2825 Commits

Author	SHA1	Message	Date
Joe Taylor	a8200fdfdd	Rempove remaining QRA64 code and direcories.	2021-02-23 09:20:46 -05:00
k9an	4b1077f897	Fix a mis-spelled warning message.	2021-02-20 08:39:53 -06:00
Joe Taylor	7192f53e68	Code cleanup.	2021-02-19 09:47:30 -05:00
Joe Taylor	b156ecd3a1	New method for measuring SNR of decoded Q65 signals.	2021-02-19 09:44:03 -05:00
Joe Taylor	23eb287449	Correct a typo.	2021-02-16 12:55:59 -05:00
Joe Taylor	10e08b0e94	Much better algorithm for detecting and removing birdies in Q65.	2021-02-12 13:26:22 -05:00
Joe Taylor	1aad2fdbda	Q65: Supppress birdies and don't display decode lines with blank message.	2021-02-11 11:30:35 -05:00
Joe Taylor	3962a1eeba	Merge branch 'release-2.4.0' of bitbucket.org:k1jt/wsjtx into release-2.4.0	2021-02-10 13:56:56 -05:00
Joe Taylor	6395087e6f	Tweak the averaging scheme: now straight avg up to N=4, then expenential with ntc=4.	2021-02-10 11:14:46 -05:00
Charlie Suckling	b58ce90a9f	Now generates a filename for 15s periods that works with averaging	2021-02-09 13:47:40 +00:00
Charlie Suckling	d91404bbb1	Modification to make filenames compatible with Q65 averaging	2021-02-09 13:47:33 +00:00
Joe Taylor	a6779ecda3	Use exponential message averaging (ntc=3) rather than simple sums in Q65.	2021-02-08 11:10:02 -05:00
Joe Taylor	3ccf949691	Save nrc (number of q[0124] iterations) to q65_decodes.dat.	2021-02-08 09:24:16 -05:00
Joe Taylor	f9a771b422	Q65 parameter ibwa should not be larger in Normal/Deep than in Fast. Corrected.	2021-02-06 09:43:14 -05:00
Joe Taylor	ade40fabf5	Be sure that Q65 nsps cannot remain undefined, and can't stop jt9[.exe].	2021-02-05 14:14:22 -05:00
Joe Taylor	7180c40754	Fix the source of a bounds error in JT4 decoder. (Why was this not a problem in v2.3?)	2021-02-05 14:03:18 -05:00
Joe Taylor	ebd41fabcc	Don't use CWD writes such as 'fort.33' in production versions!	2021-02-03 15:20:37 -05:00
Joe Taylor	9ca57f9a5c	Protect against a possible bounds error.	2021-02-02 11:39:29 -05:00
Joe Taylor	14d18306e6	Use single_decode and lagain as intended in Q65.	2021-02-02 09:58:04 -05:00
Joe Taylor	5850317de0	Speed improvements for Q65 multi-decode.	2021-02-02 09:27:40 -05:00
Joe Taylor	7d069c5617	Don't get too close to the Wide Graps frequency limits.	2021-02-01 15:52:25 -05:00
Joe Taylor	34cb4c502d	Eliminate some redundant candidates.	2021-02-01 14:23:49 -05:00
Joe Taylor	72a4158898	First working multi-decode procedure for Q65.	2021-02-01 13:28:59 -05:00
Joe Taylor	86ada8c9d4	Working toward multi-decode capability.	2021-02-01 10:54:04 -05:00
Joe Taylor	6519c07a71	Working toward a good multi-decode solution ...	2021-01-31 15:10:50 -05:00
Joe Taylor	f7cde117fe	Extens orange sync surve to the full displayed frequency range.	2021-01-31 13:41:32 -05:00
Joe Taylor	cae3095174	Allocate and use 2d CCF only within q65_ccf_85().	2021-01-31 11:39:04 -05:00
Joe Taylor	ec6fe749f9	Update the Q65 test program.	2021-01-31 09:06:20 -05:00
Joe Taylor	b8ea4c1c2c	Zero the red curve in we're not trying for q3 decodes.	2021-01-30 15:23:30 -05:00
Joe Taylor	9e822d1806	Clean up possible cases where rms evaluates to 0.0.	2021-01-30 15:10:05 -05:00
Joe Taylor	2f5c368aad	Remove a diagnostic write.	2021-01-30 14:59:55 -05:00
Joe Taylor	14c4164651	Remove the if(snr1.lt.2.8) test, for now.	2021-01-30 14:52:53 -05:00
Joe Taylor	7bd1249452	Code cleanup involving ccf1, ccf2, and plotting thereof.	2021-01-30 14:12:02 -05:00
Joe Taylor	1e7c53e0e4	Fix several comments.	2021-01-30 11:49:08 -05:00
Joe Taylor	791b39fce5	Estimate rms of ccf2 without using the 2-dimensional ccf().	2021-01-30 11:36:43 -05:00
Joe Taylor	316f41996e	Correct a flaw in setting ipk. Better loop ordering in ccf_22.	2021-01-30 10:59:46 -05:00
Joe Taylor	330ff03079	Move writing the red/orange sync curves into a subroutine.	2021-01-29 14:43:06 -05:00
Joe Taylor	f056fc99ec	Write parameters for successful decodes to file q65_decodes.dat.	2021-01-28 13:01:52 -05:00
Joe Taylor	48713f0f38	Merge branch 'feat-fst280' of bitbucket.org:k1jt/wsjtx into feat-fst280	2021-01-28 07:43:14 -05:00
Joe Taylor	4e27b009f0	Code cleanup, looking toward possible implementation of -120F and -300F submodes.	2021-01-27 13:19:39 -05:00
Joe Taylor	29a2737cf7	Clean up spec64, now that QRA64 mode is gone.	2021-01-27 11:12:57 -05:00
Bill Somerville	bd02e367f2	Leave global FFTW3 clean up to main programs	2021-01-26 21:34:10 +00:00
Joe Taylor	fb13265f37	Make ibwb larger by 1.	2021-01-23 14:14:17 -05:00
Joe Taylor	1489513803	Decodes without "newdat" should not add s1 into s1a again.	2021-01-23 10:58:28 -05:00
Bill Somerville	899a7fd895	Merge branch 'develop' into feat-fst280	2021-01-22 22:27:34 +00:00
Joe Taylor	b64520591f	Don't smooth at all if nsmo = 1: i.e., set it to nsmo=0.	2021-01-22 15:16:21 -05:00
Joe Taylor	8a436b63dd	Separate message averaging for Odd and Even sequences.	2021-01-19 16:06:10 -05:00
Joe Taylor	0cc67bbc59	Separate accumulations of odd/even sequences in s1a. Still need to fix navg ==> navg(0:1).	2021-01-19 15:30:17 -05:00
Joe Taylor	39024253d2	Fix the oddball values of DT and f0 sometimes being displayed.	2021-01-19 14:11:21 -05:00
Joe Taylor	d815eb7592	Replace the 'nsmo' smoothing: it is indeed useful, sometimes. (Can we have it both ways?)	2021-01-19 09:24:30 -05:00