Decibel

History

The decibel originates from methods used to quantify signal loss in telegraph and telephone circuits. The unit for loss was originally Miles of Standard Cable (MSC). 1 MSC corresponded to the loss of power over a 1 mile (approximately 1.6 km) length of standard telephone cable at a frequency of 5000 radians per second (795.8 Hz), and matched closely the smallest attenuation detectable to the average listener. The standard telephone cable implied was "a cable having uniformly distributed resistance of 88 ohms per loop mile and uniformly distributed shunt capacitance of 0.054 microfarad per mile" (approximately 19 gauge).

In 1924, Bell Telephone Laboratories received favorable response to a new unit definition among members of the International Advisory Committee on Long Distance Telephony in Europe and replaced the MSC with the Transmission Unit (TU). 1 TU was defined such that the number of TUs was ten times the base-10 logarithm of the ratio of measured power to a reference power level. The definition was conveniently chosen such that 1 TU approximated 1 MSC; specifically, 1 MSC was 1.056 TU. In 1928, the Bell system renamed the TU into the decibel, being one tenth of a newly defined unit for the base-10 logarithm of the power ratio. It was named the bel, in honor of the telecommunications pioneer Alexander Graham Bell. The bel is seldom used, as the decibel was the proposed working unit.

The naming and early definition of the decibel is described in the NBS Standard's Yearbook of 1931:

Since the earliest days of the telephone, the need for a unit in which to measure the transmission efficiency of telephone facilities has been recognized. The introduction of cable in 1896 afforded a stable basis for a convenient unit and the "mile of standard" cable came into general use shortly thereafter. This unit was employed up to 1923 when a new unit was adopted as being more suitable for modern telephone work. The new transmission unit is widely used among the foreign telephone organizations and recently it was termed the "decibel" at the suggestion of the International Advisory Committee on Long Distance Telephony.

The decibel may be defined by the statement that two amounts of power differ by 1 decibel when they are in the ratio of 10^0.1 and any two amounts of power differ by N decibels when they are in the ratio of 10^'N(0.1). The number of transmission units expressing the ratio of any two powers is therefore ten times the common logarithm of that ratio. This method of designating the gain or loss of power in telephone circuits permits direct addition or subtraction of the units expressing the efficiency of different parts of the circuit...

In April 2003, the International Committee for Weights and Measures (CIPM) considered a recommendation for the inclusion of the decibel in the International System of Units (SI), but decided against the proposal. However, the decibel is recognized by other international bodies such as the International Electrotechnical Commission (IEC) and International Organization for Standardization (ISO). The IEC permits the use of the decibel with field quantities as well as power and this recommendation is followed by many national standards bodies, such as NIST, which justifies the use of the decibel for voltage ratios. The term field quantity is deprecated by ISO 80000-1, which favors root-power. In spite of their widespread use, suffixes (such as in dBA or dBV) are not recognized by the IEC or ISO.

Definition

The ISO Standard 80000-3:2006 defines the following quantities. The decibel (dB) is one-tenth of a bel: 1 dB = 0.1 B. The bel (B) is  ¹⁄₂ ln(10) nepers: 1 B =  ¹⁄₂ ln(10) Np. The neper is the change in the level of a field quantity when the field quantity changes by a factor of e, that is 1 Np = ln(e) = 1, thereby relating all of the units as nondimensional natural log of field-quantity ratios, 1 dB = 0.11513… Np = 0.11513…. Finally, the level of a quantity is the logarithm of the ratio of the value of that quantity to a reference value of the same kind of quantity.

Therefore, the bel represents the logarithm of a ratio between two power quantities of 10:1, or the logarithm of a ratio between two field quantities of √10:1.

Two signals whose levels differ by one decibel have a power ratio of 10^1/10, which is approximately 1.25892, and an amplitude (field quantity) ratio of 10^1⁄₂₀ (1.12202).

The bel is rarely used either without a prefix or with SI unit prefixes other than deci; it is preferred, for example, to use hundredths of a decibel rather than millibels. Thus, five one-thousandths of a bel would normally be written '0.05 dB', and not '5 mB'.

The method of expressing a ratio as a level in decibels depends on whether the measured property is a power quantity or a field quantity; see Field, power, and root-power quantities for details.

Power quantities

When referring to measurements of power quantities, a ratio can be expressed as a level in decibels by evaluating ten times the base-10 logarithm of the ratio of the measured quantity to reference value. Thus, the ratio of P (measured power) to P₀ (reference power) is represented by L_P, that ratio expressed in decibels, which is calculated using the formula:

L P = 1 2 ln ( P P 0 )   N p = 10 log 10 ( P P 0 )   d B .

The base-10 logarithm of the ratio of the two power levels is the number of bels. The number of decibels is ten times the number of bels (equivalently, a decibel is one-tenth of a bel). P and P₀ must measure the same type of quantity, and have the same units before calculating the ratio. If P = P₀ in the above equation, then L_P = 0. If P is greater than P₀ then L_P is positive; if P is less than P₀ then L_P is negative.

Rearranging the above equation gives the following formula for P in terms of P₀ and L_P:

P = 10 L P 10 d B P 0 .

Field quantities and root-power quantities

When referring to measurements of field quantities, it is usual to consider the ratio of the squares of F (measured field) and F₀ (reference field). This is because in most applications power is proportional to the square of field, and it is desirable for the two decibel formulations to give the same result in such typical cases. Thus, the following definition is used:

L F = ln ( F F 0 )   N p = 10 log 10 ( F 2 F 0 2 )   d B = 20 log 10 ⁡ ( F F 0 )   d B .

The formula may be rearranged to give

F = 10 L F 20 d B F 0 .

Similarly, in electrical circuits, dissipated power is typically proportional to the square of voltage or current when the impedance is held constant. Taking voltage as an example, this leads to the equation:

G d B = 20 log 10 ( V V 0 )   d B ,

where V is the voltage being measured, V₀ is a specified reference voltage, and G_dB is the power gain expressed in decibels. A similar formula holds for current.

The term root-power quantity is introduced by ISO Standard 80000-1:2009 as a substitute of field quantity. The term field quantity is deprecated by that standard.

Conversions

Since logarithm differences measured in these units are used to represent power ratios and field ratios, the values of the ratios represented by each unit are also included in the table.

Examples

All of these examples yield dimensionless answers in dB because they are relative ratios expressed in decibels. The unit dBW is often used to denote a ratio for which the reference is 1 W, and similarly dBm for a 1 mW reference point.

Calculating the ratio of 1 kW (one kilowatt, or 1000 watts) to 1 W in decibels yields:

The ratio of √1000 V ≈ 31.62 V to 1 V in decibels is

(31.62 V/1 V)² ≈ 1 kW/1 W, illustrating the consequence from the definitions above that G_dB has the same value, 30, regardless of whether it is obtained from powers or from amplitudes, provided that in the specific system being considered power ratios are equal to amplitude ratios squared.

The ratio of 1 mW (one milliwatt) to 10 W in decibels is obtained with the formula

The power ratio corresponding to a 3 dB change in level is given by

A change in power ratio by a factor of 10 corresponds to a change in level of 10 dB. A change in power ratio by a factor of 2 is approximately a change of 3 dB. More precisely, the factor is 10^3⁄₁₀, or 1.9953, about 0.24% different from exactly 2. Similarly, an increase of 3 dB implies an increase in voltage by a factor of approximately √2, or about 1.41, an increase of 6 dB corresponds to approximately four times the power and twice the voltage, and so on. In exact terms the power ratio is 10^6⁄₁₀, or about 3.9811, a relative error of about 0.5%.

Calculations

In order to add or subtract levels the values that are expressed in decibel first must be divided by 10 (or 20). The definition of the decibel in ISO 80000-3 assigns a value to the decibel (approximately 0.11513) thus the division by 10 in formulas for the addition or subtraction of levels should be a division by 10 dB. Ignoring that can cause a mistake of a factor of 8.686 in the number of decibels. None of the formulas published in ISO standards regarding noise and acoustics take this into account and they shall be used as if the decibel has no value (or is unity).

Properties

The decibel has the following properties:

The logarithmic scale nature of the decibel means that a very large range of ratios can be represented by a convenient number, in a similar manner to scientific notation. This allows one to clearly visualize huge changes of some quantity. See Bode plot and semi-log plot. For example, 120 dB SPL may be clearer than "a trillion times more intense than the threshold of hearing".

Level values in decibels can be added instead of multiplying the underlying power values, which means that the overall gain of a multi-component system, such as a series of amplifier stages, can be calculated by summing the gains in decibels of the individual components, rather than multiply the amplification factors; that is, log(A × B × C) = log(A) + log(B) + log(C). Practically, this means that, armed only with the knowledge that 1 dB is approximately 26% power gain, 3 dB is approximately 2× power gain, and 10 dB is 10× power gain, it is possible to determine the power ratio of a system from the gain in dB with only simple addition and multiplication. For example: A system consists of 3 amplifiers in series, with gains (ratio of power out to in) of 10 dB, 8 dB, and 7 dB respectively, for a total gain of 25 dB. Broken into combinations of 10, 3, and 1 dB, this is: With an input of 1 watt, the output is approximately Calculated exactly, the output is 1 W x 10^25⁄₁₀ = 316.2 W. The approximate value has an error of only +0.4% with respect to the actual value which is negligible given the precision of the values supplied and the accuracy of most measurement instrumentation.

Supporting arguments

According to Mitschke, "The advantage of using a logarithmic measure is that in a transmission chain, there are many elements concatenated, and each has its own gain or attenuation. To obtain the total, addition of decibel values is much more convenient than multiplication of the individual factors."

The human perception of the intensity of sound and light approximates the logarithm of intensity rather than a linear relationship (Weber–Fechner law), making the dB scale a useful measure.

Decibels are still the commonly used units to express ratios in a number of fields, even when the original meaning of the term is obscured. Decibels are the traditional way of expressing gain or margin in such diverse disciplines as control theory, antenna and radio frequency transmission theory, and even assessment of nuclear hardness.

Criticism

Various published articles have criticized the unit decibel as having shortcomings that hinder its understanding and use: According to its critics, the decibel creates confusion, obscures reasoning, is more related to the era of slide rules than to modern digital processing, are cumbersome and difficult to interpret.

Representing the equivalent of zero watts is not possible, causing problems in conversions. Hickling concludes "Decibels are a useless affectation, which is impeding the development of noise control as an engineering discipline".

A common source of confusion in using the decibel occurs when deciding about the use of 10 × log or 20 × log. In the original definition, it was a power measurement, and as employed in that context, the formulation 10 × log should be used, as deci means one tenth. The user must be clear whether the quantity expressed is power or amplitude. It is useful to consider how power or energy is expressed, e.g., current × current × resistance,  ¹⁄₂ × velocity × velocity × mass. Where the power is a square function of a field variable (such as voltage, current, or pressure), then 10 × log is the correct expression for the square, or 20 × log for the field variable itself.

Quantities in decibels are not necessarily additive, thus being "of unacceptable form for use in dimensional analysis".

For the same reason that humans excel at additive operation over multiplication, decibels are awkward in inherently additive operations: "if two machines each individually produce a [sound pressure] level of, say, 90 dB at a certain point, then when both are operating together we should expect the combined sound pressure level to increase to 93 dB, but certainly not to 180 dB!" "suppose that the noise from a machine is measured (including the contribution of background noise) and found to be 87 dBA but when the machine is switched off the background noise alone is measured as 83 dBA. ... the machine noise [level (alone)] may be obtained by 'subtracting' the 83 dBA background noise from the combined level of 87 dBA; i.e., 84.8 dBA." "in order to find a representative value of the sound level in a room a number of measurements are taken at different positions within the room, and an average value is calculated. (...) Compare the logarithmic and arithmetic averages of ... 70 dB and 90 dB: logarithmic average = 87 dB; arithmetic average = 80 dB."

Acoustics

The decibel is commonly used in acoustics as a unit of sound pressure level. The reference pressure for sound in air is set at the typical threshold of perception of an average human and there are common comparisons used to illustrate different levels of sound pressure. Sound pressure is a field quantity, therefore the field version of the unit definition is used:

L p = 20 log 10 ( p r m s p r e f )   d B ,

where p_ref is the standard reference sound pressure of 20 micropascals in air or 1 micropascal in water.

The human ear has a large dynamic range in sound reception. The ratio of the sound intensity that causes permanent damage during short exposure to that of the quietest sound that the ear can hear is greater than or equal to 1 trillion (10¹²). Such large measurement ranges are conveniently expressed in logarithmic scale: the base-10 logarithm of 10¹² is 12, which is expressed as a sound pressure level of 120 dB re 20 μPa.

Since the human ear is not equally sensitive to all sound frequencies, noise levels at maximum human sensitivity, somewhere between 2 and 4 kHz, are factored more heavily into some measurements using frequency weighting. (See also Stevens' power law.)

Electronics

In electronics, the decibel is often used to express power or amplitude ratios (gains), in preference to arithmetic ratios or percentages. One advantage is that the total decibel gain of a series of components (such as amplifiers and attenuators) can be calculated simply by summing the decibel gains of the individual components. Similarly, in telecommunications, decibels denote signal gain or loss from a transmitter to a receiver through some medium (free space, waveguide, coaxial cable, fiber optics, etc.) using a link budget.

The decibel unit can also be combined with a suffix to create an absolute unit of electric power. For example, it can be combined with "m" for "milliwatt" to produce the "dBm". Zero dBm is the level corresponding to one milliwatt, and 1 dBm is one decibel greater (about 1.259 mW).

In professional audio specifications, a popular unit is the dBu. The dBu is a root mean square (RMS) measurement of voltage that uses as its reference approximately 0.775 V_RMS. Chosen for historical reasons, the reference value is the voltage level which delivers 1 mW of power in a 600-ohm resistor, which used to be the standard reference impedance in telephone circuits.

Optics

In an optical link, if a known amount of optical power, in dBm (referenced to 1 mW), is launched into a fiber, and the losses, in dB (decibels), of each component (e.g., connectors, splices, and lengths of fiber) are known, the overall link loss may be quickly calculated by addition and subtraction of decibel quantities.

In spectrometry and optics, the blocking unit used to measure optical density is equivalent to −1 B.

Video and digital imaging

In connection with video and digital image sensors, decibels generally represent ratios of video voltages or digitized light levels, using 20 log of the ratio, even when the represented optical power is directly proportional to the voltage or level, not to its square, as in a CCD imager where response voltage is linear in intensity. Thus, a camera signal-to-noise ratio or dynamic range of 40 dB represents a power ratio of 100:1 between signal power and noise power, not 10,000:1. Sometimes the 20 log ratio definition is applied to electron counts or photon counts directly, which are proportional to intensity without the need to consider whether the voltage response is linear.

However, as mentioned above, the 10 log intensity convention prevails more generally in physical optics, including fiber optics, so the terminology can become murky between the conventions of digital photographic technology and physics. Most commonly, quantities called "dynamic range" or "signal-to-noise" (of the camera) would be specified in 20 log dB, but in related contexts (e.g. attenuation, gain, intensifier SNR, or rejection ratio) the term should be interpreted cautiously, as confusion of the two units can result in very large misunderstandings of the value.

Photographers typically use an alternative base-2 log unit, the stop, to describe light intensity ratios or dynamic range.

Related units

mBm

mB(mW) – power relative to 1 milliwatt, in millibels (one hundredth of a decibel). 100 mBm = 1dBm. This unit is in the Wi-Fi drivers of the Linux kernel and the regulatory domain sections.

Np or cNp

Another closely related unit is the neper (Np) or centineper (cNp). Like the decibel, the neper is a unit of level. The linear approximation 1cNp =~ 1% for small percentage differences is widely used finance. 1   N p = 20 log 10 ⁡ e   d B ≈ 8 . 685889638   d B

Fractions

Attenuation constants, in fields such as optical fiber communication and radio propagation path loss, are often expressed as a fraction or ratio to distance of transmission. dB/m means decibels per meter, dB/mi is decibels per mile, for example. These quantities are to be manipulated obeying the rules of dimensional analysis, e.g., a 100-meter run with a 3.5 dB/km fiber yields a loss of 0.35 dB = 3.5 dB/km × 0.1 km.