Unit
- Dynamic range definition: The ratio between the brightest and darkest parts of a scene that a sensor or film can capture in a single exposure. Measured in stops, typical cameras achieve 10-14 stops, while the human eye perceives about 20-24 stops simultaneously.
- Scene versus capture range: A sunlit landscape may have over 20 stops of brightness difference. Without HDR, either highlights blow out to pure white or shadows crush to pure black, losing texture and detail.
- Histogram as a tool: A camera histogram shows the distribution of tones from black (left) to white (right). A high-contrast scene produces peaks at both ends, signaling lost detail. HDR aims to bring those peaks into the visible range.
- Stop increments: Each stop represents a doubling or halving of light. A three-exposure bracket at -2, 0, +2 EV captures a total range of approximately 6 stops wider than a single shot, depending on the camera.
- Sensor limitations: Digital sensors have a linear response to light, unlike film. This linearity means that most exposure data is concentrated in the highlights, leaving shadows with fewer bits and more noise.
- Importance of raw capture: Raw files contain 12-14 bits of luminance data per channel, versus 8 bits in JPEG. This extra information is essential for HDR processing to reconstruct smooth tonal gradations.
- Dual-ISO and sensor innovations: Some cameras use dual-ISO sensors or staggered readout to extend dynamic range in a single exposure, but traditional multi-exposure HDR still offers superior control for extreme contrasts.
- Perceptual versus measured dynamic range: The human eye adapts to different brightness levels across the scene, while a camera sensor captures a fixed range per exposure. HDR mimics this adaptation by merging multiple exposures.
- Exposure bracketing technique: A typical bracket of three or five exposures at one or two stop intervals ensures coverage of the full scene range. Set the camera to continuous high speed and use a remote shutter to minimize movement between frames, which reduces alignment errors in post processing.
- Alignment and ghost removal: Even with a tripod, slight subject motion or wind can cause ghosting in merged HDR images. Most HDR software offers automatic alignment and deghosting tools that detect and blend only the sharpest pixels from each exposure, preserving detail where movement occurred.
Design Principle
Principle: Tonal Compression and Detail Preservation
Definition: Tonal compression reduces the overall brightness range of a scene so that both highlights and shadows fit within the output medium limited dynamic range, typically 8 stops for a standard monitor. The goal is to preserve as much visible detail as possible across the entire tone scale.
Application: When processing an HDR image, use a tone mapping operator that applies local or global adjustments. For example, the photographic tone mapper simulates the human visual system by compressing highlights more aggressively than shadows, maintaining a natural look. Adjust the radius parameter to control halo artifacts.
Quick Revision
- Dynamic range is the brightness ratio between darkest and brightest scene areas, typically 20+ stops in nature but only 10-14 stops captured by cameras.
- HDR uses multiple exposures to extend the captured range, with histogram analysis guiding bracket selection.
- Raw files are essential for HDR because they preserve the bit depth needed for smooth tone mapping.
- Tonal compression reduces the scene range to fit the output medium while maintaining visible detail.