The logarithm of the senses — why two 60 dB speakers aren't 120 dB

Picture a recording studio. A sound engineer switches on a single monitor and sets a pure tone to exactly 60 decibels. Then they turn on a second, identical speaker beside it — also 60 decibels. Intuition suggests a simple sum: 60 + 60 = 120 dB. And 120 dB is the level of a jet taking off, the threshold of pain, a risk of instant hearing damage. Yet the studio meter will read about 63 decibels, and a person will register only a barely perceptible change in loudness. No shockwave.

This gap between physical energy and what we actually hear is the heart of the puzzle behind the decibel. Because the decibel is not a unit in the classic sense — you don't measure it the way you measure meters or kilograms. It is a dimensionless, logarithmic record of the ratio between two quantities, always tied to a fixed starting point. To understand why physics, medicine, and engineering all use this seemingly odd scale, we have to look both at the anatomy of the ear and at the mathematics of logarithms.

The ear as a biological compressor

Human hearing is one of the most sensitive instruments nature ever built. The threshold of hearing for a 1 kHz tone corresponds to an intensity on the order of 10⁻¹² W/m² (one picowatt per square meter). Sounds strong enough to cause pain and damage inner-ear structures have an intensity of about 1 W/m². The extreme ends of human hearing are therefore separated by a factor of 10¹² — a trillion.

If you built a noise meter with a linear scale, its dial would have to span from 1 to 1,000,000,000,000. That would be hopelessly impractical. Evolution solved the problem differently: the ear and brain respond not to absolute increments of energy but to their ratio. This is the Weber-Fechner law — the smallest perceptible change in a stimulus is proportional to the intensity of the stimulus already present. For a sound to seem clearly louder, its energy must grow many times over, not by a fixed amount. A logarithmic scale captures this perfectly, compressing twelve orders of magnitude into a handy range from 0 to 120 decibels.

The basis for this "compression" is the anatomy of the ear itself. The ear canal acts as a resonator, adding roughly 10–15 dB of gain in the 2000–4000 Hz band, exactly where we hear best. The middle ear — the hammer, anvil, and stirrup — works as a mechanical impedance transformer: the eardrum's surface is about 17 times larger than the oval window, which dramatically raises the pressure delivered to the fluids of the inner ear. Without this mechanism, almost all the energy (99.9%) would reflect off the air–fluid boundary, a loss of about 30 dB.

In 1936 Stanley Smith Stevens refined Fechner's theory with a power law of loudness: subjective loudness (in sones) grows in proportion to intensity raised to a power of about 0.3. The practical consequence is striking — for a sound to seem twice as loud, the level must rise by roughly 10 dB. And a 10 dB increase means a tenfold rise in the physical energy of the wave. To the brain, a ten-times-stronger sound is merely "twice as loud."

Power versus amplitude: the mechanics of the decibel

Formally, the decibel (dB) is a unit outside the SI system, but accepted for use with it by the International Bureau of Weights and Measures. It is one tenth of a bel (B) — the bel turned out to be too large in practice, so its submultiple almost entirely replaced it.

How you write a decibel depends on what you measure. For power-type quantities (energy, intensity, flux), the level is defined through the base-10 logarithm of the power ratio: L = 10 · log₁₀(P / P₀). But many quantities are amplitudes — sound pressure, voltage, current. Wave power is proportional to the square of the amplitude (P ∝ p²), so after substituting:

L = 10 · log₁₀(p² / p₀²) = 20 · log₁₀(p / p₀)

That is why the factor in front of the logarithm is sometimes 10 and sometimes 20. These two formulas give the iron rules of the scale:

The key point: a decibel always needs a defined reference value (P₀ or p₀). A bare number of decibels with no starting point is physically useless — hence the suffixes like dB SPL, dBA, dBm, or dBV.

Why 60 dB + 60 dB isn't 120 dB

Back to the two speakers. The mistake is trying to add logarithmic levels linearly. Two independent, incoherent sound sources (that is, most real machines and instruments) combine not by their levels but by their physical intensities.

Taking the threshold of hearing I₀ = 10⁻¹² W/m², from the equation 60 = 10 · log₁₀(I / I₀) we find the intensity of one speaker:

I₁ = I₀ · 10⁶ = 10⁻¹² · 10⁶ = 10⁻⁶ W/m²

The second speaker gives the same, so the total intensity is I = 2 · 10⁻⁶ W/m². Convert it back to decibels:

L = 10 · log₁₀(2 · 10⁻⁶ / 10⁻¹²) = 10 · log₁₀(2 · 10⁶) = 10 · (0.301 + 6) ≈ 63.01 dB

Doubling the acoustic power in a space yields an increase of just 3 dB — regardless of the starting level. Adding a second identical source to 40, 60, or 100 dB always lifts the background by the same 3 dB. To get a real 120 dB out of 60 dB speakers, you would have to run a million (10⁶) of them at once — only a millionfold rise in power produces a 60 dB increase.

The same logic dismantles a related myth, that "doubling the decibels doubles the loudness." Going from 40 to 80 dB is a rise of 40 dB, meaning ten thousand times more wave power (10⁴). Subjectively? The sound will seem only about sixteen times louder (2⁴ = 16), because each 10 dB is just one "doubling" in perception.

The bel and the neper: two Scots behind the scale

The logarithmic description of signals did not appear from nowhere — and two of its units carry the names of Scottish scholars.

The bel honors Alexander Graham Bell, inventor and telecommunications pioneer. Working on voice transmission, Bell wrestled with signal attenuation in long copper cables. His interests were deeply tied to the physiology of hearing: his mother was profoundly deaf, and his wife, Mabel, lost her hearing at the age of five after scarlet fever. The bel itself was later born in the Bell Telephone laboratories as a tool for describing power losses in telephone lines.

The neper (Np) owes its name to John Napier (Ioannes Neper), the 16th-century Scottish aristocrat who in 1614, in Mirifici Logarithmorum Canonis Descriptio, introduced the concept of the logarithm to science. Unlike the base-10 bel, the neper rests on the natural logarithm (base e ≈ 2.71828) and refers directly to the amplitude ratio: L = ln(A / A₀).

Converting the two scales follows from changing the logarithm base. A change of 1 neper corresponds to an amplitude ratio of exactly e, and in decibels:

1 Np = 20 · log₁₀(e) ≈ 8.686 dB

and conversely 1 dB ≈ 0.1151 Np. Although the decibel rules everyday acoustics and electronics, the neper remains important in transmission-line theory (cables, optical fibers), where attenuation in Np/m follows directly from the exponential solutions of the wave-propagation equations.

What "0 decibels" really hides

It is widely believed that 0 dB means absolute silence. That is a misconception. 0 dB SPL (Sound Pressure Level) is merely a conventional reference — a sound pressure of exactly 20 μPa (2·10⁻⁵ Pa), set in the mid-20th century as the average hearing threshold of a healthy young person for a 1 kHz tone.

Since it is only a reference point, negative values are entirely natural. If the measured pressure is smaller than p₀, the fraction inside the logarithm drops below 1, and the logarithm of a number in the range (0, 1) is negative. People with exceptionally acute hearing can, in the most sensitive 2–4 kHz band, catch sounds at −5 or −10 dB SPL.

There is, however, a physical limit to silence in Earth's atmosphere. Air molecules are in constant thermal motion, and their collisions (Brownian motion) generate a chaotic background noise estimated at about −23 to −24 dB SPL. Below that limit you would have to pump the air out — and in a vacuum, sound does not propagate at all. The quietest places on Earth are anechoic chambers: Microsoft's laboratory in Redmond measured −20.35 dBA, and the record belongs to the Orfield Laboratories chamber in Minneapolis — an astonishing −24.9 dBA. In such silence the brain dramatically raises its sensitivity: you can hear your own heartbeat and the rush of blood, while the absence of sound reflections destroys any sense of spatial orientation. The record for time spent inside the Orfield chamber is 1 hour and 26 minutes.

The scale of noise and the limits of safety

The logarithmic nature of the decibel matters enormously for workplace-safety standards. Noise-induced hearing loss depends on the total absorbed energy — the louder it is, the shorter you may stay in the noise. The measure is the "exchange rate": the level increase that halves the permissible exposure time. Here regulators differ: OSHA uses 5 dB (90 dBA for 8 h), while NIOSH and WHO use an uncompromising 3 dB grounded in physics (85 dBA for 8 h; every +3 dB, a doubling of energy, halves the time).

A measure that demands relational thinking

The decibel is a brilliant tool: instead of juggling numbers with a dozen digits, physics uses a compact logarithmic scale perfectly matched to the physiology of the senses. You only need to remember a few rules. A decibel is never an absolute value — it always describes a ratio to a chosen reference point. A rise of 3 dB is a doubling of signal power. A rise of 10 dB is a tenfold increase in energy that a person perceives merely as "twice as loud." And logarithmic levels must never be added linearly, because doubling the number of sources raises the combined level by only 3 dB.

Using the decibel therefore requires abandoning linear intuition in favor of relational thinking — and that is exactly the key to understanding both the physics of sound and the biology of our perception. The ear doesn't so much "lie" as quietly take a logarithm, sparing us the madness of a trillion orders of magnitude.

Further reading

• Rufin Makarewicz, Dźwięki i fale (Adam Mickiewicz University Press) — an accessible account of acoustic-wave physics and perception, free of needless jargon.
• Zbigniew Żyszkowski, Miernictwo akustyczne (Scientific-Technical Publishers) — a classic of Polish acoustic metrology.
• Jerzy Sadowski, Akustyka w urbanistyce, architekturze i budownictwie (Arkady) — the mechanisms of noise propagation and methods of reducing it.
• Andrzej Dobrucki, Przetworniki elektroakustyczne (PWN Scientific Publishers) — how mechanical vibration becomes an electrical signal.
• BIPM, The International System of Units (SI Brochure), 9th ed. (2019) — the status of the decibel and the neper relative to the SI.