On my phone here, so will be brief. Believe that I sent a link to a wav file a while back. It occurs in a large fraction of 20m files, so shouldn't be hard to find. Up to 15 successive nearly identical vectors differing only in one or two symbols.
> On Oct 1, 2015, at 11:47 AM, Joe Taylor <[email protected]> wrote: > > Hi Steve, > > On 10/1/2015 9:22 AM, Steven Franke wrote: >>> Meanwhile I've been gaining some experience with WSJT-X r5950 on >>> 20-40-80-160 meters. (Yes, using the "wrong" metrics.) Performance >>> generally seems good, though I have seen some pretty long decoding times. >> >> I think that a large part of this will be those strings of nearly >> identical undecodable “nuisance” vectors… > > Thanks for reminding me about this issue, which you had mentioned > before. I haven't paid attention to it, up to now. > > These are surely not seen in the simulated test data, right? > > Are you saying that with normal HF-type data (say, 14.076 MHz) we > frequently present many copies of identical (or nearly identical) > vectors mrsym() to the sfrsd (or kvasd) decoder ? How many is "many? > Do you have example .wav files for a few such cases? Or do they happen > so frequently that almost any .wav file should have examples? > > -- Joe > > ------------------------------------------------------------------------------ > _______________________________________________ > wsjt-devel mailing list > [email protected] > https://lists.sourceforge.net/lists/listinfo/wsjt-devel ------------------------------------------------------------------------------ _______________________________________________ wsjt-devel mailing list [email protected] https://lists.sourceforge.net/lists/listinfo/wsjt-devel
