ANSI S12.60 Compliance Guide for Conference Rooms

By Marcus Chen · February 24, 2026

ANSI S12.60 Compliance Guide for Conference Rooms

ANSI/ASA S12.60 is best known as a classroom acoustics standard, but the performance targets map cleanly onto conference rooms where speech intelligibility is the entire product. This guide shows you how to evaluate a typical meeting room and bring it into a S12.60-style compliance zone: controlled background noise, controlled reverberation, and predictable speech levels across the listening area. You’ll learn how to measure the room, interpret results, prioritize fixes, and verify outcomes with repeatable numbers—not guesswork.

Prerequisites / Setup

Measurement mic: Class 1 or good-quality Class 2 measurement microphone (e.g., Earthworks, iSEMcon, miniDSP UMIK-1). Calibration file preferred.
Audio interface or recorder: With phantom power if needed. Stable gain control.
Software: REW, Smaart, or similar that can measure RT60/RT20, RTA, and SPL. For background noise, an ANSI-style sound level meter is ideal, but software + calibrated mic can be acceptable for field work.
Test signal / source: Small loudspeaker for swept sine/pink noise (or your installed ceiling speaker system if it can reproduce cleanly). For speech level checks, use a reference talker playback track.
Room info: Dimensions, ceiling height, surface materials, HVAC type and diffuser locations, occupancy expectations.
Targets to aim for (conference-room practical equivalents):
- Background noise: ≤ 35 dBA preferred (or ≤ NC-30/RC-30). For high-stakes boardrooms: 30–33 dBA.
- Reverberation time: 0.4–0.6 s in the 500 Hz–2 kHz bands for small/medium rooms; up to ~0.7 s for larger rooms if speech is still clear.
- Speech level: 65–70 dBA at listener positions for reinforcement, with a comfortable margin over noise (aim for +15 dB SNR or better in the 1–4 kHz region).

Step-by-Step Process

1) Define the room use case and performance targets

Action: Write down how the room is used and choose measurable targets.

What to do and why: A “conference room” can mean anything from a 6-person huddle space to a 30-seat training room. ANSI S12.60 focuses on speech intelligibility, so you need to decide what success means: natural speech only, speech reinforcement, or far-end conferencing with loudspeakers and microphones. Targets should reflect the strictest mode. If the room hosts hybrid calls, you’re designing for both local intelligibility and clean mic pickup.

Specific values to set now:
- Background noise target: 35 dBA max (30–33 dBA if executives complain about “hiss” or HVAC roar).
- RT target: 0.5 s (±0.1 s) average in 500 Hz, 1 kHz, 2 kHz octave bands.
- Reinforced speech target: 67 dBA slow at seating positions with system at nominal.
Common pitfalls: Vague goals like “reduce echo” lead to random treatments and no verification. Another pitfall is optimizing for the far-end (microphone clarity) while ignoring near-end comfort (overly dead room feels unnatural). Solve both with controlled RT and controlled noise.
2) Survey the room and flag the three usual offenders

Action: Walk the room and document surfaces, openings, and noise sources.

What to do and why: Measurements tell you what is wrong; a survey tells you why it’s wrong. In conference rooms, problems typically come from (1) too much glass/drywall (high RT), (2) HVAC diffuser noise or duct rumble (high noise floor), and (3) leaky partitions/doors (intruding noise).

What to record:
- Surface areas: glass wall square footage, bare gypsum, hard ceiling, carpet vs. hard floor.
- Ceiling type: ACT tiles (NRC often 0.55–0.75) vs. open ceiling (usually reflective unless treated).
- HVAC: diffuser count and type, return grille location, VAV box placement.
- Door and seals: undercut gap, perimeter gasket, door closer slams.
Common pitfalls: Ignoring table reflections. A big wood table acts like a reflector right under the talker, boosting early reflections into microphones and reducing clarity.
3) Measure background noise correctly (the make-or-break metric)

Action: Measure A-weighted SPL and spectrum with the room in “quiet use” condition.

What to do and why: If background noise is too high, no amount of EQ fixes intelligibility. S12.60’s spirit is that the room should be quiet enough that normal speech is understandable without strain. For conference rooms, HVAC is the dominant continuous noise; intermittent noise comes from corridors, neighboring rooms, and outside traffic.

Procedure:
- Set HVAC to normal occupied mode (not night setback).
- Turn off program audio and any masking system unless it’s always on during meetings (if it is, measure with it on).
- Mic height: 1.2 m (seated ear height). Use at least 3 positions: near center, near door, near loudspeaker.
- Measure LAeq for 60 seconds each position. Also log 1/3-octave spectrum from 50 Hz–10 kHz.
Targets and interpretation: If you see strong energy at 63–125 Hz, suspect duct rumble or rooftop units. If 1–4 kHz is elevated, suspect diffuser hiss or electronic noise (DSP/amps).

Common pitfalls: Measuring with HVAC temporarily off and calling it “compliant.” Another is measuring near a laptop fan and thinking it’s the room. Keep personal devices away from the mic.

Troubleshooting: If results vary wildly between positions, you likely have a localized source (diffuser or door leak). Move the mic in 0.5 m increments to “hot spot” the noise.
4) Measure reverberation time in the speech bands

Action: Capture RT (RT60 or T20/T30) and focus on 500 Hz–2 kHz.

What to do and why: RT determines how long speech energy hangs around, smearing consonants. Conference rooms often have acceptable “loudness” but poor clarity because RT is high above 1 kHz (glass, drywall) or uneven (flutter echo between parallel walls).

Procedure (practical):
- Use a small loudspeaker at typical talker location (near display end or center of table).
- Play a swept sine (preferred) or pink noise and measure impulse response in REW/Smaart.
- Mic positions: at least 3 seats. Use 6 if the room seats 12+.
- Calculate T20 or T30 if possible (T20 is common in smaller rooms). Average results.
Targets: Aim for 0.4–0.6 s in 500 Hz, 1 kHz, 2 kHz. If 2 kHz is 0.9 s while 500 Hz is 0.5 s, high-frequency absorption is missing (often bare glass).

Common pitfalls: Measuring RT with people in the room one day and empty the next. Occupancy affects RT, especially above 1 kHz. For a real-world conference room, measure empty (worst case) and estimate improvement with typical occupancy.

Troubleshooting: If RT results look unstable, check your noise floor. If background noise is within 20 dB of your decay tail, RT estimates can be wrong. Increase test level (without distortion) or measure at quieter time.
5) Fix noise first: HVAC, isolation, and electronic gain structure

Action: Reduce background noise sources before adding acoustic panels.

What to do and why: Absorption improves RT but doesn’t reduce HVAC hiss at the source. S12.60-style success requires a low noise floor so speech has headroom. Start with the things that move the needle the most per dollar.

Specific interventions:
- Diffuser hiss: Verify airflow. If diffuser face velocity is high, switch to a larger diffuser or add more diffusers to reduce velocity. As a rule of thumb, lowering velocity reduces broadband noise.
- Duct rumble (63–125 Hz): Check for flex duct “drumming,” sharp turns, or undersized ducts. Add duct liner or a silencer where appropriate.
- Door leakage: Add perimeter seals and an automatic door bottom. A 6–10 mm undercut can leak surprising mid/high noise.
- Electronic noise: If you measure hiss near loudspeakers, check gain staging: set DSP output and amplifier gain so typical speech sits around -20 to -12 dBFS in DSP meters; avoid maxing amp gain with low DSP output (raises noise floor).
Common pitfalls: Treating the room with absorption while ignoring a diffuser blasting directly over the table. That’s how you get “dead but still loud” rooms.

Troubleshooting: If you cannot change HVAC, consider sound masking only as a last resort for privacy—not intelligibility. Masking raises noise floor, which works against speech clarity unless the room is already very quiet and privacy is the priority.
6) Control reverberation with targeted absorption (not random foam)

Action: Add absorption to hit RT targets while keeping the room natural.

What to do and why: You want fewer late reflections and controlled early reflections. For speech, absorption should be effective from ~250 Hz to 4 kHz, with special attention to 500 Hz–2 kHz. Thin foam mainly works above 1–2 kHz and often disappoints.

What to install (practical specs):
- Wall panels: 50 mm (2 in) mineral wool or fiberglass, fabric-wrapped, density ~48–96 kg/m³. Place at first reflection points on side walls and rear wall.
- Ceiling clouds (if hard ceiling): 50–100 mm thickness with 100 mm air gap. This increases low-mid absorption and reduces “table-to-ceiling” flutter.
- Glass walls: If you can’t cover glass, add heavy curtains with 2x fullness. Curtains help mid/highs; combine with other treatments for full-band control.
Placement guidance: Prioritize surfaces within 1–4 m of the talker and microphone pickup area. Rear wall absorption is often critical in shallow rooms where reflections return quickly.

Common pitfalls: Over-treating only the front wall and leaving the rear wall reflective, causing strong slapback into microphones. Another is installing panels too high above seated ear level; keep key absorption between 0.9–2.1 m height where speech energy travels.

Troubleshooting: If RT drops but intelligibility still feels poor, you may have strong early reflections (table, display glass, side wall near talker). Add a ceiling cloud above the table or treat the first reflection points more aggressively.
7) Verify speech levels and coverage with the installed system

Action: Set speech reinforcement to consistent, comfortable levels and confirm SNR.

What to do and why: Even a quiet, well-treated room fails if listeners can’t hear remote talkers or if local reinforcement is uneven. The goal is not “loud,” it’s “clear with margin above noise.”

Procedure:
- Play a calibrated speech track (or pink noise shaped to speech) through the conferencing loudspeaker path.
- Measure 67 dBA slow at seated positions (±3 dB across the table is a solid target).
- Check that speech is at least 15 dB above measured background noise in the 1–4 kHz range (use 1/3-octave RTA difference as a practical field metric).
Common pitfalls: Chasing “warmth” with EQ and accidentally boosting 125–250 Hz, which reduces clarity and headroom. For speech, avoid broad boosts below 250 Hz; use high-pass filters as needed.

Troubleshooting: If you hear feedback or far-end echo, don’t immediately lower levels. First confirm mic/speaker geometry, AEC reference routing, and that loudspeaker coverage isn’t blasting directly into table mics.
8) Re-measure and document: prove compliance, don’t assume it

Action: Repeat the same measurements and compare to your targets.

What to do and why: The point of standards-based thinking is repeatability. Use the same mic positions, same HVAC mode, and same measurement duration. Your “before vs after” is only meaningful if it’s measured the same way.

Deliverables to capture:
- Background noise: LAeq (60 s) at each position + average, plus 1/3-octave plots.
- RT: averaged 500 Hz, 1 kHz, 2 kHz results across positions.
- Speech level: dBA at each seat with system at nominal.
Common pitfalls: Making multiple changes at once and not knowing which one helped. When possible, fix noise, re-measure, then treat RT, re-measure, then tune levels.

Before and After: What You Should Expect

Here’s a realistic “field” example for a 6 m x 4 m x 2.7 m (20 ft x 13 ft x 9 ft) glass-heavy conference room:

Before: Background noise 42 dBA (diffuser hiss), RT at 1 kHz = 0.95 s, speech reinforcement needed at 74 dBA to feel intelligible, frequent “what did they say?” moments on hybrid calls.
After: Background noise 33–35 dBA (diffuser swap + airflow adjustment + door seals), RT at 1 kHz = 0.55 s (four 2” wall panels + 2” ceiling cloud over table), speech reinforcement comfortable at 67–69 dBA, clearer consonants, less listener fatigue, better AEC stability.

If your after-results don’t look like this, it usually means one of two things: the noise source wasn’t actually addressed (it only got masked), or absorption was added in the wrong places/bandwidth.

Pro Tips to Take It Further

Track octave-band RT, not just a single number: If 250 Hz is high but 1–2 kHz is fine, consider thicker treatment (100 mm) or air gaps, not more thin panels.
Use early reflection control for microphone performance: A ceiling cloud above the table often improves far-end clarity more than adding panels randomly, because it reduces strong early returns into boundary/table mics.
Confirm isolation with a corridor test: Measure background noise in-room with the corridor quiet, then with corridor active. If the in-room level jumps by >5 dB, partition/door leakage is dominating—treat that before adding more absorption.
Don’t EQ your way out of acoustics: EQ can shape tonality, but it can’t remove late reflections. Use EQ after RT and noise are under control.
Build a repeatable “room report” template: Date/time, HVAC mode, mic positions, LAeq values, RT per band, speech SPL at seats, photos of treatment locations. This speeds up future rooms and makes your results defensible.

Wrap-Up

ANSI S12.60-style compliance for conference rooms comes down to disciplined measurement and the right order of operations: get the noise floor down, bring RT into a speech-friendly range, then verify speech level and coverage. Do it a few times and you’ll start predicting outcomes before you even open the measurement software. Practice by measuring one room you know well, making one change at a time, and re-measuring until the numbers—and the listening experience—line up.