No. 194 · Acoustic Bulletin
Cosmic Hypothetical
Pressure Ceiling · 101.3 kPa
Series · Edition Two Filed under: Acoustics, Atmospheres, Catastrophe One sheet · Hand-made

A question of loudness.

A whisper is 30 dB. A jet engine is 140 dB. Krakatoa in 1883 was 310 dB at the source, heard 3,000 miles away. But there's a hard ceiling at 194 dB where the wave stops moving and starts breaking. Move the dial. Watch air give up.

Sound Pressure Level
110 dB
Peak Pressure (Pa)
6.32 Pa
Mode
Wave
VACUUM · 0 Pa abs ATMOSPHERIC · 101.3 kPa
REC · 4s history · 180 sps
RECORDING · WAVE
y: pressure · x: time
amp: 6.32 Pa
OVERPRESSURE · WAVE BROKEN
⚠ Air has exceeded its acoustic limit ⚠
· · · 1 atm above continuous record · pen at right 1 atm below · · ·
Drive 110.0 dB
i
Sound Pressure Level
110 dB
a rock concert
ii
Peak Pressure
6.32 Pa
62 ppm of atmospheric
iii
Intensity vs. Speech
100,000×
five orders of magnitude
iv
Effect on a Body
Painful
damage in minutes
Diagnosis Damaging

A rock concert from the floor. Your ears will recover from one night, mostly. Three nights in a row and they won't. The wave is still a wave, the air is still air, and the math is still benign. We are nowhere near the ceiling.

§ I

Working it out.

Live values · 110 dB
i.The Scale
The decibel
L  =  20 log10 (P / P0)

A logarithm of pressure ratios. The reference P0 = 20 µPa is the quietest pressure a healthy young human ear can detect at 1 kHz, defined in the 1930s. Every 20 dB is a 10× increase in pressure; every 10 dB, a 10× increase in intensity. The scale is logarithmic because human hearing is.

Worked out (current)
P = 20 µPa × 10(110/20) = 20 µPa × 316,228 = 6.32 Pa
ii.The Ceiling
Why air breaks at 194
Lmax  =  20 log10 (Patm / P0)

A sound wave oscillates above and below atmospheric pressure (101,325 Pa at sea level). The trough cannot dip below zero, because there is no such thing as negative absolute pressure. So the maximum amplitude a real wave can have is exactly one atmosphere. Plug that in, and the answer comes out a hair above 194.

Worked out (the ceiling)
Lmax = 20 log10(101,325 / 20×10⁻⁶) = 20 log10(5.07 × 10⁹) = 194.09 dB currently 84.1 dB below ceiling
iii.Comparison
Intensity vs. conversation
R  =  10(LLconv) / 10

Sound intensity is power per unit area, proportional to the square of pressure. So a 10 dB jump is exactly 10× the intensity; a 60 dB jump (whisper to chainsaw) is one million times. The reference here is normal speech at 60 dB. The scale is unforgiving once you start counting orders of magnitude.

Worked out (current)
R = 10(110 − 60)/10 = 105.0 = 100,000×
iv.Distance
How far before it's safe
dsafe  =  d0 · 10(L − 85) / 20

A point source spreads energy over a sphere, so intensity falls as 1/r². That means the level drops 6 dB for every doubling of distance. Solving for when it falls to 85 dB (the OSHA 8-hour exposure limit) gives the radius of permanent hearing damage. Atmospheric absorption isn't included, so this is a lower bound.

Worked out (from 1 m)
dsafe = 1 m × 10(110 − 85)/20 = 101.25 m = ≈ 17.8 m stand back about 18 m to be safe long-term
§ II

For context.

An aside / Why we're here
A · What sound is

A pressure wave that needs a medium

Sound is a longitudinal compression wave in some material: gas, liquid, solid. Air at sea level carries it at 343 m/s. The molecules don't travel; they jiggle, bumping their neighbors, and the disturbance propagates. Loudness is the size of those jiggles, frequency is how fast. Without a medium, no jiggling, no sound. Space is silent because space is empty. This is why the ceiling exists at all: the medium has limits.

B · The decibel

Named for Alexander Graham Bell

Engineers at Bell Labs needed a unit for telephone signal loss in the 1920s. They settled on the bel, base-10 log of a power ratio, then divided it by ten because whole bels were too coarse. The reference 20 µPa was chosen because that's what the average healthy ear can just detect at 1 kHz. So 0 dB isn't silence; it's the threshold of human perception. Negative dB exists, you just can't hear it.

C · Where this came from

A question about sonic booms

The page started with a thread that wandered: what's the loudest possible sound, then do sonic booms count, then what about a nuclear bomb. Two answers emerged. There is in fact a hard ceiling, set by the pressure of the atmosphere itself. And there are several ways past it, but none of them are sound anymore. The dial above is the conversation, made visible. Move it. Watch air give up.

§ III

Across the spectrum.

Pitch, plotted · 28 octaves end to end
∿ · ∿ · ∿

Pitch has its own ceilings

Loudness is bounded by atmospheric pressure · Pitch is bounded by black holes and atoms
Infrasound
below 20 Hz
Audible
20 — 20,000 Hz
Ultrasound & beyond
above 20 kHz
↑ Cosmic floor Logarithmic frequency · 28 octaves Atomic ceiling ↑
§ IV

The receipts.

References & assumptions

Numbers used

Atmospheric ceiling: 194.09 dB
P_ref = 20 µPa, P_atm = 101,325 Pa. 20·log₁₀(101325 / 20×10⁻⁶) = 194.09. Standard atmospheric acoustics. Above this, the rarefaction phase would require negative absolute pressure, which is unphysical.
en.wikipedia.org/wiki/Sound_pressure
Krakatoa: ~310 dB at source
Inferred from worldwide barograph readings and documented overpressure: ~172 dB at 100 miles, audible 3,000 miles away in Mauritius. Pressure wave circled Earth 3-4 times. Modern reconstructions place source intensity well above the acoustic ceiling, meaning it was a propagating shock for the first hundreds of km.
en.wikipedia.org/wiki/1883_eruption_of_Krakatoa
Tsar Bomba: ~224 dB at 1 km
50-megaton thermonuclear test, USSR, 30 October 1961. Largest weapon ever detonated. Source overpressure inferred from blast data; the precise dB value depends on distance and direction and is itself a shock measurement.
en.wikipedia.org/wiki/Tsar_Bomba
Perseus B-flat: ~10⁻¹⁵ Hz
Discovered 2003 by Andrew Fabian's team using Chandra X-ray imagery of pressure ripples in the Perseus cluster's hot intracluster gas, driven by the central black hole. Period ~10 million years; pitch 57 octaves below middle C. NASA description names the note B-flat.
chandra.harvard.edu/photo/2003/perseus
Atomic ceiling: ~10¹³ Hz
A wave can't have a wavelength shorter than the spacing between the particles carrying it. In typical solids that's ~0.2 nm at speed-of-sound ~5 km/s, giving ~25 THz max. Above this, "sound" stops being meaningful and you're describing thermal phonons.
Speech reference: 60 dB at 1 m
Standard reference value for normal conversational speech at conversational distance. ANSI/IEEE conventions. Pain threshold: ~120 dB. OSHA 8-hour exposure limit: 85 dB. NIOSH recommendations are similar.
osha.gov/noise

Honest caveats

On the wave clipping.
The visualization simplifies. A real shock wave isn't a clipped sine; it's an asymmetric pressure pulse with a steep leading front and a gradual tail. The moment of breakage at 194 dB is real, but what the wave looks like above the ceiling depends on the source. The clipped sine here is a teaching diagram, not a measurement.
On values above 194 dB.
Decibel readings beyond the acoustic ceiling are reported as overpressure, which is mathematically still a logarithm of a pressure ratio but no longer describes a wave. Strictly, you've left acoustics. The eardrum doesn't care about the distinction.
On the inverse-square distance.
Treats the source as a point and the air as lossless. Real sound also undergoes atmospheric absorption, which is frequency-dependent and significant for high pitches over distance. For the kinds of low-frequency, high-amplitude events at the top of the scale, shock decay also kicks in much faster than 1/r². Use the formula for a vibe, not a survey.
For the curious, not the rigorous.
Editorial science writing. Where the literature disagreed, the friendlier or more familiar number was preferred. Where assumptions were necessary, they're stated. Corrections welcome; pretensions to peer review, not.