Fine business Alfred, I am happy the mystery is solved. It is good to work together on these issues, and compare results.
Thank Steve for your profiling, the acquisition algorithm could use some more work, but I'm more interested in 700C performance at low SNR right now. Cheers, David On 10/02/17 00:33, wully wrote: > Hi David > > I have to apologize! After extensive timing tests, I have found out, > that in the case of 8000 I made an error with the buffer size counting > when not in sync. This produced the forementioned irregular sounds. The > times for both methods (7500 and 8000) are absolutely comparable, as > your results confirm. > > So now, I will work only with 8000. No need for 7500 any more. > > I will make further tests with Helmut, dc6ny, to find out more about the > behavior of 700c. > > Greetings, Alfred > > > > On 09.02.2017 08:09, wully wrote: >> Hi David >> >> Thank you for your extensive test. So, your result suggests, that the >> Downsampling just makes jump the freedv_rx() over the time limit in my >> implementation. I can see, that the cpu cycles jump up also when using >> cohpsk directly. But for the time beeing, I will stay with 7500, because >> the nasty sound is unacceptable. When you find a quicker accquisition, I >> can try again with 8000. >> >> Greetings Alfred >> >> >> >> > > > ------------------------------------------------------------------------------ > Check out the vibrant tech community on one of the world's most > engaging tech sites, SlashDot.org! http://sdm.link/slashdot > _______________________________________________ > Freetel-codec2 mailing list > [email protected] > https://lists.sourceforge.net/lists/listinfo/freetel-codec2 > ------------------------------------------------------------------------------ Check out the vibrant tech community on one of the world's most engaging tech sites, SlashDot.org! http://sdm.link/slashdot _______________________________________________ Freetel-codec2 mailing list [email protected] https://lists.sourceforge.net/lists/listinfo/freetel-codec2
