Before you can read the parent note, you must own every symbol it throws at you. This page builds each one from nothing — a plain meaning, a picture, and the reason the topic needs it. Read top to bottom; each brick rests on the one below it. This is the foundation for Acoustic Loads — SPL, Octave Band Analysis.
Picture a drum skin. Air molecules are constantly bumping into it from both sides. The steady background pushing is atmospheric pressure — about 105 Pa (100,000 Pa). That never changes fast, so it does no shaking.
Sound is a tiny wobble on top of that steady value. When a sound wave passes, the pressure at one spot rises a little, then falls a little, then rises again — over and over. We give that wobble its own symbol.
Look at the figure: the flat dashed line is steady atmosphere; the wavy coral curve is the sound pressure riding on top. Why the topic needs this: everything — loudness, force on a panel, damage — comes from the size of that coral wobble, not the flat line underneath.
The wobble spends as much time positive as negative, so its plain average is zero — useless for measuring loudness. We need a number that says "how big is the swing" without cancelling itself out.
In the figure, the coral curve is p(t); the lavender curve is p(t)2 (always above zero); the mint dashed line is its average; and the butter line marks prms — the square root of that average.
Where does the 2 come from? Take a clean wave p(t)=ppeaksin(t). To get the "mean of the squares" we average ppeak2sin2(t) over one full wobble. The only quantity we need is the average of sin2(t). Look at the lavender curve in the figure: sin2 bounces between 0 and 1, and — because sin2+cos2=1 and both look identical just shifted — it spends exactly half its time above and half below, so its average is 21. Doing the honest integral confirms it:
mean of sin2=2π1∫02πsin2(t)dt=21.
So the mean of the squares is ppeak2×21, and the root of that is:
Why the topic needs it: the loudness formula, the intensity formula, and every force calculation use prms — it is the single honest measure of "how big the squeeze."
Recall Why can't we just use the peak?
Peak works too, but RMS ties directly to energy (energy goes as pressure squared), and loudness/damage are about energy delivered. RMS is the natural bridge. ::: RMS connects directly to energy because energy ∝ pressure², so RMS is the physically meaningful average.
Human ears (and rocket fairings) handle pressures from a whisper (20×10−6 Pa) to a jet (200 Pa). That is a ten-million-fold range. Writing "0.00002 Pa to 200 Pa" is clumsy. We want a scale where each step means "×10 bigger."
The figure shows the same pressure axis two ways: on top, a linear ruler where the whisper is crushed invisibly against zero; below, a log ruler where whisper, speech, and jet spread out into readable, evenly-spaced marks.
Now we assemble the symbols. Loudness is a log of the ratio of your pressure to a reference whisper.
Reading the number: every +20 dB means ×10 more pressure; every +6 dB means ×2 pressure (because 20log102≈6). Why the topic needs it: fairing specs are always quoted in dB, so you must fluently convert both directions.
The parent note derives where pref=20μPa comes from, using four more symbols. Meet them.
Chaining these gives the key bridge the parent uses:
I=ρcprms2
Recall Why does intensity go as pressure squared, not pressure?
Intensity is pressure × velocity, and velocity is itself proportional to pressure (v=p/ρc). Multiply pressure by pressure → pressure². ::: Because I=pv and v∝p, so I∝p2.
Two rocket engines, 145 dB each, roaring together. You cannot add the dB numbers. Here is the machinery, using every symbol above. First, a shorthand: from now on we write L for any single SPL value (a "level," in dB) — so L1, L2 are the levels of source 1 and source 2, and Ltotal is the level of the two combined. L is nothing new; it is just SPL wearing a shorter name.
Since doubling intensity is +10log10(2)≈+3 dB, two equal uncorrelated sources = +3 dB, never +145. This same "un-log, add energy, re-log" recipe is exactly how octave bands are summed into one overall level.
When many bands or sources combine, the same recipe gives the OASPL (Overall Sound Pressure Level) — the single dB number that summarises the whole sound after all its parts are energy-added:
LOASPL=10log10(∑i10Li/10) dB.
This is the quantity the prerequisite map below points to, and the parent note computes it for real octave-band data.
A real launch sound is not one clean wave — it is a jumble of slow and fast wobbles mixed together. We sort them by frequency.
The band runs from flower=fc/2 up to fupper=fc2, so fupper/flower=2 — exactly one octave.
Recall Why the
2 factors instead of just ± half?
So the center is halfway in a logarithmic (multiplying) sense: fc=flowerfupper, the geometric mean. On the log ruler from Section 3, fc sits dead-center. ::: Because on a log/frequency-ratio scale the natural midpoint is the geometric mean, giving fc/2 and fc2.
Why the topic needs it: octave bands convert one messy sound into a short table of numbers (one SPL per band) that engineers compare against the panel's natural frequencies — and that connects straight to 3.6.11 Random Vibration — PSD, Miles' Equation.