The Snorri software has been developed at CRIN. It provides a means of editing and analysing a speech signal. The aim of Snorri is to increase the efficiency of our research in acoustic-phonetic decoding. With Snorri the user can record and play back sentences, compute and display spectrograms as well as label speech utterances. It also enables global investigation of speech labelled corpus and extraction of all the occurrences of a given phonetic pattern. More-over, coarticulation phenomena can be studied with other tools like formant tracking displayed in the F1-F2 plane. Snorri is hence an efficient and an easy-to-use tool for speech recognition.