display input all support the intended HDR mode. An HLG, PQ or SDR signal interpreted as the wrong type will not display correctly.

HDR looks too dark

PQ content may appear dark on a display with limited luminance or poor tone mapping, especially in bright ambient light. Verify the correct picture mode, disable aggressive energy-saving settings and check whether the content and display share a supported HDR format.

How to judge whether HDR is working

Good HDR does not make the entire picture brighter. It creates room for highlights while protecting shadow detail and midtone contrast. Look for natural specular reflections, texture in clouds, controlled bright signs, dimensional lighting and smooth color gradients. Faces should remain believable, and dark scenes should stay dark without losing intentional detail.

For photography, compare the result with the middle RAW exposure at a normal viewing size. If the HDR version only adds halos and saturation, the merge is not helping. For displays, use known HDR material and confirm that the device actually enters its HDR mode. Marketing labels alone are insufficient.

How HDR evolved from photography to modern displays

High dynamic range imaging did not begin as a television feature. Its foundations are tied to the much older problem of representing scenes whose brightness range exceeds the capacity of one recording medium. Photographers solved versions of this problem through exposure control, dodging and burning, graduated filters and careful printing. Digital imaging turned those craft techniques into measurable capture and processing pipelines.

A major milestone arrived in 1997, when Paul Debevec and Jitendra Malik described a practical method for recovering a camera response curve and combining differently exposed photographs into a radiance map. Their work made it possible to estimate scene radiance from ordinary photographs rather than merely blending attractive portions of several images. That distinction matters: a radiance map can preserve values far above the white point of a conventional image, making it useful for research, visual effects and image-based lighting as well as photography.

Early consumer HDR became associated with heavily tone-mapped JPEG images. Local operators revealed texture everywhere, but aggressive settings often produced bright halos, gray blacks and surreal saturation. This recognizable style caused many people to treat HDR as an effect. Modern HDR is broader. Phone cameras merge bursts automatically, professional pipelines retain scene-referred floating-point data, and HDR displays can reproduce highlights that would previously have required compression into SDR.

The history therefore has three overlapping phases: extending capture through multiple exposures, compressing wide-range data for ordinary output, and delivering a genuinely wider-range signal to capable displays. Current workflows often use all three.

Choosing the right HDR capture method

There is no single best way to capture HDR. The correct method depends on subject movement, camera support, required precision and final output. Before bracketing, determine whether a single RAW exposure already protects the highlights while keeping shadow noise acceptable. Modern sensors can recover substantial shadow detail at base ISO, and unnecessary brackets increase storage and introduce alignment problems.

Single RAW exposure

Use one RAW frame when the scene fits within the sensor’s usable range or when motion makes a sequence unreliable. Expose to protect important highlights without pushing the remaining image so dark that shadow noise and color errors become objectionable. A single frame is generally the cleanest option for people, wildlife, waves and handheld action.

Automatic exposure bracketing

Use automatic exposure bracketing for static scenes that exceed one exposure. Interiors with bright windows and architectural scenes at sunrise are classic examples. Three frames two stops apart are a practical starting point, but the histogram should determine the range. The darkest exposure must retain highlight color and texture; the brightest must lift important shadows above the noise floor.

Computational burst capture

Phones and some cameras capture a rapid burst before and after the shutter press, align selected frames and combine them using spatially varying weights. Short exposures protect highlights and reduce motion blur, while longer or accumulated exposures improve darker areas. Semantic processing may treat skies, faces and foliage differently. This can produce a natural result from one shutter action, but it is not literally one exposure.

Dual-gain and dual-conversion capture

Some sensors read information using different gain paths. One path protects bright areas while another improves darker signal quality. When combined, the result can extend usable dynamic range without the time gap of conventional bracketing. Implementation details vary by sensor, and additional range may involve resolution, readout or noise trade-offs.

Graduated filters and controlled lighting

HDR processing is not always the best solution. A graduated neutral-density filter can reduce a bright sky in a landscape, while fill light can reduce contrast in a portrait or interior. These methods change the captured scene rather than reconstructing it later and can prevent difficult transitions around moving subjects.

An end-to-end HDR photography workflow

A reliable HDR workflow preserves flexibility until the final output. Begin by shooting RAW with a fixed white balance, fixed focus and the lowest practical ISO. If using a tripod, disable stabilization modes that can introduce movement when the camera is locked. Use a remote release or short timer when long exposures could be affected by vibration.

Import the frames without applying radically different adjustments to individual exposures. Lens-profile corrections and chromatic-aberration removal are usually safe before merging, but strong local edits should wait. Ask the merge software to align the sequence and inspect the result at 100 percent. Alignment failures appear as doubled edges, while deghosting errors often look like translucent people, broken leaves or repeating water patterns.

Perform global tonal work before local enhancement. Establish a convincing black point, protect specular highlights and keep midtone contrast strong enough that the image does not look flat. Then use selective masks for areas that genuinely need attention. Local contrast should reinforce the visual hierarchy rather than reveal maximum texture in every surface.

Color requires separate judgment. Lifting deep shadows can expose weak chroma data, and highlight compression can push highly saturated colors outside the destination gamut. Check skin, skies, foliage and artificial lights for hue shifts. Soft proofing or an output preview helps identify colors that will clip on an SDR web image or print.

Keep a high-bit-depth master—typically a 16-bit TIFF, PSD or floating-point EXR depending on the workflow—and create delivery versions from that master. Do not repeatedly edit and resave an 8-bit JPEG. A separate SDR rendition remains important because many websites, messaging applications and display paths still do not preserve HDR metadata or gain maps consistently.

HDR video: capture, grading and delivery

HDR video is a complete production chain. A camera records a wide-range signal, the footage is transformed into a defined working space, a colorist grades it on a reference display, metadata and mastering decisions are encoded, and the playback device maps the signal to the viewer’s screen. A failure at any point can produce clipped highlights, raised blacks, washed-out color or an image that is simply too dark.

ITU-R BT.2100 defines key image parameters for HDR television and includes two principal transfer-function approaches. Perceptual Quantizer, or PQ, represents display-referred absolute luminance and is used by HDR10 and Dolby Vision workflows. Hybrid Log-Gamma, or HLG, is scene-referred and was designed with live production and broadcasting in mind. They are not interchangeable labels; they embody different assumptions about mastering and display adaptation.

HDR10 uses static mastering information, meaning one set of metadata describes the program as a whole. HDR10+ and Dolby Vision can provide scene-level or frame-level guidance that helps compatible displays adapt content more precisely. Dynamic metadata does not guarantee a better picture: the quality of the grade, encoding and television’s tone-mapping implementation remains decisive.

Consumer displays rarely match a mastering monitor’s exact capabilities. A television therefore performs display mapping based on its peak brightness, black level and color volume. If a 1,000-nit master is shown on a device capable of a much smaller highlight level, the device must compress part of the signal. Good mapping preserves highlight relationships and overall contrast; poor mapping either clips bright detail or darkens the entire presentation.

Bit depth, color gamut and luminance

HDR is often discussed alongside wide color gamut and higher bit depth because the three work together, but they describe different properties. Dynamic range concerns luminance span. Color gamut describes the range of chromaticities a system can represent. Bit depth determines how many discrete code values are available for encoding tonal and color transitions.

Increasing luminance range without sufficient precision can make banding more visible, particularly in skies, gradients and shadow transitions. Ten-bit delivery provides substantially more code values per channel than eight-bit delivery and is standard in mainstream HDR video pipelines. However, a 10-bit file is not automatically HDR, just as an image tagged with a wide color space does not necessarily contain wide-gamut colors.

BT.2100 commonly operates with the BT.2020 color primaries as a container, while actual mastered colors may occupy a smaller practical gamut such as P3. Displays differ in how much of that volume they reproduce at different luminance levels. This is why two televisions with similar peak-brightness specifications may render saturated bright colors very differently.

HDR quality assessment

Assessing HDR requires more than checking whether detail is visible. A successful result should retain the scene’s intended hierarchy: important highlights should attract attention, blacks should remain convincing where appropriate, faces should look natural and local contrast should not create distracting artifacts.

Objective measurements can identify clipping, noise, color error and differences from a reference, but a single score cannot fully describe appearance. Metrics such as HDR-VDP-2 model aspects of human visibility across a wide luminance range. Tone-mapped image metrics attempt to predict structural fidelity or naturalness after compression. Their usefulness depends on the viewing assumptions and the availability of a meaningful reference.

Subjective evaluation remains essential. View the image at an appropriate distance under controlled ambient light, compare it with a neutral reference rendering, and allow observers time to adapt. A bright showroom mode in a lit room and a calibrated display in a dark grading environment create different perceptions. Evaluation conditions should therefore be recorded when results are compared.

For practical reviews, inspect several content types: a dark scene with small highlights, a bright outdoor scene, saturated objects, skin tones and smooth gradients. Look for black crush, blooming, posterization, color clipping and abrupt tone-mapping changes. On a phone, also check whether the image changes when opened in another application or shared through a service that may remove HDR data.

Publishing HDR images on the web

Web delivery is complicated because capture, browser, operating system and display support do not advance together. A conventional JPEG or WebP rendition offers broad compatibility but shows only the tone-mapped SDR interpretation. Newer workflows can pair an SDR base with additional gain-map data so compatible software reconstructs a brighter HDR presentation while older software displays the base image.

When publishing, test the page on both HDR and SDR screens. The SDR fallback should look intentional rather than dull, and essential information must never depend on extreme highlight brightness. Use descriptive alt text because dynamic range cannot compensate for inaccessible content. Keep file dimensions and compression sensible: an unnecessarily large HDR asset can damage page speed, which is especially counterproductive for a search-focused article.

Metadata preservation is critical. Image editors, social platforms and optimization plugins may strip profiles, gain maps or auxiliary data. Verify the final downloaded asset rather than assuming the uploaded master survived the publishing pipeline. When consistent HDR delivery is not possible, a carefully tone-mapped SDR image is preferable to an unpredictable result.

HDR for printing

Print does not emit light, so it cannot reproduce display HDR in the same physical way. Paper white depends on illumination, and the darkest printable black reflects some light. The available contrast is therefore much smaller than the range stored in an HDR master. Printing requires tone and gamut mapping tailored to the paper, ink, printer and viewing conditions.

A high-quality HDR capture still benefits print because it gives the editor more information from which to build the final rendering. Highlight separation, shadow color and local contrast can be allocated deliberately instead of being lost during capture. Use a calibrated monitor, the correct printer profile and soft proofing. Matte and glossy papers need different expectations, particularly for black depth and highlight brilliance.

Common HDR myths

“More exposures always produce better HDR”

Extra frames help only when they record useful information beyond the sensor’s existing range. Redundant exposures increase processing time and the chance of movement artifacts. The required endpoints matter more than the frame count.

“The brightest display has the best HDR”

Peak luminance is only one component. A display with high brightness but weak blacks, coarse local dimming or inaccurate tone mapping may look less convincing than a lower-peak display with stronger contrast control and color volume.

“HDR means oversaturated and unrealistic”

That appearance comes from particular processing decisions. A well-managed HDR workflow can look more natural because it avoids clipping and preserves highlight relationships that SDR must compress.

“RAW and HDR are the same”

RAW is a capture format containing minimally processed sensor data. It may provide wide editing latitude, but it does not automatically represent the full range of a difficult scene or define an HDR display signal.

Frequently asked questions

What is high dynamic range imaging?

High dynamic range imaging is a group of methods that capture, store, process or display a wider range between dark and bright image values than a conventional standard-dynamic-range workflow.

Is HDR the same as 4K?

No. 4K describes spatial resolution, while HDR describes luminance and tonal range. A display can be 4K without producing good HDR, and HDR can be delivered at resolutions below 4K under BT.2100.

Does HDR require multiple photos?

Not always. Exposure bracketing is one method. Modern sensors, RAW files, dual-gain capture and computational bursts can create useful HDR information from a single shutter action.

What is the difference between HDR and HDR10?

HDR is the broad concept. HDR10 is a widely supported video-delivery format that uses PQ, 10-bit signals and static metadata.

Is Dolby Vision better than HDR10?

Dolby Vision can use dynamic metadata and a managed playback ecosystem, giving compatible displays more information for tone mapping. The visible result still depends on the master, player and display; a well-mastered HDR10 presentation can look better than a poor Dolby Vision one.

What is tone mapping?

Tone mapping converts a wide-range image or video signal to the smaller range a particular screen or print can reproduce. It decides how highlights, shadows, contrast and color are compressed.

What is an HDR gain map?

A gain map stores instructions for increasing luminance over an SDR base image. Compatible displays reconstruct an HDR appearance, while unsupported software can show the conventional base image.

Why do some HDR photos look fake?

Aggressive local contrast, saturation and shadow lifting can create halos, flat blacks and unnatural color. These are processing choices, not unavoidable properties of HDR.

Can every screen show HDR?

No. A compatible operating system, application, signal path and HDR-capable display are required for genuine HDR output. Otherwise, the content is converted or shown as SDR.

Conclusion

High dynamic range imaging is best understood as a pipeline, not a single effect. The scene contains a range of light; a camera captures part of it; processing and file formats preserve or compress it; and the display maps the result to its own capabilities. HDR succeeds when every stage protects meaningful highlight, shadow, color and contrast information without making the image look artificial.

For photographers, the practical lesson is simple: bracket only when one RAW exposure is not enough, merge carefully and tone map with restraint. For display buyers, evaluate black level, sustained brightness, local dimming, color volume and tone mapping rather than relying on an HDR logo. For imaging professionals, scene-referred floating-point data, defined transfer functions and perceptually grounded appearance models provide the precision needed to move HDR reliably between capture and display.

Selected references

1. ITU-R. “Recommendation BT.2100-3: Image parameter values for high dynamic range television for use in production and international programme exchange.” 2025. https://www.itu.int/rec/r-rec-bt.2100
2. Society of Motion Picture and Television Engineers. “2021 High-Dynamic Range Progress Report.” SMPTE Motion Imaging Journal 130(8). https://journal.smpte.org/periodicals/SMPTE%20Motion%20Imaging%20Journal/130/8/32/
3. Google Research. “Introducing the HDR+ Burst Photography Dataset.” 2018. https://research.google/blog/introducing-the-hdr-burst-photography-dataset/
4. Android Developers. “Edit Ultra HDR images.” Updated 2026. https://developer.android.com/media/grow/ultra-hdr/edit
5. OpenEXR. “Technical Introduction to OpenEXR.” https://openexr.com/en/latest/TechnicalIntroduction.html
6. Adobe. “High Dynamic Range — Explained.” 2023. https://blog.adobe.com/en/publish/2023/10/10/hdr-explained
7. Debevec, P. E., and J. Malik. “Recovering High Dynamic Range Radiance Maps from Photographs.” SIGGRAPH, 1997. https://www.pauldebevec.com/Research/HDR/debevec-siggraph97.pdf
8. Reinhard, E., M. Stark, P. Shirley, and J. Ferwerda. “Photographic Tone Reproduction for Digital Images.” ACM Transactions on Graphics, 2002.
9. Durand, F., and J. Dorsey. “Fast Bilateral Filtering for the Display of High-Dynamic-Range Images.” ACM Transactions on Graphics, 2002.
10. Fattal, R., D. Lischinski, and M. Werman. “Gradient Domain High Dynamic Range Compression.” ACM Transactions on Graphics, 2002.
11. Mertens, T., J. Kautz, and F. Van Reeth. “Exposure Fusion.” Pacific Graphics, 2007.
12. Mehmood, I., X. Shi, M. U. Khan, and M. R. Luo. “Perceptual Tone Mapping Model for High Dynamic Range Imaging.” IEEE Access 11, 2023, 110272–110288.
13. Mehmood, I., M. U. Khan, and M. R. Luo. “Generic Color Correction for Tone Mapping Operators in High Dynamic Range Imaging.” Optics Express 32(16), 2024, 27849–27866.
14. Mehmood, I., X. Liu, M. U. Khan, and M. R. Luo. “Method for Developing and Using High Quality Reference Images to Evaluate Tone Mapping Operators.” JOSA A 39(6), 2022, B11–B20.
15. Mehmood, I., M. Zhou, M. U. Khan, and M. R. Luo. “CIECAM16-Based Tone Mapping of High Dynamic Range Images.” Color and Imaging Conference 31, 2023, 102–107.
16. Mehmood, I., M. U. Khan, and M. R. Luo. “Adaptive Chroma Correction of Tone Mapping Operators for Natural Image Appearance.” CIC32, 2024.
17. Mehmood, I., M. U. Khan, M. F. Mughal, and M. R. Luo. “A Tone Mapping Model Based on Receptive Field for HDR Images.” CIC28, 2020, 100–104.
18. Mehmood, I., and M. R. Luo. “Developing HDR Tone Mapping Operators Based on Uniform Colour Spaces.” 2021.
19. Khan, M. U., I. Mehmood, M. R. Luo, and M. F. Mughal. “No-Reference Image Quality Metric for Tone-Mapped Images.” CIC27, 2019, 252–255.
20. Mehmood, I., M. U. Khan, M. R. Luo, and M. F. Mughal. “Tone Mapping Operators Evaluation Based on High Quality Reference Images.” CIC27, 2019, 268–272.
21. Mantiuk, R., K. J. Kim, A. G. Rempel, and W. Heidrich. “HDR-VDP-2: A Calibrated Visual Metric for Visibility and Quality Predictions in All Luminance Conditions.” ACM Transactions on Graphics, 2011.