More NCURSES Tweaking
-tweak array, seems to update smoother now, immediate
-more concise, cleaner names for sync types
-array updates on DMR, P25P1, NXDN, PV types only currenlty (no dstar etc yet)
More P25 resetState tweaking/testing
Other under the hood tweaking
resetState and Minor Ncurses Printer tweak
don't run with -n yet, resets state, using to test to fix when input type changes, but won't decode properly
Change NXDN to C4FM Modulation and change RF_mod type.
Not sure if it actually affects performance, but its the correct type.
Decoding relies more on samplesPerSymbol than anything.
Add Github Build Version
Work on Pulse Audio System (Testing 2 Channel Stereo Output)
Work on NCurses Terminal (Still Work to Do)
Include install.sh and rebuild.sh scripts
Initial NCURSES Support
Its in there, and can be called with -N, but it doesn't display much yet aside from program name, mbelib, and time.
Just want to upload it so I can test it vs autobuild with cmake requiring ncurses libraries again
Had to fix line 644 with this
Had wrong detection type
else if ((strncmperr (synctest18, INV_NXDN_BS_VOICE_SYNC, 18, 1) == 0) || (strncmperr (synctest18, INV_NXDN_MS_VOICE_SYNC, 18, 1) == 0)) //supposed to be voice, not data?
Had to fix line 651 with correct detection type
else if ((strncmperr (synctest18, INV_NXDN_BS_VOICE_SYNC, 18, 1) == 0) || (strncmperr (synctest18, INV_NXDN_MS_VOICE_SYNC, 18, 1) == 0))
More Pulse Audio Tinkering
Had to disable Monitor Source Input
(Was working in Arch, but with some latency/buffer issues, makes LM pulse server crash/restart?)
Fix Monitor Source Audio Issue
Forgot to add 'now' variable to all sync types, would attempt to play source audio and decoded audio at same time on most types.
Add Source Audio Monitoring - But its a Work in Progress
Seriously, though, depending on which setting you have on, it sounds terrible.
Also, make ncurses no longer a requirement for the time being. Still want to work that into this version eventually.
NXDN48 Voice Decoding is Good Now
Just use squelch if using SDR++ or GQRX to feed it. RTL-SDR input mode and NXDN still makes tons of false positives, and doesn't work well at all. May just be a bad signal issue but still think the RTLSDR input uses wrong center freq/tuned freq/"optimizations", need to get back to that some time, maybe also implement squelch settings on rtl_fm mode.
Thanks to LouisErigHerve for his 'Improved NXDN Detection' update on his version, implemented it and works SO MUCH BETTER!
May still disable the if ((state->lastsynctype == 9) || (state->lastsynctype == 17)) condition in dsd_frame_sync.c if I think it leads to more consistent decoding vs false positives. USE SQUELCH!
