Perceived difference between 42 db(A) and 43 db(A)

[Dave Plowman (News)]

From someone sitting next to you on the tube.

[Huge]

Console yourself with the thought that they'll be as deaf as a post in a few years.

[<me9]

3 dB is about the smallest difference that can be detected. 1 dB would be undedtectable, unless the nature of the noise was different.

[Andrew Gabriel]

Which is where you get by working backwards from the sone and phon units, which are units of perceived loudness and doubling of loudness.

Rather reminds me of a conversation I overheard in a car parts shop. Customer had upgraded from 100W to 200W speakers to make the louder, and was very dissapointed to find they were noticably quieter ;-) I don't suppose he payed much attention in Physics lessons...

[<me9]

1dB may be detectable with good ears on a steady note. Better than 3dB is difficult to discern on random noise that may be of a different nature.

[tony sayer]

In article , DanR writes

Naff all that you'd notice....

[Dave Plowman (News)]

1dB is the just detectable unit. 3dB clearly detectable.

Here's a reasonable explanation.

[Roger]

The message from "Dave Plowman (News)" contains these words:

That would surely depend on the state of the listeners hearing.

Tnp claimed that that was the definition of 1 dB but while 0 dB is based on the threshold of sound any connection between 1 dB and the smallest detectable change has to be fortuitous.

centimetre! Whatever next - lbf per square inch? :-)

[Andy Hall]

Did he have similar difficulties with the fluorescent tubes for under the car and the Wayne and Sharon banner for the windscreen?

[Andy Hall]

Not since they changed the firmware. There's now a parental lock on max volume if you want.

[Andy Hall]

Assuming they survive summer on the tube and don't end up as a dessicated pile in the corner by the escalator.

[Dave Plowman (News)]

Pardon?

[James Salisbury]

That speaks volumes about Which.... Have a look at

have a 41 dB bosch and live in flat and bedroom next to kitchen. I have had to get up at night to turnoff the buzzing mechanical clock on the cooker, but never had a problem with the fridge....

[djc]

As others have said it is unlikely to be noticable. Even if compared side by side the variation from one sample to another would probably be more important than the paper spec. Also parts wear over time and they get noisier, and loose bottles etc in the fridge can be noisy. For me the intermittent nature of the sound from a fridge matters most: a fridge running can be ignored likew much background noise, whereas the sound of one starting in an otherwise quiet room can be annoying. I chose a Liebherr on the basis that the noise level was 36-38dB. It is just audible when running and I am sitting 2m away, but usually drowned out by the traffic outside.

[Andy Wade]

Er, what? How can that be? The decibel is defined as 10 times the common logarithm of a power ratio and the A-weighting is just a frequency response curve, defined in terms of a static (time and level invariant) transfer function[1]. That 1 dB corresponds to a just-audible difference is surely no more than a rather convenient coincidence.

[1] See

[Dave Plowman (News)]

It may or may not be. I'm not sure why the Bel was chosen as the original unit. Things like feet and yards had some relativity to the body. The fact that the practical auditory scale ranges over approx 100 dB may not be coincidence.

[tony sayer]

In article , Dave Plowman (News) writes

Just tried this last nite I can determine a .8 dB difference on tone but only managed 1.5 dB on music and it didn't seem to matter what that was classic, dub, house, R&B or rock!....

[Huge]

You might be amused to look up the definition of the "Dol", also.

[Andy Wade]

It was originally called a "transmission unit" and originated in telephone transmission theory in the 1930s (at Bell Labs, not surprisingly).

Or perhaps the other way round: the fact that such coincidences existed made the dB into a practically useful unit - and hence its widespread use. (c.f. the Neper - same idea, but using natural logs - which is little used by comparison).

[The Natural Philosopher]

All right, is defined as something that 'just happens to be'.... I've certainly sat in front of loudspeakers for a few years of my life, and 3dB difference is very detectable. 1dB is detectable if you switch between. Any less than 1dB is possible to detect, but its not obvious.