What can be measured
This tool is a simple application that utilizes standard browser audio processing interfaces to analyze audio waveform data in real time, captured from a microphone built into the user's device or an external microphone connected to it.
What is achieved here is not the measurement of physical sound pressure in a strict sense, but rather a visualization of relative changes in the sound volume captured by the microphone, presented in the format of decibels to provide reference values for easily grasping the volume level of the current location.
The main purpose of this tool is to allow users to instantly display environmental volume fluctuations as meters and graphs on the screen simply by accessing a web page and granting microphone permission, without needing to purchase dedicated measurement equipment or install special software.
The target of this system's processing is strictly the digitized time-domain waveform data passed from the microphone to the browser.
The audio stream is always calculated entirely locally in the device's memory; the process never involves sending the captured waveform data to external servers or saving it as audio files on the device.
Thanks to this locally-contained design, the component calculates the waveform entirely locally within the browser. However, the numbers visualized by this tool are uncalibrated relative values derived from the calculation processes described later, and they do not equal the absolute sound energy levels in physical space.
Due to the constraints of running on a browser, it specializes in the process of extracting purely the amplitude information of the waveform and updating the screen rendering based on it; no detailed analysis per pitch or correction processing based on specific evaluation criteria is performed.
Therefore, what this tool calculates is "the relative amplitude of the waveform captured by that device's microphone at that precise moment," which can be positioned as a result translated into an easy-to-understand visual indicator.
RMS and fixed offset
The calculation process used to derive the decibel values displayed on the screen from the acquired audio waveform data occurs in two stages: calculating the Root Mean Square (RMS) and adding a fixed offset value to the result.
First, it uses features provided by the browser's web audio programming interface to retrieve an array of time-domain data with a certain length from the audio stream.
Specifically, a data array containing 256 sample points is extracted for each finely divided frame, and each sample is represented as a relative amplitude value centered on a state of silence.
All 256 of these amplitude data points are first normalized to a range of minus 1 to plus 1 centered on zero. Then, each value is squared to invert negative values to positive while determining the energy amount of the amplitude, and the sum of all these values is calculated.
The resulting sum is divided by 256, the total number of samples, to derive the average energy, and finally, the square root is calculated to establish the RMS value indicating the average amplitude size of the entire waveform.
Since this RMS value is a very small decimal close to zero, a logarithmic calculation is performed to convert it into a decibel numerical format that is more intuitively understandable to humans.
The mathematical formula applies a common logarithm to the calculated RMS value and multiplies it by 20. However, because the calculation result at this point often becomes a relative negative value based on the electronic characteristics of the microphone input, this tool adopts a design that simply adds a fixed offset value of "90" internally within the program.
This fixed value of 90 is an expedient benchmark set so that the meter needle moves within an appropriate range suited to the sense of volume in typical living environments, and it is not based on a reference sound pressure.
Furthermore, to prevent the calculation results from reaching extreme values and breaking the screen display, the final number is forcibly rounded to fit within a minimum of 0 and a maximum of 120.
Thus, you must understand that the decibel display of this tool is a virtual number based on a combination of RMS calculation and a fixed offset.
Graphs and aggregation
In this tool, instead of simply displaying the decibel value calculated in the previous chapter as text, multiple visual user interfaces and aggregation logic are implemented so that users can intuitively grasp trends in volume changes.
The main display is a semicircular gauge modeled after an analog meter and a digital readout of the current value. The angle of the meter needle updates in real time according to the calculated value from 0 to 120, allowing you to visually capture the current sense of sound volume.
In addition, to visualize volume fluctuations over time, a historical trend graph of the most recent 150 frames is drawn using a canvas element.
The process of capturing audio from the microphone and calculating values is repeatedly executed at short intervals synchronized with the browser's screen update cycle, and each newly calculated value is appended to the historical data array.
This history array is designed to hold a maximum of 150 data points, and operates on a mechanism where the oldest data is removed from the front of the array every time new data is added.
By drawing lines connecting these 150 frames of data from left to right on the canvas, an animated graph is realized where volume changes flow like waves.
Furthermore, separately from the current value trend graph, two statistical aggregation metrics—the maximum value and the arithmetic mean—are calculated in real time and displayed on the screen to show the overall trend from the start of measurement to the present.
The maximum value is compared to past records each time the measurement loop runs, and by only updating when a higher number is detected, it continually records the largest volume peak during the observation period.
The arithmetic mean, on the other hand, is calculated by simply adding all the decibel values from all frames computed since measurement started, and dividing that by the total number of executed frames using a cumulative average logic.
These graph displays and aggregation features are valid only during an ongoing session from when the start button is pressed until the stop button is pressed; if you stop and restart the measurement, the history data, maximum value, and average value are all reset to their initial state, and aggregation begins anew from zero.
Device differences
A crucial factor in understanding the characteristics of the numbers displayed by this tool is the existence of significant "device differences" caused by the hardware and software environments used by the user.
As mentioned above, this tool utilizes the audio waveform data input from the device's microphone via the browser directly as the basis for calculation.
However, the microphone modules built into smartphones, tablets, and laptops worldwide each possess completely different sensitivities, response characteristics, and directivities.
At a hardware level, one device's microphone may pick up minute sounds very sensitively, while another device's microphone may physically struggle to react unless the sound is loud.
Additionally, modern operating systems, browsers, and many audio-related drivers apply automatic correction processing to audio input to enhance user convenience.
Typical examples include automatic gain control and noise cancellation features. These dynamically and irreversibly alter the amplitude of the audio waveform, such as automatically amplifying the volume if the input sound is too quiet, or conversely reducing the volume to prevent audio clipping if it is too loud.
What this tool can access is the waveform data after such preliminary corrections by the operating system or hardware have already been applied.
Therefore, even if you run this tool simultaneously on two different devices against the exact same sound source in the exact same room, they will acquire different waveform amplitudes, routinely causing the decibel reference values displayed on the screen to differ significantly.
Furthermore, internal processing timing also varies depending on the type and version of the web browser, which is another factor that creates numerical variance.
Due to these combined reasons, the numbers calculated by this tool are relative results that only hold true under the highly specific, unique environment of the "combination of the device and browser currently in use," and you must always keep in mind that they are not suitable for comparing with numbers from other devices or diverting as absolute standards.
How to use
The procedures for using this tool are designed to be extremely simple and require no complex settings or preparation. When you access the page, a button saying "Start noise measurement" is located in the center of the screen.
Clicking or tapping this button serves as the starting point for all processing. Once the start button is pressed, the browser immediately displays a security confirmation prompt requesting microphone access permission from the user.
When the user allows microphone usage here, an initialization process runs within the tool, and it begins acquiring the audio stream from the microphone.
In the unlikely event that microphone access is denied or no microphone is connected to the device, an error message appears on the screen, and the process is safely aborted.
Once the microphone stream is successfully obtained, the waveform data analysis loop starts spinning immediately, and the on-screen indicators begin to move.
The main meter displays large digital numbers indicating the current reference value, and its needle swings smoothly left and right in response to volume changes.
Also, depending on the size of the displayed value, the environmental guideline text below the meter switches dynamically. For example, if the calculated value is under 50, green text such as "Quiet" appears, and as the number increases, it changes to emphasized orange or red text like "Normal" or "Noisy," conveying a guideline for the current acoustic situation along with a visual color code.
On the right side or bottom of the screen, the maximum and average values since the start of measurement are updated and displayed continuously, and the 150-frame waveform graph showing recent fluctuations continues to be drawn with animation.
Users can leave the tool running in this state and observe environmental volume changes. When you wish to end the measurement, press the red "Stop measurement" button.
When the stop button is pressed, the internal audio processing loop is immediately canceled, and the connection stream with the microphone is severed and released.
Consequently, the meter needle returns to its initial position, and the on-screen text reverts to the standby notation. If you want to measure again, pressing the start button once more allows you to begin a new session from a state where all history and aggregated data have been reset.
Situations requiring calibrated equipment
As detailed in previous chapters, this tool is an application that simply visualizes relative fluctuations within the device by adding a fixed offset to the RMS value of the waveform acquired from the microphone.
Therefore, the numbers obtained from this tool cannot be used as official records or strict criteria for judgment, and there are situations where you must absolutely use dedicated measurement equipment specifically designed to achieve particular objectives.
The most typical examples are assessments of health and safety in workplace environments and judgments regarding standards to protect workers' health.
For evaluations to determine whether noise levels exceed certain lines in factories or construction sites, and to impose obligations for appropriate soundproofing measures or the wearing of protective gear, dedicated measuring instruments that meet the specifications established by official guidelines are indispensable.
The same applies when recording situations as objective evidence in neighborhood disputes or environmental issues accompanying construction work, in order to use them as a basis for consulting administrative agencies or formal procedures.
In situations that involve such proof, reproducibility and objectivity—meaning "the same result is obtained no matter who measures it, with whichever device"—are absolute requirements.
Dedicated measuring equipment undergoes strict calibration at the manufacturing stage using known reference sound sources, and accurately implements complex filtering circuits aligned with international standards that consider human auditory characteristics.
On the other hand, this tool does not implement such calibration features based on reference sound pressure or complex correction logic to comply with specific evaluation standards, and is limited to simple rounding calculations of time-domain waveform amplitude.
Therefore, it is inappropriate to use this tool in any situation that requires objective judgment based on absolute physical quantities, such as verifying whether regulations are being followed or comparing the accurate decibel notation of volume emitted from a specific device with catalog specifications.
This tool should only be utilized as an everyday reference indicator to casually enjoy relative changes, such as "it seems louder now than before," entirely within the scope of personal interest.