[Elecraft] Re: roofing filters - when to use them

[Bill Tippett]

N9VV:
 >I have never understood the
obsession of the modern CW ops with *narrow* filters.
I use my brain and don't seem to miss any QSOs ;-)

         Ken there are two issues here.  You are
talking about how the ear/brain processes signals
in noise.  Indeed there are two different ways
people prefer to do that.  One is to use the brain's
DSP and a wide audio bandwidth.  The other is to
use a narrow audio bandwidth (i.e. the radio's DSP).
I also prefer relatively wide audio bandwidths and
use my brain's ~50 Hz filter to separate the signals.

         The second issue (the real one which
roofing filters address) is the generation of
spurious products within the receiver itself
which can interfere with the desired signal.
If you have a IMD mixing product zero beat with
the signal you are trying to copy, your ear/brain
will not be able to separate them.  I'm sure you
have heard these products which are produced when
two adjacent strong signals mix and you hear the
result as phantom bleeps and bloops in your
passband.  It sounds like CW tones but in some code
other than Morse.

         This second issue is really what roofing filters
address...i.e. eliminating the phantom interfering
products generated in the receiver itself.  This
becomes a big deal especially when you have many
very strong signals spaced very closely, such as in
a contest or a big pileup when trying to copy a
weak signal close by.

                                 73,  Bill  W4ZV