|[AMRadio] AM Broadcast Quality|
k4kyv at charter.net
Tue Dec 17 13:29:59 EST 2013
From: Larry Szendrei <ne1s at securespeed.us>
>>On 12/16/13 9:56 PM, Bry Carling wrote:
> I don't see what is so objectionable about a 8 or 10kHz wide signal if it
isn't bothering anyone else.
Me, neither. As Don had said, adjust bandwidth according to band congestion.
There are many facets of amateur radio, and it's all good as long as we
apply common sense and be considerate of one another. It is not
inconsiderate to occupy 20 KHz in an otherwise empty band, if that's what
floats yer boat.
"Communications quality" advocates claim that something like 300-2700 Hz is
all that is necessary for voice communications and that everything else
should be filtered out. That is not true. A lot of intelligibility is lost
when everything above 2700~ or even 3000~ is cut off. Bell Telephone labs
and others have demonstrated that frequencies in the 3000-4000 Hz range
convey a substantial amount of the intelligibility contained in the human
voice. Cutting off the highs at 3000~ or below leaves gaps in the data
stream received by our ears and processed by our brain, which subconsciously
fills in the missing information. Our brain's ability to fill in the gaps
works fairly well, but intelligibility is still compromised when the
information is re-constituted by subconscious guesswork, and not actually
present in the sounds we receive. Suppression of frequencies in the
2700-3500 Hz range reduces or eliminates the distinctions among many of the
consonant sounds, for example: s/c, th as in 'thin', f ; v, z, th as in
'the'; p and t, etc. That explains why phonetics are so important with SSB
communication, notably during contests, and why "zed" is the preferred
pronunciation of the last letter of the alphabet instead of "zee".
The low end of the frequency range below about 250~ should not be emphasised
so much that it overpowers the rest of the voice spectrum (as often happens
with the ESSB crowd and with the so-called 'East Coast Sound'), but those
frequencies still need to be there since they are essential for giving the
voice a natural quality and allow us to distinguish one person's voice from
another. Most natural human voices do not carry much energy below about 90
Hz or above about 3500 Hz, so using a transmitter whose audio frequency
response extends out to 5 kHz or beyond causes relatively little harmful
interference to adjacent channels. These artefacts of articulation may
sometimes be annoying, but the most objectionable interference results from
spurious sideband splatter caused by distortion and overmodulation, not from
the upper frequency components of a clean, well-modulated phone signal whose
highs are not boosted to an unreasonable level.
I often find voice signals over the air more intelligible when using the 8
kHz mechanical filter in my 75A-4 than when using the stock 6 kHz one, even
in the presence of substantial QRM.
A friend of mine whose native language is French but who was a student of
English, once asked me to transcribe the lyrics to a George Harrison song
she had on a 45 RPM single. I played and re-played the song and attempted to
write down the words, but was amazed at how many of the words I couldn't
understand because the singer slurred them or they were overpowered by the
instrumentation in the background. I hadn't previously noticed any problem
understanding the song when I heard it, but this made me aware that I was
getting the gist of the lyrics without actually understanding every word,
some of which were utterly incomprehensible in the recording. My brain had
always filled in the gaps to make me thing I was understanding everything.
This is the same phenomenon we experience when listening through static and
interference, and to excessively restricted "space shuttle quality" voice
communication, and why we often resort to the use of phonetics when giving
This email is free from viruses and malware because avast! Antivirus protection is active.
More information about the AMRadio mailing list
This page last updated 20 Nov 2017.