Hi Glen and David, In my testing of "parrot" derived from "freebeacon", mode 700D, I'm "seeing" sync holding up when it shouldn't.
I'm running relatively low input levels, similar to what an FM receiver might do when the mute is closed. state: Rx Sync peak: 1981 sync: 1 SNR: 16.3 triggered: 1 recordany: 0 rxDataCnt 9492 ncodec 14 <- valid signal state: Rx Sync peak: 2215 sync: 1 SNR: 16.2 triggered: 1 recordany: 0 rxDataCnt 9590 ncodec 14 state: Rx Sync peak: 124 sync: 1 SNR: 9.9 triggered: 1 recordany: 0 rxDataCnt 9632 ncodec 14 <- signal lost background noise into mic. state: Rx Sync peak: 126 sync: 1 SNR: 7.9 triggered: 1 recordany: 0 rxDataCnt 9632 ncodec 14 state: Rx Sync peak: 105 sync: 1 SNR: 7.0 triggered: 1 recordany: 0 rxDataCnt 9632 ncodec 14 state: Rx Sync peak: 118 sync: 1 SNR: 7.5 triggered: 1 recordany: 0 rxDataCnt 9632 ncodec 14 state: Rx Sync peak: 120 sync: 1 SNR: 8.6 triggered: 1 recordany: 0 rxDataCnt 9632 ncodec 14 state: Rx Sync peak: 113 sync: 1 SNR: 8.3 triggered: 1 recordany: 0 rxDataCnt 9632 ncodec 14 state: Rx Sync peak: 129 sync: 1 SNR: 5.7 triggered: 1 recordany: 0 rxDataCnt 9632 ncodec 14 state: Rx Sync peak: 119 sync: 1 SNR: 3.7 triggered: 1 recordany: 0 rxDataCnt 9632 ncodec 14 state: Rx Sync peak: 713 sync: 0 SNR: -7.0 triggered: 1 recordany: 0 rxDataCnt 9632 ncodec 14 <- me whistling to get it out of this silly state state: Rx Maybe UnSync peak: 130 sync: 0 SNR: -7.7 triggered: 1 recordany: 0 rxDataCnt 9632 ncodec 0 At this point, I'm hoping the SSB receiver will feed in sufficient noise and "sync" will "behave". Alan VK2ZIW On Wed, 15 Jul 2020 06:09:44 +0930, David Rowe wrote > Hi Glen, > > Yes cohpsk_ch could be used to apply amplitude variations. > Replaying a chunk of your stored samples that demonstrates the > issue is also a good approach. > > You can also run the Octave version of the 700D demod, which will > give a better view of the modems internal states. You won't be able > to measure BER, but you'll be able to see if the LDPC decoder is > converging, and if the demod is falling over in the fast ALC condition. > > - David > > On 14/7/20 5:09 pm, glen english wrote: > > OK, in addition to my actual recordings, > > > > that would be using : > > > > https://github.com/drowe67/codec2/blob/master/src/cohpsk_ch.c > > > > together with > > > > https://github.com/drowe67/codec2/blob/master/octave/cohpsk_ch_fading.m > > > > ? > > > > On 7/14/2020 5:23 PM, David Rowe wrote: > >> Hi Glen, > >> > >>> /* update mean amplitude estimate for LDPC decoder scaling */ > >>> > >>> ofdm->mean_amp = 0.9f * ofdm->mean_amp + 0.1f * sum_amp / > >>> (ofdm->rowsperframe * ofdm->nc); > >> Yes, that's used for LDPC decoding so could be related. > >> > >> It would also be useful to have a repeatable unit test to explore the > >> issue, e.g. a repeatable fading profile running with the command line > >> version of the modem. > >> > >> Cheers, > >> David > >> > >> > >> > >> > >> _______________________________________________ > >> Freetel-codec2 mailing list > >> Freetel-codec2@lists.sourceforge.net > >> https://lists.sourceforge.net/lists/listinfo/freetel-codec2 > > > > _______________________________________________ > Freetel-codec2 mailing list > Freetel-codec2@lists.sourceforge.net > https://lists.sourceforge.net/lists/listinfo/freetel-codec2 --------------------------------------------------- Alan VK2ZIW OpenWebMail 2.53, nothing in the cloud.
_______________________________________________ Freetel-codec2 mailing list Freetel-codec2@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/freetel-codec2
