Understanding Octave Band Analysis: Why Single-Number Ratings Hide the Truth About Your Room

Imagine going to the doctor for a check-up. She runs your blood work, glances at the results, and says: "Your average blood value is normal. You are healthy." You leave relieved. But buried in the lab report — the one she averaged into a single number — your cholesterol is dangerously high and your iron is critically low. Two serious problems, both invisible because they cancelled each other out in the arithmetic.

Single-number acoustic ratings do exactly the same thing to your room.

When someone tells you a room has an RT60 of 0.6 seconds, or that an acoustic panel is rated NRC 0.75, they are handing you an average. They are telling you the patient is fine based on the mean of six very different measurements. And just like that blood test, the average can hide the fact that one frequency band is catastrophically out of control while another is perfectly behaved. The room "passes" on paper and sounds terrible in person. The specification is met and the occupants complain.

Octave band analysis is the full blood panel. It breaks the acoustic assessment into six separate frequency measurements, each telling its own story. And it is the only way to know what is actually happening in your room.

What Is an Octave?

Before we get to the analysis, we need to understand the unit it is built on: the octave.

An octave is a doubling of frequency. If you start at 125 Hz and double it, you get 250 Hz. Double again: 500 Hz. Again: 1000 Hz. Then 2000 Hz. Then 4000 Hz. Each of these doublings is one octave. The term comes from music — on a piano keyboard, the note A at 440 Hz and the note A one octave higher at 880 Hz sound like the "same" note, just higher. Your brain perceives them as fundamentally related because one vibrates exactly twice as fast as the other.

In building acoustics, we use six octave bands centered on these frequencies:

125 Hz - 250 Hz - 500 Hz - 1000 Hz - 2000 Hz - 4000 Hz

These six bands span the range from deep bass rumble (125 Hz) to the bright shimmer of cymbals and sibilant consonants (4000 Hz). Together, they cover nearly everything that matters for speech intelligibility, music clarity, and noise control in buildings. Standards like ISO 3382-2:2008 specify that reverberation time measurements must be reported across this exact set of octave bands — not as a single number, but as six individual values.

The reason we use octave-width bands rather than, say, bands of equal width in Hertz is that human hearing is logarithmic. The perceptual difference between 125 Hz and 250 Hz feels the same as the difference between 2000 Hz and 4000 Hz, even though one jump is 125 Hz wide and the other is 2000 Hz wide. Octave bands respect how we actually hear the world.

The Standard Octave Bands for Building Acoustics

Each octave band is named by its center frequency, but it actually covers a range. The lower edge of the band sits at the center frequency divided by the square root of 2 (roughly 0.707 times the center), and the upper edge sits at the center frequency multiplied by the square root of 2 (roughly 1.414 times the center). Here is the full picture:

Notice the range column. The 125 Hz band does not just cover sound at exactly 125 Hz — it covers everything from 88 Hz to 177 Hz. A truck idling outside at 100 Hz, an air handling unit humming at 160 Hz, and the lowest note of a male baritone voice at 120 Hz all fall into this band. When we say "the room has a problem at 125 Hz," we mean the room has a problem across this entire 88-177 Hz range.

This matters because different building materials behave completely differently at different frequencies. A 25 mm mineral fiber ceiling tile might absorb 90% of sound at 1000 Hz but only 8% at 125 Hz. A concrete wall reflects almost everything at every frequency. A heavy curtain absorbs well above 500 Hz but does almost nothing below 250 Hz. These frequency-dependent behaviors are invisible when you collapse everything into a single number.

Why Single-Number Ratings Fail

The NRC Problem

The Noise Reduction Coefficient, defined in ASTM C423, is calculated as:

NRC = (a250 + a500 + a1000 + a2000) / 4

It is an arithmetic average of absorption coefficients at four octave bands — and it has two gaping blind spots. First, 125 Hz is excluded entirely. The lowest frequency band in the NRC calculation is 250 Hz, which means the entire range of deep bass where HVAC noise, traffic rumble, and male voice fundamentals live is invisible. Second, 4000 Hz is also excluded. A material's high-frequency performance, which matters for speech clarity and electronic sound reproduction, simply does not factor in.

But the more fundamental problem is the averaging itself. Consider two acoustic panels:

Both panels are rated NRC 0.75. They are interchangeable in any specification that says "NRC >= 0.75." But Panel A absorbs only 10% of sound at 125 Hz while Panel B absorbs 45%. In a room with a bass reverberation problem — and most rooms have one — Panel B is four and a half times more effective at the frequency that matters most. The NRC number conceals this difference completely.

The Single-Number RT60 Problem

The same averaging trap applies to reverberation time. When someone says "the room has an RT60 of 0.55 seconds," they almost always mean the average of RT60 at 500 Hz and 1000 Hz. Some consultants average all six bands. Either way, the single number hides what might be happening at the extremes.

Here is a real scenario. A medium conference room — 7m x 5m x 2.8m, volume of 98 m3 — fitted with a suspended acoustic ceiling, carpet flooring, and painted plasterboard walls. The acoustic consultant measures RT60 at each octave band and gets:

The average of the 500 Hz and 1000 Hz values is (0.55 + 0.45) / 2 = 0.50 seconds. This is excellent. It passes WELL v2 Feature S07 (formerly F74), which requires RT60 below 0.80 seconds for rooms under 570 m3. It passes the corporate fit-out specification. It passes every standard that only checks the mid-frequency average.

But look at 125 Hz. The RT60 is 2.8 seconds — more than five times the mid-frequency value. At this frequency, every sound bounces around the room for nearly three seconds before dying away. Bass energy from the HVAC system piles up. The room feels boomy and oppressive. Male voices sound muddy because their fundamental frequencies persist long after the consonants that carry intelligibility have faded. Video calls are compromised because the microphone picks up the low-frequency reverberation and remote participants hear an echoey, cavernous quality.

The room passes the test and fails the people who use it. The single-number RT60 hid a catastrophic problem at 125 Hz because it was averaged away by excellent performance at 500 Hz and above.

How Octave Band RT60 Calculation Works

The reason RT60 varies so dramatically across frequencies is that absorption coefficients are frequency-dependent. Every surface in a room absorbs a different percentage of sound at each octave band. To calculate RT60 properly, you must run the entire Sabine or Eyring calculation six separate times — once for each octave band — using the specific absorption coefficient of every surface at that frequency.

A Worked Example

Consider a room 6m long, 5m wide, 3m high — a volume of 90 m3. The surfaces are:

• Ceiling: 30 m2 of mineral fiber acoustic tile
• Floor: 30 m2 of commercial carpet on concrete slab
• Walls: 66 m2 of painted plasterboard

Now we calculate the total absorption A (in sabins, or square meters of equivalent open window) at each frequency by multiplying each surface's area by its absorption coefficient and summing:

A = sum of (S_i x a_i) for all surfaces

Look at the RT60 column. At 500 Hz and above, this room is well-controlled — RT60 between 0.32 and 0.47 seconds, excellent for speech. But at 125 Hz, the RT60 is 1.38 seconds, nearly three times the mid-frequency value. At 250 Hz it is still over a second.

The acoustic ceiling tile — the workhorse of the design — absorbs only 8% of bass energy at 125 Hz despite absorbing 90% at 1000 Hz. The carpet contributes almost nothing at low frequencies. And the plasterboard walls, which actually show their highest absorption at 125 Hz (0.10) due to panel resonance, cannot compensate for 126 m2 of surfaces that are nearly transparent to bass.

If you averaged the mid-frequency bands, you would report RT60 = 0.42 seconds and declare victory. Octave band analysis reveals that the room has a serious bass reverberation problem that will make it sound muddy and boomy despite excellent mid-to-high frequency control.

This is not an unusual outcome. It is the default outcome in rooms treated with thin, lightweight absorbers. Standard acoustic ceiling tiles and foam panels are designed to address mid and high frequencies. Without dedicated bass treatment — thick absorbers, tuned panel absorbers, or membrane absorbers — the 125 Hz band almost always runs away.

Third-Octave Bands: When Six Bands Are Not Enough

Sometimes you need finer resolution than six octave bands provide. Third-octave band analysis divides each octave into three sub-bands, giving you 18 measurement points across the same frequency range instead of six.

Within the "125 Hz octave," for example, you get three separate measurements at 100 Hz, 125 Hz, and 160 Hz. Within the "1000 Hz octave," you get 800 Hz, 1000 Hz, and 1250 Hz. The preferred center frequencies for third-octave bands are defined in ISO 266:1997, and they follow the same logarithmic spacing as full octave bands — each third-octave step is a frequency ratio of 2^(1/3), approximately 1.26.

Third-octave bands are used in several specific contexts:

Sound insulation testing (STC and Rw). When you measure the sound transmission loss of a wall or floor assembly, you measure it in third-octave bands from 100 Hz to 3150 Hz (or wider). The single-number rating — STC in North America per ASTM E413, Rw internationally per ISO 717-1 — is derived by fitting a standardized reference contour to the third-octave data. The contour-fitting process means that deficiencies at specific third-octave frequencies are penalized individually, not averaged away. A wall with a sharp dip at 200 Hz will score lower than one with uniform performance, even if their average transmission loss is the same.

Environmental noise assessment. Regulatory noise limits are often specified in third-octave or even narrower bands. Traffic noise regulations may set limits at specific third-octave frequencies where tire noise or engine noise concentrates. Industrial noise assessments per ISO 9612 use third-octave analysis to identify tonal components that carry additional annoyance penalties.

Detailed room acoustic analysis. When a room exhibits flutter echoes, resonance problems, or unusual modal behavior, third-octave analysis can pinpoint the problematic frequency with three times the resolution of octave-band analysis. A room mode at 140 Hz would show up as a spike in the 125 Hz octave band, but with third-octave analysis you can see whether the energy is concentrated at 100, 125, or 160 Hz — information that changes the treatment strategy.

For most architectural acoustic design — meeting rooms, offices, classrooms, restaurants — full octave band analysis at six frequencies is sufficient. Third-octave analysis adds value in specialized applications and in diagnostic work where you are trying to identify a specific problem rather than characterize general room behavior.

How to Read an Octave-Band Chart

When you run a simulation in an acoustic design tool or receive measurement results from an acoustic consultant, you will typically see octave-band data presented as a bar chart or "spectrum" plot. Here is how to interpret it.

The horizontal axis shows the six octave bands: 125, 250, 500, 1000, 2000, 4000 Hz, progressing from left (low frequency, bass) to right (high frequency, treble). The vertical axis shows the acoustic parameter — usually RT60 in seconds, but sometimes absorption coefficient, noise level in dB, or another metric.

What a Good Room Looks Like

A well-designed room has a relatively flat RT60 profile across all six bands. The bars should be roughly the same height from 125 Hz through 4000 Hz. In practice, perfect flatness is nearly impossible — some gentle rise at 125 Hz is normal and acceptable. A variation of plus or minus 20-30% from the mid-frequency average is typical in well-treated rooms.

For a meeting room targeting RT60 = 0.50 seconds at mid-frequencies, a good octave-band profile might look like:

The 125 Hz value is higher than the rest, but only by about 30%. The room will sound clear and balanced, with a slight warmth in the bass that most listeners find pleasant.

What a Problem Room Looks Like

A room with inadequate bass treatment shows a dramatically different profile — a steep downward slope from left to right, with the 125 Hz bar towering over the others:

The 125 Hz RT60 is more than four times the mid-frequency value. This room will sound boomy and muddy. Male voices will lack articulation. HVAC noise will linger. The mid-frequency average of 0.47 seconds looks perfectly acceptable, but the room's occupants will report discomfort, poor speech intelligibility, and difficulty concentrating.

The shape of the octave-band profile — flat versus tilted, smooth versus spiked — tells you more about a room's acoustic character than any single number ever could.

Standards That Require Octave-Band Analysis

The most important international standards in building acoustics require frequency-dependent analysis. A single-number RT60 is not compliant with any of them.

ISO 3382-2:2008 (Reverberation Time in Ordinary Rooms)

Section 6.2 of this standard specifies that "reverberation time shall be determined as a function of frequency in octave bands with center frequencies from at least 125 Hz to 4000 Hz." The standard is unambiguous: a single averaged RT60 value does not constitute a valid measurement. The result must be reported as six values (minimum), one per octave band, plus the measurement uncertainty at each frequency. Spatial averaging across multiple measurement positions is required, and each position must report its own octave-band set.

WELL v2 Feature S07 (Sound — formerly Feature 74)

The WELL Building Standard checks acoustic compliance per octave band. The reverberation time limit — typically 0.60 seconds for rooms smaller than 570 m3 — must be met at each octave band independently, not just on average. A room that achieves RT60 = 0.45 seconds at 500 Hz but 1.2 seconds at 125 Hz does not pass WELL S07, even though its average is well below the threshold. This is one of the most common reasons WELL acoustic assessments fail: the specifier checked the mid-frequency average, assumed compliance, and was surprised when the assessor tested all six bands.

DIN 18041:2016 (Acoustic Quality in Rooms)

The German standard for room acoustic quality specifies frequency-dependent RT60 targets with distinct limits at each octave band. Notably, DIN 18041 sets tighter limits at 125 Hz than at mid frequencies for certain room types, recognizing that low-frequency reverberation causes disproportionate problems for speech intelligibility. A room designed to meet DIN 18041 must demonstrate controlled bass reverberation — you cannot achieve compliance by treating only the mid and high frequencies and hoping the 125 Hz bar stays below the limit.

BB93:2015 (Acoustic Design of Schools — UK)

The UK standard for school acoustics specifies RT60 requirements per octave band, with limits that must be met at each frequency independently. Classrooms must achieve RT60 at or below the target across the entire 125-4000 Hz range. The standard also specifies background noise limits per octave band, recognizing that HVAC noise at 125 Hz can mask speech even when the overall dBA level is acceptable.

ASTM E413 / ISO 717-1 (Sound Insulation Ratings)

While these standards produce single-number ratings (STC and Rw respectively), the underlying data is always third-octave band sound transmission loss. The single number is derived from the frequency-dependent data by contour fitting, not by simple averaging. Deficiencies at specific frequencies are penalized. The raw third-octave data is always reported alongside the single-number rating, and experienced acousticians always examine the curve shape, not just the number.

The Most Common Issue: 125 Hz Going Red

If there is one takeaway from octave band analysis, it is this: the 125 Hz band is almost always the problem.

The physics are simple. Low-frequency absorption requires thick, heavy, or resonant treatment. Standard acoustic ceiling tiles (typically 15-25 mm thick mineral fiber) are effective above 500 Hz but nearly transparent at 125 Hz. Foam panels of typical commercial thickness (25-50 mm) are similarly ineffective at bass frequencies. Carpet absorbs well at 2000-4000 Hz but does almost nothing below 250 Hz.

Meanwhile, the hard surfaces that dominate most rooms — plasterboard walls, concrete floors, glass windows — reflect low-frequency sound almost perfectly. The result is a room where mid and high-frequency sound is well-controlled but bass energy has nowhere to go. It bounces back and forth between surfaces, decaying slowly, building up a muddy, reverberant quality that single-number ratings miss entirely.

The fix requires dedicated bass treatment: thick porous absorbers (100 mm or more), tuned panel absorbers (membrane absorbers designed to resonate at specific low frequencies), or Helmholtz resonators (cavity-based absorbers tuned to particular bass frequencies). These solutions are bulkier and more expensive than standard acoustic tiles, which is why they are so often omitted from specifications. The specifier checks the NRC box, the ceiling tile goes up, and the 125 Hz problem remains hidden until someone actually listens to the room.

Octave band analysis catches this before construction. Six bars on a chart, one of them towering above the rest, is an unmistakable signal that the treatment plan is incomplete.

Why AcousPlan Shows Octave-Band Analysis by Default

Every simulation in AcousPlan produces results across all six octave bands — 125, 250, 500, 1000, 2000, and 4000 Hz. The results page shows six bars, each color-coded: green if the RT60 at that frequency meets the target for your selected room type, amber if it is borderline, and red if it exceeds the limit.

This is not optional. There is no "show me just the average" mode. Because the average lies. It smooths over the 125 Hz spike. It conceals the 250 Hz weakness. It produces a number that looks compliant while the room sounds wrong.

When you add or change materials in AcousPlan's room builder, you can watch the six bars shift independently. Swap a standard 20 mm ceiling tile for a 50 mm high-density panel and the 500 Hz bar barely moves — it was already well-controlled — but the 250 Hz bar drops noticeably. Add a bass trap in the room corners and the 125 Hz bar finally comes down from its red-flagged position into the green zone.

This frequency-by-frequency feedback is what makes the difference between a specification that looks good on paper and a room that sounds good in practice. The six octave bands tell you the truth. The single number tells you a story.

What to Do Next

If you have been working with single-number ratings — specifying materials by NRC, checking RT60 as a single averaged value — start looking at the individual octave bands. Every material datasheet from a reputable manufacturer includes absorption coefficients at all six standard frequencies. Every acoustic standard requires frequency-dependent reporting. The data is there. It has always been there. The industry just developed a habit of averaging it away for convenience.

Run your room through AcousPlan's acoustic simulator. Enter your dimensions, select your surfaces, and look at all six bars. Pay special attention to the 125 Hz bar. If it is red while the others are green, you have found the problem that a single-number rating would have hidden from you.

That is what octave band analysis does. It tells you the truth about your room, one frequency at a time.