More range. Not simply brighter.

HDR is not defined by an export switch. A camera measures a scene, processing renders an image, a transfer function encodes the signal and the viewing chain presents it on a screen. A mismatch at any boundary matters even when the file extension and HDR flag look right.

SDR also needs viewing assumptions. A 100 cd/m² reference-white example is useful, but is not a physical ceiling for every SDR display. Compare coordinated chains with known white, black and surroundings rather than two application windows at maximum brightness.

BT.2100 specifies HDR television parameters and two signal systems: PQ and HLG. It covers more than curves, including colourimetry, component representation, depth and digital levels. BT.2020, PQ and HLG answer different questions.

Rec.2020 alone does not imply HDR; ten bits define neither gamut nor luminance. Check primaries, transfer function and code range together. RGB/Y′CbCr conversion adds a matrix choice, which is not a substitute for a transfer-function setting.

Operational HDR guidance uses 203 cd/m² as nominal HDR Reference White. It helps align graphics, diffuse white and SDR/HDR interchange. It is not a compulsory luminance for every white object, face or shot.

A white object in shadow may be much darker, while an emitter may be brighter. The lab separates white from highlights to demonstrate range allocation, not to replace an agreed production reference with an arbitrary slider position.

A broad white surface and a small sun reflection need not share a luminance. HDR can distinguish them without raising the whole image. A higher peak alone says nothing about bright-area size or duration.

Increasing highlight area in the test chart changes its mean while keeping its maximum fixed. Real viewing comfort and display capability are more complex: field size, duration and surroundings also matter.

PQ defines an absolute reference relationship between normalised signal and display light. HLG describes relative scene encoding plus a display system. Calling HLG “PQ without metadata” misses that distinction.

Conversion requires the source rendering, reference levels and target viewing assumptions. Changing a transfer tag does not transform sample values. The following chapters let you inspect both systems numerically rather than judging HDR from SDR swatches.

Colour volume combines chromaticity and luminance. A flat primary triangle does not show which saturated colours a display can reproduce at different light levels. A higher peak therefore does not guarantee saturated highlights.

Tone mapping reshapes tonal relationships; gamut mapping fits colours to target limits. These tasks interact in an output transform. Correct RGB matrix conversion alone cannot decide how a neon emitter should look on a smaller display.

Mastering conditions describe the reference display; image levels describe the programme. A 1000 cd/m² mastering display does not require every shot to reach 1000, let alone the whole frame. Device description and content measurement are not interchangeable.

The lab separates peak, mean and area above a chosen ceiling. The latter illustrates hard-clipping risk. A real display may map the signal differently, while video analysis must inspect the sequence and defined measurement rules.

HDR judgement needs both scopes and appropriate viewing conditions. The application, operating system, output interface and screen must interpret the signal coherently. An ordinary screenshot or sRGB window does not document actual HDR luminance.

“1000” in this lab is a computed value, not a command to your monitor. Final review requires compatible output, the intended display mode and known surroundings. Without HDR monitoring, structural and numerical checks are possible, but visual approval remains outstanding.

A white title is defined in an encoding such as sRGB. Decode it appropriately and place it at the agreed graphics-white level. Interpreting code 1 as the maximum of every HDR output is not a general import rule.

Keep separate control of text, logos and footage where appropriate. A restrained title need not receive the same creative highlight treatment as a bright window. Compositing order and conversion boundaries, including transparency, must still be explicit.

Colour tags tell a receiver how to interpret a stream. Additional mastering or content information can assist output adaptation. These are different layers: metadata presence does not prove pixel correctness.

Do not fill fields with guessed “HDR values” without delivery requirements and analysis. HLG does not require mastering metadata for its reference display system, but still needs correct transfer, primary and range signalling.

Reducing range requires choices about midtones, highlight detail, saturation, local contrast and intent. Global scaling can fit the maximum while darkening faces and graphics; clipping can keep the middle but discard highlight differences.

Review the SDR result as a separate deliverable. BT.2446 describes multiple conversion approaches rather than one universal recipe. Expanding already clipped SDR cannot restore lost texture.

In Resolve, identify camera log, RAW, finished PQ/HLG and SDR graphics before choosing a working representation. Decide whether project colour management or an explicit conversion graph owns the output rendering.

Do not add a CST merely because another graph has one. Duplicating a project-owned output changes the image twice. Multiple transforms are legitimate when each has a distinct purpose and matched input/output definitions.

After rendering, inspect codec, depth, range, primaries and transfer rather than only an HDR checkbox. Reimport a short test to expose duplicate transforms or range errors. Metadata records a declaration, not proof that it is true.

Review shadows, saturated highlights, graphics and shot transitions on the target chain. Record project settings and delivery requirements. A checklist is useful only when each item is backed by an observation rather than assumed defaults.