Topband
[Top] [All Lists]

Topband: On FT8, noise bandwidths, filters, and signal vs noise.

Subject: Topband: On FT8, noise bandwidths, filters, and signal vs noise.
From: Tim Shoppa <tshoppa@gmail.com>
Date: Thu, 20 Dec 2018 12:10:57 -0500
List-post: <mailto:topband@contesting.com>
I broadly enjoy the digital modes, especially RTTY, and have been using
some FT8 outside of contests.

A signal power that FT8 reports as being at -15dB, is easily heard and
copied by ear by any decent CW operator. I think a really good CW operator
could pull (maybe with a few repeats) callsigns out at strengths that FT8
would classify as weaker than -20dB.

This is a nice presentation about noise estimation in WSJT especially FT8:
https://tapr.org/pdf/DCC2018-KC5RUO-TheReal-FT8-JT65-JT9=SNR.pdf

The subject of noise bandwidth has come up several times and is important
to how the digital mode guys "push" their newest modes. There's a lot of
traps with this way of rating things, especially when you compare modes
(like FT8) that allow no human smarts to improve copy, against other modes
(like SSB, CW and in fact RTTY) that allow human smarts to improve copy.
There was a SUPERB article in NCJ several years back by a ZS RTTY operator
about all the tricks to use your brain to improve RTTY copy.

But also note that there's a lot of reasons why an operator might not want
to use a narrow filter for CW.

Especially on the low bands, in the face of lightning crashes, many of us
prefer to use wider bandwidths. In part this is because historically many
rigs had narrow filters that rung like the dickens.  (This was the trap of
"shape factors" as a selling point for a few decades). But even with modern
rigs with more optimal phase response a lot of us are more comfortable
especially putting in long hours with wider rather than narrower filters.

In VHF/UHF and EME weak-signal CW work, a lot of operators also liked
listening with wider RX filters too, often preferring Gaussian filter
shapes, and letting their ear pull the signal out of the noise.

The psychoacoustics of listening to narrow bandwidths for extended periods
leads to a lot of these preferences. There's an "acoustical paranoia" that
sets in if you listen to noise too long with a narrow bandwidth and don't
change things up (bandwidth, center frequency, something!) at least
occasionally. When you find yourself in that paranoia it seems like the
signal you are trying to hear is exactly the same shape and bandwidth of
your narrow filter.

And finally: one man's noise is another man's signal. Or even a Nobel
prize. This is the story of how NBS scientists unintentionally discovered
radio emissions from Jupiter by tracking down their noise source:
https://www.nist.gov/news-events/news/2017/08/field-dreams

Tim N3QE


>
>
_________________
Searchable Archives: http://www.contesting.com/_topband - Topband Reflector

<Prev in Thread] Current Thread [Next in Thread>