I have been playing around with the Codec2 vocoder block. I have a
flowgraph where I put a C2 encoder and a C2 decoder back to back.
I’m puzzled with the expected stream format. In the coder, the intent
to create a vector of 50 bits (or bytes where the LSB contains the bit
value) as output. If we look closer, the output is packed, so the 7
bytes contain actual data and the 43 others are 0.
Is there a rationale about the codec stream format (shall it be packed
unpacked) ? Is is supposed to be uniform among all codecs ?
I’ve seen that the GSM coder uses a single byte output for the stream.
What is the advantage of using a single byte output compared to
an array ?