A pronunciation dictionary is basically a mapping between the orthographic form and the phonetic form of word entities. Fundamental descriptors for pronunciation dictionaries are the coding of the orthographic form, the phoneme inventar and the coding of the phonetic form and - of course - the size and coverage.
To be useful for technical applications such as Speech Recognition or Speech Synthesis the dictionary must be in digital and formally regular form. It may be augmented by additional information such as: occurance statistic based on a certain speech resource, syntactic labels, origin etc.

Pronunciation dictionaries are needed for a number of technical applications such as: Automatic Speech Recognition, Speech Synthesis, Media Indexing, Forensic Phonetics.

Phonetic coding of so called canonic pronunciation requires a strict code of conventions to achieve reliable results. The BAS has developed a BAS Standard for German Pronuciation that simplifies the coordinated work of larger groups of phonetitians on the same pronunciation dictionary.