Why Tone Mapping Is Needed
Tone mapping adapts a picture's brightness range to what the output display can reproduce.
An HDR source asks for a 1,000-nit reflection beside a 100-nit surface. A destination with a 100-nit output peak has room for the surface, but cannot reproduce the reflection at its requested absolute light level. Tone mapping chooses how both parts fit into the available picture.
There are two simple solutions with visibly different costs:
| Choice in this example | Surface | Reflection | Result |
|---|---|---|---|
| Clip at 100 nits | 100 | 100 | Brightness separation is lost |
| Divide all light by ten | 10 | 100 | The whole lower range is darkened |
A nonlinear mapping can give ordinary levels more room while compressing highlights more strongly. It redistributes contrast according to the source and destination instead of making only an all-or-nothing cutoff decision.
Why this is an image operation
Changing resolution alters the spatial sample grid. Changing bitrate alters how the encoder describes the picture. Neither operation, on its own, changes the light interpretation enough to solve this range mismatch.
A transfer-function label describes the meaning of the samples. Relabeling HDR as SDR would ask the player to apply another meaning to the existing values. Correct conversion first interprets the source, adapts its brightness and colors, then prepares suitable output values and signaling.
The HDR-to-SDR walkthrough follows those steps. Curves show the brightness relationship; gamut and color-volume handling also matter for a bright colored object.
Place the mapping in the playback path
Emby Server can process pictures during a video transcode. A renderer can also adapt them locally; the Windows app documentation describes presentation-stage tone mapping. Direct Play identifies delivery from the server, while later app and display processing still determines the light produced.
Mapping can also be needed between two HDR ranges, not only from HDR to SDR. A source peak above an HDR display's usable output presents the same fitting problem, even when the destination remains HDR.
The useful comparison is the relationship among objects in the same scene: ordinary surfaces, dark detail, and bright texture. A good peak fit is one part of that relationship. The mapping's available range and its chosen distribution explain the rest.