摘要 |
<p>Speech recording is effected in a GSM phone handset (100) by storing in a memory (116) speech frames during the presence of speech, one or more SID frames during the absence of speech, and data representative of the duration of the absence of speech. In this way memory (116) does not store silent speech frames, and utilisation of memory space is therefore particularly efficient. In addition, items such as a voice activity detector and a comfort noise estimator, which are already provided in the handset as part of the GSM system, are "re-used" by the invention, thereby making efficient use of already-provided hardware/software. <IMAGE></p> |