Hi Now, everything is ok. Analysis gave the correct relation between modemcount (nin) and the speechoutputcount also on the receiving side, so, forget my previous mail.
The code is now running perfectely with 48kS/S-testdata, but a test on the air has not yet been possible, since the current conditions on 40m did not allow a qso with dc6ny. Unfortunately, I did not find another ham who could make a freedv-qso in 700-mode. I will report on our experience. 73, Alfred On 21.08.2015 16:05, wully wrote: > Hi David > > I have realized the code. The tests with files runs ok. But on the > outputstream on the rx-side, I have to upsample again to 48kHz (from > 8kHz after codec2-decoding). But here I get again 640 Samples, which, > wenn upsampled will only give 3840 S. But on the input, I have to use > 4096 Samples to get the 640 from freedv. So, there is again a mismatch. > Currently, I fill in 0-samples, when the supply from the freedv does not > fulfil the requests. This is not a critical problem, since it is on the > audio output side, but I want to be exact. > > But probably, I miss again a point. > > Alfred > > > > ------------------------------------------------------------------------------ _______________________________________________ Freetel-codec2 mailing list [email protected] https://lists.sourceforge.net/lists/listinfo/freetel-codec2
