[CQ-Contest] Contesting Technology - Phone Skimmer Nears Beta Test

[Jim Brown]

On Fri,4/3/2015 11:25 AM, Marijan Miletic wrote:
> Five years ago there was a hope that mature speech recognition technology
> would be applied to limited vocabulary hamradio contesting.  It has become
> April fools joke instead sadly illustrating dumbing of our hobby.

Having just endured another weekend of truly awful audio with WPX SSB, I 
couldn't even get my ear/brain to decode many of them. I heard at least 
50 stations with audio so badly distorted, muffled, or wildly 
overprocessed that I could not copy them, and at least a hundred more 
where the poor audio quality made me take twice as long to complete the 
QSO.

I'm not suggesting that we don't use processing/eq/compression -- these 
are excellent techniques if done WELL, but they destroy speech 
intelligibility if done badly. I run my K3 with the three lowest 
frequency bands set for maximum cut and some cut of the fourth (400 Hz) 
band. The remaining bands are set flat. I have compression set for 10 dB 
on peaks, and I carefully set computer playback level so that the radio 
is not overdriven.

PLEASE -- the next time you are setting up for a SSB contest, either 
listen to your own audio on another rig or get another ham to listen 
critically to your audio.

73, Jim K9YC