Transcriptions should normally be held as simple text files, with annotations (as above) clearly identified in ISO 8879 (SGML) compatible notation. Software that will align the digitally recorded sound wave with an orthographic transcription is available in prototype for several languages, and it is expected that it will shortly be on the market, and extended gradually to all the languages of Europe.
Databases are not recommended for the storage of language corpus data because of the heavy overhead of entering the material, and the slow performance of retrieval software.