Hi Jose,
We will be starting with tests of ROS 1 baud tomorrow but I will not
have any results until next week, after we have been able to make tests
over several days and under many different conditions. The tests with
ROS 16 baud have been finished and our results are as I have already
reported.
Perhaps if the spreading were much wider, say as much a 10 kHz or 20
kHz, the result "might" be better, but then nobody on UHF SSB has an IF
filter wider than 2.5 kHz anyway. It would probably take at least a SDR
on both ends, I think, but so far those are still rare, even though they
make excellent IF's for VHF and UHF transverters. So, wider spreading is
just not practical.
Whatever it is that is causing a "raspy" CW note, and "raspy" sounding
ROS tones, must be destroying the data modulation on the carriers, but I
do not know enough about the modulation technique or the autocorrelation
function that ROS uses to understand why that is causing ROS to fail.
Perhaps it is because EVERY tone in the bandpass is so badly distorted
that autocorrelation is not possible and decoding fails (i.e. is the
Doppler shift perhpas moving the carriers outside some very narrow DSP
filter?). As best I can remember from my college days (50 years ago!),
autocorrelation will only work if reoccurring signals are identified
among random noise, but if the tones are distorted so they appear too
much like the noise, correlation may not be possible. I am sure
experienced communications theorists can make a better guess than I
can! The Olivia tones are also "raspy" sounding, but Olivia survives
and ROS does not. When the tones sound pure, ROS does OK, but that does
not happen very often at fringe area reception on UHF, and mostly only
when there is propagation enhancement.
73 - Skip KH6TY
I promised to post the results of our attempts to use ROS on UHF on
this reflector, and this is what we have found. So, it looks like
Olivia is currently still the best digital mode to use on UHF, VHF,
or HF for normal (not EME) digital QSO's.
Skip, please do tell us. I am particularly quite curious about the
results of your tests.
73,
Jose, CO2JA