This article is the first part in a series on decibels, excerpted from Bob Katz’s book Mastering Audio: The Art and The Science.
So many of us take our meters for granted—after all, recording is simple: all you do is peak to 0 dB and never go over!
But things only appear that simple until you discover that with the same material, one machine says that it peaks to -1 dB, another machine shows an OVER level, and yet your workstation tells you it just reaches 0 dB!
To make things worse, among the expensive digital meters available, only a handful accurately convey the
information we really need to know.
In this chapter we will explore the different types of meters, the concept of the digital OVER, analog and digital headroom, gain staging, loudness, signal-to-noise ratio and we will also take a fresh look at the common practices of dubbing and level calibration.
Bob’s Top 10 List of Slippery & Confusing Audio Terms
10. INTENSITY… is a measure of energy flow per unit area. For practical purposes, sound intensity is the same as SPL (see below).
9. LEVEL…is a measure of intensity, but when used alone, because it can mean almost anything, it means absolutely nothing! To avoid confusion, the level figure should always be qualified by a ‘unit’ term, e.g. voltage level, sound pressure level, digital level.
Examples: 40 dB SPL, -20 dBu, -25 dBFS. Each suffix defines the reference. SPL (sound pressure level) is a measure of the amplitude or energy of the physical sound present in the atmosphere. 40 dB SPL and 0.002 Pa (Pascals) are the same pressure, the first expressed in decibels relative to 0 dB SPL, the second in absolute pressure units.
8. DECIBEL (dB)… is a relative quantity; it is always expressed as a ratio, compared to a reference. For example, what if every length had to be compared to one centimeter? You’d say, “this piece of string is ten times longer than one centimeter.”
It’s the same thing with decibels, though sometimes the reference is not explicitly stated but just implied. +10 dB means “10 dB more than my reference, which I defined as 0 dB.” Decibels are logarithmic ratios, so if we mean “twice as much voltage,” we say “6 dB more” [20 * log (2) = 6]. DBU, DBM, DB SPL, DBFS… are ratios with predefined references, so they can be converted to absolute values in volts, power, etc.
I believe the term dBu was coined in the 1960’s by the Neve Corporation; it means decibels unterminated, compared to a voltage reference of 0.775 volts.
dBm means decibels compared to a power reference of one milliwatt.
dBFS means decibels compared to full scale PCM; 0 dBFS represents the highest digital level we can encode.
Plurals. We do say “two decibels”, but we do not pluralize the abbreviation. We do say “two dee bee”, but we do not say “two dee bees”.
7. GAIN or AMPLIFICATION… is a relative term expressed in plain decibels with no suffix: it is the ratio of the amplifier’s output level to the input level. If an amplifier receives an input of -23 dBu and puts out an output of +4 dBu, it has 27 dB gain (without any suffix). See Sidebar.
6. ATTENUATION…when expressed in dB is an optional term for negative gain, e.g. a loss. Examples: 20 dB attenuation is the same as -20 dB gain.
5. SOUND PRESSURE LEVEL (SPL)… is a measure of sound pressure in dB relative to 0.0002 dyne/cm2 (0 dB SPL). 74 dB SPL is the typical level of spoken word 12 inches away, which increases to 94 dB SPL at one inch distance.
While we often see language like 95 dB SPL loud, this is both inaccurate and ill-defined, as loud refers to the user’s perception, and SPL to the physical intensity.
4. LOUDNESS… is used specifically and precisely for the listener’s perception. Loudness is much more difficult to represent in a metering system, in fact, it’s best presented as a series of numbers rather than as one overall figure of “loudness.”
Two pieces of music that measure the same on a flat level meter can have drastically different loudness. A true loudness meter makes a complex calculation using SPL, frequency content, and duration.
Exposure time also affects our perception; after a five minute rest, the music seems much louder, but then we get used to it again—another reason why it is wise to have an SPL meter around to keep us from damaging our ears.
3. INTRINSIC LOUDNESS…In the first edition I invented a term “absolute loudness”, but it wrongly gives
the impression that we somehow have control over the consumer’s volume control.
So I’m replacing that with a new term intrinsic loudness, which I define as the loudness of a program before the level is adjusted using the monitor control.
Since there is no SPL reference in a digital file, intrinsic loudness has no absolute units, but the term can be used in a relative way.
We can compare two programs’ intrinsic loudness by switching between them, adjusting the monitor control until they sound equally loud, and noting the decibel difference in the monitor positions.
Then we can say that program 1 has “2 dB more intrinsic loudness than program 2” though, for brevity, I may say program 1 is 2 dB “louder”, using quotation marks.
When I use the term hot CD or hot master, I am referring to a recording which has a high intrinsic loudness. Our perception of the program’s loudness is also affected by the behavior of the monitor DAC.
For if a program is so distorted that its analog reconstructed level would cause a certain DAC to overload, this DAC may appear louder due to the high frequency distortion.
Intrinsic loudness would not have been a meaningful term in the analog era because analog tapes and LPs do not have a consistent reference, but with digital 0 dBFS is always the same.
2A. AVERAGE VS. PEAK LEVEL…Which of these two waves is louder?
The answer is: Both have the same loudness. Both have the same average and maximum peak level.
The first wave is identical to the second except its polarity is exactly reversed.
Think of this picture as the graph of the movement of a drumhead seen from the front and back side.
The terms average and peak are a bit misleading if we consider the upgoing direction as positive and the down-going as negative, because the top wave has a maximum in the positive direction and the bottom has a maximum in the negative; however the absolute value of the greatest peak in each wave is the same, so we say each wave has the same maximum peak level.
Over a long period of time, the average of the positive and negative numbers must be zero, which is the static pressure of the atmosphere (the position of the drumhead at rest).
However, the ear generally ignores polarity when it considers loudness, so it sees this wave rectified (pictured), where all the negative segments have been converted to a positive, or more correctly, absolute value.
The average value of this rectified wave over time is somewhere between its peak value and zero; our meters determine this average depending on the method of averaging and the time length of the average.
Sometimes we set our SPL meters to read peak level, but generally we set them to read average sound pressure level because this correlates closer to how the ear hears.
For proper monitor calibration, a good SPL meter should use the RMS averaging method, as opposed to a simple averaging which can produce as much as 2 dB measurement error.
RMS ignores issues such as phase shift and tells us the true energy level of the recording. The word average in this book can refer to the true RMS level or the simple average; when it is important, I will specifically say RMS.
2B. CREST FACTOR, also known as PEAK-TO-RMS RATIO, is the difference between the RMS level of a musical passage and its instantaneous peak level.
In practice, any averaging meter may be used, e.g. if a fortissimo passage measures -20 dBFS on the averaging meter and the highest momentary peak is -3 dBFS on the peak meter, it has a crest factor of 17 dB.
It is extremely rare to encounter a piece of music with a crest factor greater than 20 dB, so this is the commonly cited maximum.
When the dynamic range (the difference between the loudest and softest passages) of a recording has been reduced, we say the material has been compressed, and that a compressed recording has a lower crest factor than an uncompressed one.
And the # 1 most confusing audio term is…
1. VOLUME… is usually associated with an audio level control, but is an imprecise consumer term. Volume is measured in quarts, liters and cubic meters! The words more properly used in our art are Level and Loudness.
The big problem is that consumers use the term ambiguously, to mean both the loudness they perceive and the position of the “volume control”—a perfect example of confusing gain with level!
So in this book I prefer to use the professional term monitor control. I rarely use the word volume except when speaking informally to clients or consumers, and occasionally succumb to saying “volume control” when referring to a consumer’s system.
Editor Note: This article is the first part in a series on decibels, excerpted from Bob Katz’s book Mastering Audio: The Art and The Science.
To acquire this book, click over to the Focal Press.
Return to article