Wow, 49 GB :-)
Here it took 4600 seconds for a full phase search and a list length of 200.
6E6 list size? The calculator page says 325804 is optimum. Wouldn't it
be more promising to try a set of noise blanker settings to gain maybe
0.1...0.2 dB from the individual QRN from last night?
BTW i use the bad wx to wind another coil on a bucket. With 5 fully
wound buckets i think i can reach < 4.5 kHz at 300 mA! The 5th coil
which is just wound to 30% must stay as it is in the moment because i
want to complete a few more messages (3 and 12 characters) on 5170 Hz
when Alex is back in a few weeks...
I can try 16K19A (when it will be available) and 500 chars at 1 seconds
symbols maybe. We are in the spectrum of the impossible anyway, so it
will work, because it can't!
Am 12.11.2017 15:02, schrieb Paul Nicholson:
> The message will be repeated today starting 17:15 UTC,
> just like yesterday...
Great. So far no result from last night, although the decoder
is still working through a full phase search with 6Meg list size
(49 Gbyte RAM).
Last night was quiet. Another night should do it if the noise
> Coding 4K19A
Pity we can't do rate 1/16, would make a significant difference
with these long messages.