View previous topic | View next topic

decibels - (no doubt again)!

Page 1 of 1

bordonbert
1314078.  Tue Feb 19, 2019 12:13 pm Reply with quote

I find it very hard to come to terms with the fact that, in a programme of QI's calibre in terms of accuracy of factual information, you consistently drop the ball when talking about decibels. The latest was a repeat programme where the delightfully chirpy yet ultra-knowledgeable Mz Toksvig talked about "sound levels of 110dB." That is an absolutely meaningless phrase to make such an intelligent lady utter.

The decibel is not a measure of sound level or intensity, it is purely a comparative measurement of anything. It measures only the ratio of two quantities. It has no units. It is not tied in to any particular field of science. It is exactly the same as saying " 'X' times as big as". It can be used in the measurement of the size of ice cream cones or sand castles if desired. Can you find any meaning in this sentence which in its entirety is "That is a sound of 100 times as big as"? That is the exact equivalent of saying "That is a sound of 20dB". There is no base reference to which the ratio can be applied. It is often inferred that the reference is "the threshold of hearing" but that would require the term "dB SPL" to be used.

The decibel MUST have a qualifier to indicate what field we are working in and what the reference level for the dB scale is. In science and technology this leads to terms such as "dBm" - decibels compared to 1mW of power, "dBM" - decibels compared to 1W of power, "dBV" - decibels referenced to 1V, and of course dB SPL - decibels relative to sound pressures of 20μPa in air or 1μPa in water. Your own situation would possibly be best referred to 'A' weighting of sounds for the human ear, in which case it should perhaps have been "dBA".

I refer you to your Nemesis, Wikipedia, where you will find a long, (too long perhaps), list of the various references which should point out to you how important this actually is to technologists.

Pedantism? Possibly, but only by layman's standards and that does not lead to accuracy and remove ambiguity. This is QI and the struggle for total accuracy must continue!

 
Jenny
1314092.  Tue Feb 19, 2019 1:24 pm Reply with quote

We love pedantry!

Welcome to the forums bordonbert :-)

 
suze
1314121.  Tue Feb 19, 2019 7:46 pm Reply with quote

Strictly speaking, bordonbert is absolutely right.

But the layperson has only ever encountered the decibel as a unit of "loudness" (properly, sound pressure), and its use as a measure of "loudness" is sufficiently common that you'd need to give explicit definitions if you were using it for anything else.

When and only when the decibel is used as a measure of "loudness", there is a formal definition of 0 dB. That formal definition is awkward to work with, so there is also a conventional approximation to it. More details on those at post 1041908.

 
bordonbert
1314178.  Wed Feb 20, 2019 7:57 am Reply with quote

But you have missed the point Suze. What does that definition of 0dB mean? "0dB" by itself is deliberately not a reference point in any form, it only means "this one is the same size as that one." 0dB = 1x as big, that's all. That is how the decibel scale is made and kept universal to any field.

What you are now saying is that "The Layperson" is the guardian of technical standards. The fact that they have never encountered the decibel used in its correct way is neither here nor there, it is still plain wrong! Not "little wrong" but "120dB wrong"! They will continue to use it incorrectly until it is laid out before them in a clear way.

You are also correct that there is a "formal" definition of loudness as in sound levels where 20uPa is used as the reference, I actually quoted the same thing myself. I also added (deliberately) the fact that this is correctly designated "dB SPL" to indicate that it is in the field of sound levels and the 0dB reference is as we have said. Without the "SPL" in it it reverts back to the bland "so many times as big as something which we don't know".

It is exactly the same thing as saying that Birmingham is 101miles. It is meaningless unless you know from what. The fact that we use London as a reference point for so much is neither here nor there, you would never say it. I reiterate, it makes no difference that this is in common use in a degenerated incorrect form, it was wrong and it is wrong! Is it so hard to add "SPL" after dB (Sound Pressure Level) to get it right?

I really feel that care should be taken with this in future and I am disappointed to find that accuracy is NOT the byword on the show. For my money it should even be added as a correction in a future programme. There are engineers, technicians, physicists, mathematicians even some musicians all over the country grinding their teeth every time this is used in this way. They are the ones who use this everyday in the simple but specific highly accurate way it was intended and their standards should be respected not the standards of the layperson.

 
dr.bob
1314258.  Thu Feb 21, 2019 5:24 am Reply with quote

bordonbert wrote:
But you have missed the point Suze. What does that definition of 0dB mean?


bordonbert wrote:
You are also correct that there is a "formal" definition of loudness as in sound levels where 20uPa is used as the reference


I think you've just answered your own question, no?

bordonbert wrote:
I also added (deliberately) the fact that this is correctly designated "dB SPL" to indicate that it is in the field of sound levels and the 0dB reference is as we have said.


Your original post complained about Sandi talking about "sound levels of 110dB." Surely the fact that she prefaced the decibel level with the word "sound levels" was a pretty massive clue that we're discussing the field of acoustics, with all the implications for reference points that that entails.

bordonbert wrote:
Is it so hard to add "SPL" after dB (Sound Pressure Level) to get it right?


I dunno, why don't you ask the National Physical Laboratory? I don't think anyone is suggesting that "The Layperson" should be the guardian of technical standards, but the NPL has a pretty good shout at fulfilling that role. Here is an instructive poster produced by the NPL to explain acoustics in which they talk about decibels at length. They choose not to add "SPL", so why should QI?

bordonbert wrote:
I really feel that care should be taken with this in future and I am disappointed to find that accuracy is NOT the byword on the show.


There's a difference between accuracy and pedantry. Whilst the people who research QI enjoy a bit of needless pedantry, it must be remembered that QI is, above all, an entertainment show. It is not the Open University.

QI has never claimed to be 100% accurate. There's no practical way it could achieve that and, if it did, it would probably be so dry and dull that it would be cancelled after 3 shows. What it does do very well is to inject a bit of curiosity about this world we live in and, hopefully, inspire people to do their own research and find out more about a subject.

 
bordonbert
1314292.  Thu Feb 21, 2019 9:25 am Reply with quote

Quote:
I think you've just answered your own question, no?

No, absolutely not! What I said is correct, in ultra-simple terms the formal definition of 0dB is "1x as big", I don't think we are challenging that are we? It is also universally true that the decibel implies absolutely no units or references as it is in your quote. The only way it can be related to sound pressure levels, just the same as any other field, is when a reference is attached, such as the 20uPa which is the norm. The point is that that is not the only reference even for sound pressure levels! Just saying "sound levels of 110dB" does indeed indicate the field, but is it dB SPL, dBA, dBC, dBZ? These are all qualifiers used in that sound field.

Quote:
The decibel scale is a logarithmic scale applicable to any parameter, used to make quantities with a wide range of values more manageable. In the measurement of sound, we are concerned with the amplitude of the acoustic pressure, measured in pascals (Pa). The range of acoustic pressures that the human ear can detect is very wide - from the lower limit of hearing at around 20 ÁPa (2 x 10-5 Pa) to the threshold of pain at around 20 Pa. This very wide range of values is unwieldy, so it is converted into a logarithmic scale. This changes the range of values shown above to the more manageable range of 0 dB to 140 dB. Thus 0 dB is roughly the lowest level a normal person can hear, but it is not the lowest level possible!
(The National Physics Laboratory poster. My bold.)

They have already taken care to establish the field of enquiry as sound and they have already taken care to establish the reference point as 20uPa. I think they are making my own point pretty clearly.

Quote:
There's a difference between accuracy and pedantry. Whilst the people who research QI enjoy a bit of needless pedantry, it must be remembered that QI is, above all, an entertainment show. It is not the Open University.


Absolutely agreed, it is an entertainment show. So it doesn't matter that this confusing, (to the layman), area which is repeatedly used absolutely incorrectly in so many fields by laymen is not made correct by the simple addition of, in this case, "SPL" or "A" or "C" or "Z"? The principle must be upheld that those letters must NOT be included just to prove the point that accuracy isn't the aim and pedantry isn't the game?

I apologise for my incorrect assumption that accuracy of information was one of the tenets of the programme. I have obviously stumbled onto a mine here and will happily withdraw any complaint a wiser man. The layman has his programme back in his safe hands. (There are no smilies to wink and smile with as I say that in order to indicate it is a lighthearted retreat and not petulance.)

 
dr.bob
1314298.  Thu Feb 21, 2019 11:08 am Reply with quote

bordonbert wrote:
The only way it can be related to sound pressure levels, just the same as any other field, is when a reference is attached, such as the 20uPa which is the norm.


(my emphasis)

When discussing sounds in air, setting the value of 0dB to 20uPa is the norm. Given that, especially for an entertainment show on the tellybox, it is surely only required to highlight when you are discussing things that vary from the commonly agreed norm.

After all, if Sandi reads out something about a pint of beer, she could specify that we're talking about UK pints which measure 568 ml rather than US pints which only hold 473 ml. She could go on to point out that we're ignoring Australian pints which can be up to 570 ml, unless you're in South Australia where a "pint" is 425 ml and is referred to as a "schooner" in the rest of Oz. She could even mention that, in Israel, while the official definition of a pint is 568 ml, pubs use the term in a cavalier manner and a "pint" can be as small as 360ml. Then she could finish by mentioning that, in Canada, "un pinte" in French is twice as large as "a pint" in English, while historically a "Scottish pint" was the equivalent of 3 imperial pints until the 19th century.

She doesn't of course, do any of this for two important reasons:

1) QI is a British programme made for a British audience. As such, the audience will already realise that the commonly accepted value of "a pint" is 568 ml and will not require further explanation.

2) If she did do that, then the credits would be rolling by the time she finished the explanation and the vast majority of the audience would've switched off long before that point.

bordonbert wrote:
Just saying "sound levels of 110dB" does indeed indicate the field, but is it dB SPL, dBA, dBC, dBZ? These are all qualifiers used in that sound field.


So now you want Sandi to spend the entire show explaining about frequency weighting of decibel measurements, why they're important, and when they're applicable? I think even the XL edition would start to run out of time for that one.

bordonbert wrote:
So it doesn't matter that this confusing, (to the layman), area which is repeatedly used absolutely incorrectly in so many fields by laymen is not made correct by the simple addition of, in this case, "SPL" or "A" or "C" or "Z"?


Sandi has already established that we're dealing with acoustics. You've already stated that, in that field, a reference point of 20uPa is the norm.

What benefit would be gained by Sandi spending a long and tedious monologue explaining the precise reasons why the use of decibels in this throwaway remark was being used in exactly the way that the vast majority of the audience had already assumed it was being used? Wouldn't the finite amount of TV time available to the show be better served by giving the audience more interesting things to think about?

 

Page 1 of 1

All times are GMT - 5 Hours


Display posts from previous:   

Search Search Forums

Powered by phpBB © 2001, 2002 phpBB Group