Roger (K8RI) wrote:
> jimlux wrote:
>> Mike Besemer (WM4B) wrote:
>>> How, it the real world of ham radio, do you experience hearing the
>>> difference of 3dB daily?
>>> I'm not talking about a lab.
>> 3dB is the difference between unreadable and barely readable against the
>> noise floor. The difference between the 3kHz and 2.4kHz IF filter on
>> SSB is a lot less than 3dB and readily audible on a weak signal (on HF).
>> It's probably the difference between somewhat noisy and full quieting
>> on a FM rig.
>> At threshold, 3dB is important.
>> At S9 or 60 over, 3dB is irrelevant.
> Not if that's your noise floor, which isn't all that uncommon up here in
> snow country when the wind blows. <:-))
> It was S7 last night and very steady, well...night before last now.
>> Now, if you start talking 1dB... 1dB is very close to a "just noticeable
>> difference" in audio levels..
> That is the definition of 1 db (just barely detectable) It just so
> happens our hearing is logarithmic so to us, the change sounds linear.
That's just coincidence, as it happens.
the dB is based on the Bel, which is the log10 of the power ratio, just
as the Neper is the ln of the power ratio.. Certainly, from a circuit
analysis or plotting standpoint, one could have used Bels or milliBels, etc.
I think deciBels got used because it's close to the jnd for sound
intensity, and there was a lot of activity in the acoustics area when
the mathematical theory around circuits and what have you was being
developed (probably because Bell Labs was doing lots of it?) I don't
recall seeing dB used in books and papers from the late 1800s, although
log scales certainly were. Even today, the use of dB seems to be
restricted to electrical engineering, acoustics, and vibration.
Structural engineers don't use log units for the most part. Astronomers
use visual magnitude, which is a totally different log base.Earthquakes
use a measure based on log 10 of the height of a trace on a standard
However... I've learned that the "Weber constant" (the ratio between two
levels that is just noticeable) isn't actually 1dB, but somewhat
different, and also it varies with the level. And, of course, the
constant varies for the sense (e.g. light intensity scales differently
from pitch which scales differently from sound intensity) and for the
actual test (you can tell the difference between levels in pure tones
differently than for filtered narrowband noise)
TowerTalk mailing list