Noise Rating Curves Compared: NC, PNC, RC, NCB and RC Mark II

Five noise rating systems have shaped building acoustic design since 1957, and choosing the right noise rating method still matters — Noise Criteria (NC), Preferred Noise Criteria (PNC), Room Criteria (RC), Balanced Noise Criteria (NCB) and RC Mark II — and each was published to correct a specific shortcoming in the one before it. Knowing which correction each represents is what tells you when to use it. Knowing which are still primary methods in the current standard is what stops a specification citing a demoted one.

Five Noise Rating Systems Compared

The right-hand column is the one most guides omit. The current standard's three primary methods are the A-weighted survey method, expanded NC, and RNC for low-frequency fluctuating noise.

Where the Lineage Starts

Every noise rating method rests on two prior developments: an instrument capable of measuring sound repeatably, and frequency discrimination through octave-band analysis. From there the line runs through equal-loudness contours (Fletcher and Steinberg, 1924; the more widely cited Fletcher and Munson, 1933), to A/B/C/D frequency weighting first standardised in ASA Z24.3-1944 — of which A-weighting and C-weighting remain in common use, A reflecting the ear's frequency response and C used mainly for peak measurements.

NC, 1957: Tangency and Wide Adoption

The first NC curve was published by Beranek in 1957, using equal-loudness contours as its reference and evaluated by the tangential method. Its adoption accelerated through inclusion in ASHRAE guidance, which is why it remains the method most mechanical and electrical engineers reach for. It is still the most extensively used and standardised rating in technical literature — covered in more depth in our article on reading an NC rating.

PNC, 1971: The Correction That Never Landed

PNC addressed a genuine NC weakness — its inability to capture "hiss," high-frequency dominant sound, or "rumble," low-frequency dominant sound. Its curve is less steep at low frequencies and steeper at high frequencies than NC, giving a better spectral balance.

It failed on economics rather than physics. The tighter low-frequency limits demanded more extensive and more expensive noise control than most buildings could justify, experienced consultants found them impractical, and PNC was never incorporated into a standard. It is the clearest case in this lineage of a technically superior method losing to a buildable one.

RC, 1981: Straight Curves and a Quality Descriptor

ASHRAE's mid-1970s background noise survey gave Blazier the data to develop Room Criteria: straight, parallel curves at a constant −5 dB per octave, designed to be perceptually neutral — free of any dominant tone.

The rating is calculated in two steps. First, the mid-frequency average level, L_MF = (L500 + L1000 + L2000) / 3, which is the RC number itself. Second, a spectral balance check: low-frequency dominance (16–500 Hz) is labelled rumble, high-frequency dominance (1,000–8,000 Hz) hiss, and a balanced spectrum neutral. The same comparison flags where low-frequency energy is likely to cause audible rattling in light building elements — suspended ceilings, light fittings, doors, windows, ductwork.

NCB, 1989: The Same Idea on a Different Basis

NCB spans 16 Hz to 8,000 Hz across NCB-10 to NCB-65, with NCB-0 representing the audibility threshold for continuous sound in a diffuse field. It differs from RC in its calculation basis: the Speech Interference Level, SIL = (L500 + L1000 + L2000 + L4000) / 4, rather than the three-band mid-frequency average. Ratings are written NCB XX(YY), using the same rumble/hiss/vibration/neutral descriptor system as RC, though the dB thresholds classifying rumble and hiss differ slightly.

RC Mark II, 1997: A Finer Description of Imbalance

RC Mark II refines RC in two ways. Its curve is flat rather than sloped between 16 and 31 Hz. And it replaces simple rumble/hiss labelling with two calculated quantities — the energy mean spectral deviation factor and the Quality Assessment Index — evaluated across low (16–63 Hz), medium (125–500 Hz) and high (1,000–4,000 Hz) regions, corresponding to rumble, roar and hiss. A QAI of 5 dB or less is labelled neutral; above 5 dB the descriptor reflects whichever region deviates most, with additional vibration descriptors where relevant.

It also adds a subjective response layer, estimating whether occupants are likely to find a given spectrum acceptable, marginal or unpleasant against recommendations for that space type. That estimate is what most distinguishes it from the original RC.

Which Noise Rating to Use Today

NC remains the default and the most standardised, and it is a primary method in the current standard. RC, NCB and RC Mark II all offer a spectral-balance descriptor NC lacks, which matters where low-frequency rumble or high-frequency hiss is the actual complaint — but all three now sit in annexes rather than the main body of ANSI/ASA S12.2-2019, so a specification that names them should say so deliberately rather than assuming co-equal status.

The method the current standard promotes alongside NC is RNC, for low-frequency fluctuating noise with surging or turbulence — the case NC handles least well. ALTA Integra selects the right noise rating method per project through its acoustic engineering and design practice.

FAQ

References

1. ANSI/ASA S12.2-2019 (R2023), Criteria for Evaluating Room Noise (supersedes S12.2-2008, which superseded S12.2-1995).
2. L. L. Beranek, "Revised criteria for noise in buildings," Noise Control, vol. 3, pp. 19–27, 1957.
3. W. E. Blazier Jr., "Revised Noise Criteria for Application in the Acoustical Design and Rating of HVAC Systems," Noise Control Eng. J., pp. 64–73, 1981.
4. W. E. Blazier Jr., "RC Mark II: A refined procedure for rating the noise of HVAC systems in buildings," Noise Control Eng. J., 1997.
5. G. C. Tocci, "Room Noise Criteria — The State of the Art in the Year 2000," Noise News International, vol. 8, no. 3, 2000.