Difference between revisions of "HDR"
Revision as of 21:39, 1 May 2006
HDR means High Dynamic Range. The term is used to refer to techniques and images that are capable of capturing and reproducing scenes with a high dynamic range.
HDR is used more or less accurately for almost any way to get more details out of the shadows and/or highlights. To understand how so-called "HDR techniques" work, it is important to differentiate between the two limitations that affects dynamic range in photography and imaging. The first limitation comes from the capture, i.e. the camera. The second limitation from the display, i.e. monitor or printer.
- Camera limitation: Standard digital cameras can capture at most a dynamic range of 1,000:1, which is much less than the dynamic range of most outdoor scenes. This limitation is usually addressed by taking several exposures of the same scene (this is the purpose of the "Automatic Exposure Bracketing" function available in many digicams).
- Display limitation: Standard monitors have a rather low dynamic range, around 100:1, which is not even enough to display correctly the RAW data captured by a standard camera. This limitation is addressed by tone mapping or by techniques that directly blend exposures. Exposure Blending can be done either manually in Photoshop, semi-manually with Photoshop actions (see Contrast Blending and Jook Leung's HDR for Dummy) or automatically in specialized software (see Photomatix or FDRtools - comparison on HDR compression).
Both camera and display limitations can be overcome with specialised equipment like the SpheroCam HDR camera and Brightside HDR Display. However, such equipment remains very expensive for the time being.
HDR images are stored in a high bit depth and/or floating point format. There are several HDR formats among them
- .hdr Radiance RGBE format
- .tif Floating point TIFF
- .pfm Portable floatmap
- .float Raw binary floating point
- .exr OpenEXR EXR format