# Talk:Decibel

## 0 dB = +8 volts in audio electronics

I believe this is notable, even though it's obscure, because it serves to document the history of audio level interconnect standards in the industry. Can you take a look at the source and help me make the below description more clear? The drawing of the meter on page 42, and on page 43 is the statement "On the scale, 0 db was chosen to be 8 volts, so -6 db represents 4 volts." which is a further basis for claims below. Oops; I just spotted the +6 dB instead of -6 dB typo on my part.

dB
dB – a 1950s consumer audio measurement scheme used without a suffix, with dB 0 calibrated to be +8 Volts (with −6 dB representing 4 volts), with a range of −22 dB/0 volts to +2 dB/10 volts.[1]

My purpose here is to find dB relationships to volts used in the audio industry that will lead to the modern defacto standard of -10 dBV as the reference level for consumer line-level audio. I'm currently reading every issue of Audiocraft in search of other relevant references.PetesGuide, K6WEB (talk) 19:12, 18 September 2018 (UTC)

Thank you, this helps. One thing that still puzzles me is the apparent claim that -22 dB corresponds to 0 V. Do you understand what that means? Dondervogel 2 (talk) 22:32, 18 September 2018 (UTC)
Oh, I see it now - the -22 dB correspondence is in the drawing. That seems dubious to me. Better to start from the stated 0 dB point (8 V) and work back to -6 dB (4 V), -12 dB (2 V) and -18 dB (1 V). But these values do not correspond to those in the drawing, so something is still not quite right.
Question: The voltage values are rms, right? (otherwise the formula for power does not make sense). Is there an implied value for the resistance?
Dondervogel 2 (talk) 10:35, 19 September 2018 (UTC)
Oh, very good catch. From a quick glance, I thought that −22 dB and 0 volts on the scale image were coincident, but they're not. 0 volts is somewhere lower than −22 dB, but I don't see a way to calculate it from the article or the meter image. I do believe the voltages are RMS. Can you describe how you equated -18 dB to 1V, when that doesn't agree with the drawing? I'm not fluent on decibel math.

dB
dB – a 1950s consumer audio measurement scheme used without a suffix, with dB 0 calibrated to be +8 Volts (with −6 dB representing 4 volts), with a range of 0 volts (lower than −22 dB) to 10 volts (+2 dB), with 1 volt being equal to −16 dB.[2]
Oh, also, do you have access to any of the ISO or IEC documents that define dBV and other audio level standards?PetesGuide, K6WEB (talk) 14:54, 19 September 2018 (UTC)
On this scale, 1 V would correspond to -18 dB, not -16 dB (each factor of 2 change in rms voltage results in a 6 dB change in level). The near-coincidence of -16 dB on the meter scale is misleading.
With this in mind how about
dB
dB – a 1950s consumer audio measurement scheme used without a suffix, with 0 dB, -6 dB and -12 dB corresponding to an rms voltage of 8 V, 4 V and 2 V, respectively.
I don't think you'll find an ISO or IEC standard defining dBV because use of that notation is deprecated by IEC. In IEC-speak the unit is the decibel (symbol dB), regardless of the physical quantity being measured.
Dondervogel 2 (talk) 16:44, 19 September 2018 (UTC)
I like it, except for not including 1 V, which I think will help readers relate this usage to other definitions that are defined relative to 1 V. And should there be some note about the alignment/misalignment of the scale?
dB
dB – a 1950s consumer audio measurement scheme used without a suffix, with 0 dB, -6 dB, -12 dB, and -18 dB corresponding to an rms voltage of 8 V, 4 V, 2 V, and 1 V, respectively.
PetesGuide, K6WEB (talk) 16:55, 20 September 2018 (UTC)
Perfect. I suggest waiting for 24 hours to allow others to comment, and if no one does, to add your latest proposal. Dondervogel 2 (talk) 18:56, 20 September 2018 (UTC)
Reading it again, I wonder if it is just an example, and not an actual indication. It mentions that they do put dB scales on voltmeters. Note that you can use such a scale to compare two values, without knowing the actual reference value, as long as both are made with the same load impedance. Even more, note that most voltmeters have more than one voltage scale, sometimes with the user supplying the decimal point. (In the one shown, that might apply for an 8V, 80V, and 800V scale.) It does mention dBm in the article. As well as I know it, audio, more specifically telephony, now uses a 600 ohm load, but that might not have been done at the time. Gah4 (talk) 19:36, 20 September 2018 (UTC)
I'm pretty sure the article does not describe a device with a switch that can be set to either dB or Volts, but we should read more carefully to be sure. Also, modern audio no longer uses 600 ohms except in very specific conditions, and only the POTS telephone system still uses it consistently. dBm is 600 ohm specific, but dBu is not, and neither is dBV. Read up on impedance bridging for a deeper understanding of why. And in particular, I think the topic of this addition may be one of the links in how we got from using dBm/dBu for audio to using dBV. PetesGuide, K6WEB (talk) 01:40, 21 September 2018 (UTC)

References

1. ^ Horowitz, Mannie. "The DB in Hi-Fi" (PDF). Audiocraft. 1: 33, 34, 42, 43.
2. ^ Horowitz, Mannie. "The DB in Hi-Fi" (PDF). Audiocraft. 1: 33, 34, 42, 43.

## optics

I believe that the distinction between amplitude and intensity, that is, square or not, in the case of optical signals needs more fixing. Since most of the time, one doens't measure the amplitude (of the electric or magnetic field) of an optical signal, but instead the power imparted to some system, often an output signal is a voltage proportional (maybe gamma corrected) to the power level. Gah4 (talk) 02:39, 11 January 2019 (UTC)

Yes, the application of decibels to images is usually quite fucked up, applied as if for a field value when the voltage really represents an optical power or energy as counted by collected photo-electrons. It's hard to find a source that admits that though. Actually, it's covered a bit in the next section, Decibel#Video_and_digital_imaging. Dicklyon (talk) 05:12, 11 January 2019 (UTC)
Maybe this is a bit tangential, but this is IMO a symptom of the messed up definition of the dB: to make it "simpler" (no factor when going from root-power to power), the unit has been ill-defined; if the factor of 2 had been included (so that e.g. the power level is the sum of the voltage level and current level), the level of any ratio would have been unambiguously defined. The optical case is a little muddy, but the quantity being measured needs to be specified, e.g. the sensor response (voltage), the intensity, etc. One weird one that is rooted in physics is pressure: is it a root-power or a power quantity? In acoustics, it is clearly treated as a root-power quantity, but in electromagnetism (as three components of the stress–energy tensor, it can only be a power quantity, since it is proportional to the square of EM field strength, a root-power quantity. —Quondum 19:04, 11 January 2019 (UTC)

## gamma correction

Continuing the previous note, some video signals are gamma corrected. A CRT monitor does not generate a signal linear in the input voltage, but some power, called gamma, of the voltage. To make CRT televisions easier to build, video signals are adjusted at the source (once), instead of in each individual receiver. Should this be discussed here? Gah4 (talk) 02:43, 11 January 2019 (UTC)

No. Nothing to do with logarithms (decibels), and covered at gamma correction. Dicklyon (talk) 05:09, 11 January 2019 (UTC)
The question is, how do you quote power levels or gains of signals that have had gamma correction applied. Gah4 (talk) 10:39, 11 January 2019 (UTC)
The gain applied to a signal doesn't depend on what the signal represents. Dicklyon (talk) 15:56, 11 January 2019 (UTC)

## re "missing something?"

With reference to this edit comment , I was alluding to that even in linear systems, there is no simple relationship P/P0 = (F/F0)2. I was sensitized to this only recently by one of the references (probably the Mills paper in Metrologia that I have). It is only true if either of two conditions holds: the impedance is frequency-independent (excluding phase), or the waveform has the same power spectrum. Either of these can be described as "idealized". I am not happy with the word "idealized", but I was finding an interim way of not reverting the revert of my edit, and simultaneously prompting a discussion. —Quondum 16:16, 1 March 2019 (UTC)

In that case I think all that is needed is the addition of some words in the body of the article, explaining the above . Dondervogel 2 (talk) 17:48, 1 March 2019 (UTC)

## Hickling

In edit on 2019-06-14 you removed a section mentioning Hickling, without a reference to who he is. Doing a [WikiBlame] I found it was added complete with reference in edit on 2013-09-06.

I'm not sure what to do with this information though... There has been talk in this talkpage about the encyclopedic merit of arguments that Decibels are confusing... Heddmj (talk) 10:41, 18 June 2019 (UTC)

That is news to me. How else can we measure sound level volume/pressure?--Mark v1.0 (talk) 16:43, 19 July 2019 (UTC)
Sound pressure, sound power and sound intensity are most easily understand when present in pascals, watts, and watts per square metre. The use of the decibel is confusing because one first needs to divide these intuitive physical quantities by some arbitrarily chosen reference value and then take the logarithm. Doing so obscures the original meaning, especially when one conveniently "forgets" to mention the reference value. Dondervogel 2 (talk) 20:01, 19 July 2019 (UTC)
On the other hand, the numerical ranges typical of dB are so much easier to deal with that measurements expressed in dB become very easy for practitioners to deal with and visualize and compare. Their use is not confusing to those who use them. Dicklyon (talk) 05:24, 20 July 2019 (UTC)
I guess it's fair to say that when used correctly, they're not confusing to those who understand them. In my experience they are rarely used unambiguously and often used by individuals with no training in acoustical or electrical engineering. In those circumstances the result is confusion. In the end though it is not my opinion that matters, but the views expressed in the publications we are citing. Dondervogel 2 (talk) 07:54, 20 July 2019 (UTC)