Dear Andy,

Thank you very much for your educational piece on the signal to noise
improvement possible for various modes.

It might interest you that I have been testing the limits of aural CW copy
using narrow analogue filters. It appears that after some exercise the
human brain-ear system can adapt itself to a bandwidth of 5 Hz for slow
speed morse code; e.g. the CW identifier of non directional beacons.
This offers a 20 dB improvement compared to a standard 500 Hz CW bandwidth
and seems to be as good as it gets.

The irony is that these tests were triggered by the appearence of narrow
DSP filters. Had I known this 40 year ago, it should have been an
excellent opportunity to show the possibilities of very low power
transmissions at HF!

Best regards,
Roelof bakker, pa0rdt