4

There is a very common decibel-based sound level scale that I see casually tossed around all over the place with 0dB often described as "the threshold of human hearing" and typical non-specific entries like this:

  • 30dB: Whisper
  • 60dB: Normal conversation
  • 110dB: Sporting event
  • 130dB: Ambulance
  • Etc.

Examples (there are many) include:

As these charts get passed around and reproduced, there seems to be a bit of a drift in the specific numeric values, as well as inconsistencies in signal description (e.g. "dB" and "dB(A)" being used interchangeably).

The Question: What are the specifications of this scale (or do none exist)?

A Note: By "specifications", I mean the typical set of expected information:

  • Does it have a technical name / standardized definition?
  • What is the actual reference level and measurement unit? (In this scale, it usually seems identified as "the threshold of human hearing", but that obviously has a lot of variance and also I'm not sure what that is.)
  • Assuming it's referring to pressure, is there some specific distance to the sound sources identified in this scale?
  • Is it meant to apply to unmodified measurements, or is it meant to use some frequency weighting?
  • When using it, is there a standardized, or at least widely-recognized, prefix/suffix to the "dB" abbreviation that indicates use of this scale?
  • Where did the typically given measurements even come from in the first place?

My guesses:

You can stop reading here if you are not interested in my guesses about answers so far. You can continue reading if you want to see my reasoning and find it helpful to build off it.

Regarding the weighting:

I often see it related to hearing in humans, and also "the threshold of human hearing" in absolute units is presumably frequency-dependent. So from that context, it seems like it would describe weighted signals -- most probably A-weighted -- but it's hard to tell because usage of "dB" vs. "dBA" in such charts seems inconsistent, and most of them seem to just use "dB". Still, given that the target audience for these charts are generally human, it would be misleading to try to relate the decibel levels to familiar events ("whisper", "refrigerator", "jack hammer") while referencing a non-weighted signal, so A-weighting does seem likely. Is that the case?

Regarding the reference level and units:

Firstly, I'm assuming that this scale has sound pressure in mind (albeit at some unknown distance). It doesn't seem like it's describing power -- that wouldn't make sense in the contexts I always see this scale in, where distance to source matters. It also doesn't seem like it's describing intensity, because ... well, just a gut feeling (plus something about basing it off the "typical human's cross-sectional ear canal area" or whatever feels odd). That said, I've never seen one of these charts refer to "dB(SPL)" or similar.

So I've noticed, in real observations, a rough correspondance to the numbers in this scale when using various SPL meters in the past. For example, I have a Reed R8070SD that registers in the 75-95 dB(A) range when I am standing ~10ft away from a busy city street during rush hour, this is roughly similar to the 80-90 "dB" typically given for "loud traffic" in these scales.

Now, the SPL meter reference level is sort of (afaict) defined in IEC 61672 as "conventionally" 20µPa, but that "convention" is solidified by IEC 60942 (sound level calibrators; when calibrating the meters), which clearly defines the same 0dB = 20µPa reference level (note: most calibrators I've seen use a 1kHz tone [not sure if that's standardized] which has identity gain in all standard frequency weightings, so the weighting is irrelevant there).

Anyways, because I've noticed that SPL meters (calibrated with IEC 60942 compliant calibrators) seem to be in the general ballpark of the numbers in the mystery scale, and also because it seems likely that the scale is describing pressure, that makes me believe that the reference level in these scales is also 20µPa. Is this correct?

2
  • 1
    Comments are not for extended discussion; this conversation has been moved to chat.
    – Rory Alsop
    Oct 22, 2022 at 17:58
  • I think Timinycricket's answer is mostly correct, or at least in the right direction identifying OSHA as the authority on this one, I just haven't had time to come back to this yet.
    – Jason C
    Oct 22, 2022 at 18:12

1 Answer 1

4

Many of the “mystery scale” charts that I have seen include distance from the source. For example OSHA has one and includes distance and says 0dB is the threshold at 1kHz which would also be at 20µPa. (I have never heard any other number stated as the threshold.) they also specify that they use dBA weighted.

as for how they originally got those measurements I would assume they paid someone to take measurements and write them down. It’s just a comparison. The same way a doctor might convey the size of a tumor or a baby developing; “it’s the size of a grape” or “a pear” are all grapes and pears the same size? No, they vary a bit, but it’s just something to grasp other than a number. Who was the originator of those charts? Not really that important or useful unless you are writing a book on it’s origins.

4
  • TBH, I'm not even sure why they use 1 kHz. I assume it's coz it's kinda in the middle of the logarithmic freq range.
    – n00dles
    Oct 10, 2022 at 12:50
  • 1
    @n00dles 1kHz is conventional; it has identity gain in in all IEC weighting curves (e.g. A-weighting from 61672) as well as ISO and ITU-R curves, which means the weighting of choice does not affect calibrations done with 1kHz baselines. I assume it's 1kHz because that's just about the peak response frequency for human ears (there's some nice charts there with tympanic electrical responses) and also it tends to survive natural hearing loss (which has a marked steep drop-off from 2kHz and above), and it's a conveniently round number.
    – Jason C
    Oct 10, 2022 at 16:13
  • 1
    @Timinycricket I'm also noticing that all my research paths seem to end up leading towards OSHA as the actual origin of this scale, with the 20µPa @ 1kHz reference as well, A-weighted, and your answer supports that, too. I'm going to hold on accepting answers until I can confirm, or at least find a document #.
    – Jason C
    Oct 10, 2022 at 16:21
  • Right -- and without a distance from the source, the terms are meaningless, in many cases. Oct 17, 2022 at 14:20

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.