On Thu, 1 Jul 2004 19:43:02 +0000 (UTC), "Reg Edwards"
wrote:

I could take the average of a few replies.

Hi All,

Perhaps an exercise useful to approximate within 2 or 3 orders of
magnitude.

The problem with measuring against the physiological response (aka
hearing, seeing, and the other senses) is that we "own" the estimate
as a personal expression of our integrity (how could be wrong?).
Problem is that such measurements are invariably fraught with the
widest error spread imaginable.

For a simple correlation (based on the cool, calm, detachment of
instrumentation) I noted the specifications for my speakers (Pioneer
vintage mid 70s) which describe they are capable of producing a SPL
(sound pressure level) of 92dB @ 1M @ 1 Watt. They also supply a
chart of relative sensitivity across the band of 20-20000 Hz, Harmonic
distortion, Impedances (room and coil). The published variation shows
that premium equipment wanders to considerable degree (Z, nominally at
10 Ohms varies from 5 Ohms to a peaked 40 Ohms). Hence nearly every
audio specification is set at either 1KHz or 400Hz and those
practitioners know full well not to abstract them throughout their
hearing range (which they also know full well varies with even much
wilder swings).

Let's take that same 92dB @ 1M @ 1 Watt and compare to the several
reports of hi-Z headsets. The first translation is of course against
application. Headphones encompass far less distance than 1M. I will
arbitrarily assign a path link of 1cM for the ear canal and re-specify
my Pioneer speakers to 112dB @ 1W (basically with the ear pressed next
to the speaker cone).

If I simply derate the power to achieve to 2dB above the threshold of
hearing (110 of that same 112dB) the power is similarly derated to
1W / 10¹¹
or
10 picoWatts

On Thu, 01 Jul 2004 21:06:10 GMT, "H. Adam Stevens"
wrote:
I can report using Sennheiser hi impedance dynamic phones to monitor 600 ohm
balanced broadcast audio signals ... The signal qualified as a good listening level as I recall.
"good" is too inspecific to derate to threshold power level.

On 01 Jul 2004 22:57:11 GMT, (JGBOYLES) wrote:
a 1000 ohm crystal earphone. The diode drop would be around 250 millivolts, and the
listening level would be fine.
"fine" is too inspecific to derate to threshold power level.

On Thu, 01 Jul 2004 23:44:01 GMT, "owen.home"
wrote:
I dont think they were as high as 2k but they were not low Z so probably 600ohms and very
sensitive ... could just hear -60dBm
"just hear" suggests threshold at 1 nanoWatt or 1000 picoWatts
or
2 orders of magnitude worse than my speakers.

On Thu, 1 Jul 2004 20:55:15 -0500, "Nick Kennedy"
wrote:
iron diaphragm, engraved on the back
"Cannonball Dixie, E.F. Cannon Co., Springwater, N.Y., U.S.A"
I measured the resistance of each can at a little over 1000 ohms.
The sound was clearly audible but not strong. ... 0.084 uW.
"clearly audible" is inspecific, however
84 nanoWatts or 84000 picoWatts
or
nearly 4 orders of magnitude worse than my speakers.

On Fri, 2 Jul 2004 13:39:59 +0000 (UTC), Harry Whitfield
wrote:
Test with S G Brown Type F Headphones
0.001 [Vrms] Threshold [@ 1KHz]
[and in a follow-up]
The DC resistance is 3860 ohm.
We are in the region of 250 pW !
[and a third follow-up]
I (also) have a very old pair of Philips Hi-Fi headphones.
The spec is given as 2 x 600 ohm (they are stereo phones).
Sensitivity is given as 112dB at 1mW.
-whew!- 2dB level at 10 femtoWatts

Harry offers the wildest variation. His first SG Browns are
25 times worse than my speakers
however his Philips Hi-Fi headphones
3 orders of magnitude better than my speakers!

Hence, as an average, response to this question spans roughly 7 orders
of magnitude.

As the quality of unzipped (and unrevealed design) software goes, such
a spread could easily satisfy anyone, even boys in their treehouse
with a string and dixie cup telephone system.

What accounts for this ENORMOUS swing of response to a simple
enquiry? The human element of owning what amounts to "dead
reckoning." The perception of sound is individual, not absolute.
Through the survey above it also suffers in reporting too few
variables. In other words, just what dB level satisfies such
qualities as "just hear" "clearly audible" "good" "fine" and any other
subjective qualifiers? The nearly 7 decades of variation cover
roughly half the hearing range between threshold and pain - 140dB.
Obviously if we trimmed 40dB from each end, the middle zone of the
remaining 60dB suddenly becomes "average" listening levels for
"someone."

Well, I hope you guys enjoyed this tour of the fantastic improbability
of human sensation rendered to a fixed number. ;-)

73's
Richard Clark, KB7QHC