Hello Tom, Not really, there's a "energy" parameter in each frame that ideally makes performance independent of the input speech energy/level.
Suggest you keep away from clipping, so +/-16000 peak is a reasonable level to shoot for. - David On 26/10/20 8:22 am, Tom Early wrote: > Is there an inherent RMS sweet spot for the incoming PCM to the 1600 and > 3200 Codec 2 methods? Let's assume the mic is linear and the speakers > are not over-powered. > _______________________________________________ Freetel-codec2 mailing list [email protected] https://lists.sourceforge.net/lists/listinfo/freetel-codec2
