From Photon to Pixel

8Image Representation: Coding and Formats

The wide range of formats used to archive and to display images constitutes an obstacle to their use by the general public. While the JPEG format has become widely used by both specialist and general users of photography, and is widely considered to represent a standard, a variety of other formats are available, and often produce better results.

The choice of format is determined by differing and often conflicting aims: reduction of image volume and/or maximum respect of properties (dynamics, chromatic palette, line geometry, quality of solid colors, etc.). Certain formats are best suited for archival purposes, while others are better for transmission, or for interoperability across a range of different treatment or visualization programs.

Images are used in a broad and varied range of applications. Some of these applications require documents to be transmitted at high speeds, whilse others are more concerned with quality, which may be judged according to different domain-specific criteria: representation of nuances or fine details, line sharpness, color reproduction, text readability, capacity for integration into a complex composition, etc.

Nowadays, even the most “ordinary” photograph is made up of at least ten million pixels, each often represented by 3 bytes, one per chromatic channel (red, green or blue). If the signal is not compressed in any way, each image requires several tens of millions of bytes of storage space in both the ...