Background: HDR Imaging (High Dynamic Range)
- 16,777,216 colors might not be enough!
What does "Dynamic Range" mean?
The dynamic range of an image is the ratio of the darkest pixel value to the brightest pixel value.
The technical quality of an image is also determined by the number of values
that can be defined in between these two extremities.
Photographers usually don't define the dynamic range as a ratio,
but prefer the use of exposure values (EV).
An exposure value defines the amount of light falling on the image sensor,
depending on the camera's aperture and exposure time settings.
This is a logarithmic value. Doubling the amout of light will increase the EV by one.
A dynamic value of 1:400 corresponds to about 12 EV.
Some rough estimates of dynamic ranges:
| Paper print, matte | 5 - 6 EV |
| Paper print, glossy | 7 - 8 EV |
| Colour negative film | 8 - 10 EV |
| Slide film | 6 - 9 EV |
| Digital image sensor | 7 - 11 EV |
| LCD screen | 10 - 18 EV |
| Daylight, sunshine | 15 EV |
| Human eye, immidiately visible | 10 EV |
| Human eye, after adaptation | 20 EV |
| HDR, definition | > 10 EV |
Image Storage Formats
Brightness values of images are usually stored as integer values.
Examples:
| Black/White | 1 bit per pixel |
| Greyscale | 8 bits per pixel |
| Palettes | 1 ... 8 bits per pixel, plus color table |
| RGB 5/6 | 5 oder 6 bits per color channel, 16 bits per pixel |
| RGB 8 | 8 bits per color channel, thus 24 bits per pixel |
| RGBA 8 | 8 bits per color channel + alpha, thus 32 bits per pixel |
| RGB 16 | 16 bits per color channel, thus 48 bits per pixel |
Today, the most commonly used format for photographs is RGB 8.
Values are to be interpreted as logarithmic. They can be transformed to physical values
by applying a so-called gamma transformation.
Unlike the formats above, HDR uses floating point numbers for storage.
This allows a large increase in dynamic range plus a finer gradation of values.
If applied in a smart way, this won't require much more space.
HDR image formats are generic and don't depend on camera manufacturers - unlike raw file formats.
In the long run, HDR could replace 16-bit integer formats in professional imaging!
Creating HDR Images
To a certain extent, HDR images can be created from camera raw files.
For high-end cameras, there's a trend to develop image sensors with higher dynamic ranges.
For real high dynamic range, images can be created by blending several single images
with different exposure settings using special software (exposure blending).
What is HDR good for?
Even in future times, HDR images will not always be displayed directly,
like on HDR enabled monitor screens or HDR enabled viewers (see below).
But if you regard such images as raw material, there are some very interesting
fields of application:
- Creation of so-called environment maps - spherical or cubic panorama images
which are used as background images for 3D rendering, like in film productions,
advertising graphics, or computer games.
- Creation of artistic images containing areas of very different brightness.
This requires the use of tonemapping tools.
