Started by phil_lancs2002 April 9, 2002

I'm working on a CELP AMR codec. The coder uses previous samples of
the input speech, for windowing or creating the past excitation.

So, my question is: what happened for the first frame ?

Thank you,


does anyone know how the network configure the AMR_HR (Half rate) rather than
AMR_FR(full rate)?
Which is the difference between them? As far as I know the AMR algorithm has 8
bit rate codec
but It is not clear to me how the channel codec mode (HR/FR) is changed from the
network and which
differences are between them.

Thanks in advance