How Loud Is a Normal Conversation?

Normal one-to-one conversation commonly falls around 55 to 75 dBA at about 1 metre, or roughly arm’s length. About 60 dBA is often used as a simple reference, but speech does not stay at one level. Strong syllables, animated speech, several talkers, background noise, and a shorter mouth-to-microphone distance can all produce higher readings. NIDCD NIDCD infographic NIOSH and CPWR

Sound levels are approximate. The same conversation can differ substantially with distance, direction, room acoustics, vocal effort, and measurement method. An A-weighted average, a Fast maximum, and a short peak are not interchangeable.

How Loud Are Common Sounds? A Context-Aware Decibel Chart

Why conversation levels vary continuously

Conversation is a variable and intermittent sound. A speaker produces brief syllables separated by quieter consonants, pauses, breaths, and periods when another person is talking. A live meter may move through a wide range even though the conversation sounds steady to the participants.

For a representative speech level, an A-weighted equivalent average such as LAeq over a defined period is usually more useful than the highest number that appears on screen. LAFmax can describe the strongest Fast-time-weighted moment, but it should not be presented as the normal level of the whole conversation.

The main variables are:

vocal effort and the speaker’s natural voice
the distance and angle between the mouth and microphone
one speaker versus several people talking at once
background noise, which often causes people to raise their voices
room size, hard surfaces, soft furnishings, and reverberation
whether pauses are included in the averaging period
frequency weighting, response time, and the app’s calculation method

Public reference values differ for these reasons. NIOSH commonly uses about 60 dBA as an educational comparison. NIDCD gives ranges around 60 to 70 dBA in one source and 65 to 80 dB at arm’s length in another. Those values can coexist without describing exactly the same speaker, room, distance, or time metric. NIDCD NIDCD infographic NIOSH and CPWR

Whispering, normal speech, loud speech, and shouting are different

A whisper is not simply normal conversation turned down by a fixed number of decibels. Whispering changes how the voice is produced, and a very soft whisper at conversational distance may be difficult to separate from the room’s background noise. Normal speech uses a comfortable vocal effort. Loud speech and shouting use more effort and are often produced in response to a noisy environment or greater distance.

A chart that labels all speech as 60 dB hides these differences. It also hides the measurement position. A phone placed 15 centimetres from a speaker’s mouth can show a much higher value than the same phone placed at the other person’s ear position one metre away.

Distance has a physical effect on sound pressure. In an ideal free field, doubling the distance from a point-like source reduces the level by about 6 dB. Human speech in a room rarely follows that rule exactly because the head and body are directional, the source is not a perfect point, and walls and ceilings return reflected sound. The rule is still useful for understanding why mouth-to-phone distance must be recorded. OSHA manual

What Is Sound Pressure Level? SPL Explained

Is normal conversation harmful to hearing?

Ordinary one-to-one conversation at a normal distance is generally not treated as a hearing hazard. Its typical range is well below the levels used in occupational noise-dose guidance. That does not mean every place where people are talking is quiet. A crowded restaurant, party, classroom, or event can be much louder because speech is combined with music, movement, ventilation, and many simultaneous voices.

OSHA and NIOSH use the need to shout to communicate with someone about one metre away as a practical warning sign that the surrounding workplace may be near or above 85 dBA. It is only a screening rule. It does not replace a measurement, and it refers to the surrounding environment rather than the sound level of one person’s ordinary voice. OSHA

Conversation can also interrupt concentration or privacy at levels far below those associated with occupational hearing risk. Annoyance and speech interference answer a different question from hearing exposure.

What Is a Safe Decibel Level?

How to measure a conversation in a useful way

First decide whose listening position the result should represent. For a normal face-to-face conversation, place the phone at about the listener’s ear height and approximately one metre from the main speaker. Do not put it directly beside the speaker’s mouth unless close-range speech is the specific question.

Use a small stand where possible. Keep the phone orientation fixed, leave its microphones uncovered, and avoid holding it against clothing or a tabletop. Record the room, distance, number of speakers, and whether doors or windows were open.

Measure a representative section rather than one word. A 30 to 60 second LAeq can be useful for a steady comparison, while a longer period may be needed if several people are taking turns or the room changes. Use the same duration for repeated measurements. Record LAFmax separately if the strongest syllables are relevant.

For comparisons, keep these details unchanged:

phone and app
calibration state
frequency weighting and response time
microphone position and orientation
speaker distance and room
measurement duration

A measurement of a quiet conversation can be limited by the phone’s own microphone noise and by background sound. A loud syllable can also be affected by automatic gain control or other processing. Android phones do not share one microphone design or one unprocessed audio path, so a reading should be treated as an estimate unless the complete setup has been validated. NIOSH app

Are Decibel Meter Apps Accurate? How to Measure Decibels With an Android Phone

Why phone position changes the result so much

Speech is directional, especially at higher frequencies. A microphone facing the speaker may receive a different spectrum and level from one turned sideways or partly blocked by the phone case. The listener’s body, a table, and nearby walls can also reflect or obstruct sound.

Two readings should not be compared when one phone was beside the speaker and the other was in the listener’s pocket. A pocket changes both distance and obstruction. The second result may be lower because fabric covers the microphone, or it may contain rubbing noise that has nothing to do with the conversation.

Displayed precision should also be kept in perspective. A result of 63.4 dBA does not show that the phone is accurate to 0.1 dB. For an everyday comparison, a documented range and a repeatable position are more useful than a highly precise-looking single number.

Compare speech with other everyday sounds

Use the Common Sounds Explorer to compare normal conversation with whispering, household appliances, traffic, and other familiar sources. Keep the stated distance and metric visible, because a 60 dBA speech average at one metre is not directly comparable with a product sound-power label or a short maximum measured beside another source.

Common Sounds Explorer

Sources

NIDCD, Noise-Induced Hearing Loss. https://www.nidcd.nih.gov/health/noise-induced-hearing-loss
NIDCD, Listen Up! Protect Your Hearing. https://www.nidcd.nih.gov/health/listen-infographic
CDC/NIOSH and CPWR, Construction Noise and Hearing Loss Prevention. https://stacks.cdc.gov/view/cdc/78630
OSHA, Occupational Noise Exposure. https://www.osha.gov/noise
OSHA Technical Manual, Section III, Chapter 5. https://www.osha.gov/otm/section-3-health-hazards/chapter-5
CDC/NIOSH, NIOSH Sound Level Meter App. https://www.cdc.gov/niosh/noise/about/app.html