Dear all, I am currently working on the GSM FR speech encoder for my student final year and I am experiencing some problem with the ETSI GSM FR DTX test vectors: Even when I use the implementation described by GSM 06.32 (section A.2.1) for LAR and xmax averaging, the FR DTX test vectors are not bit exact with my C code. As someone of you ever experienced this problem? Are all ETSI DTX test vectors using another algorithm for LAR and XMAX averaging (which is not the one disclosed in 06.32)? Is there a mistake in ETSI TV (!!!???) Moreover, the last 10 "sync" test vectors (from synch150.cod to synch159.cod) are not bit exact with my FR C code whereas the 150 first synch test vectors are OK. Note: all the other regular FR test vectors are OK. Is someone of you faced to the same issue? I am looking forward to hearing from you as my training period is ended up in few weeks. Thank you very much for your consideration. Aurelie. |