Define the Range
Enter start and end frequencies or choose a full-range, bass, midrange, treble, or subwoofer preset.
Ready — sound is off
Browser audio signal toolkit
Move a generated tone continuously between two frequencies with a defined curve, duration, direction, waveform, level, and stereo channel.
A frequency sweep generator changes an oscillator from one frequency to another over a selected time. Rather than checking isolated tones manually, you hear a continuous path through the range. Set the endpoints, duration, linear or logarithmic curve, direction, waveform, channel, and level; then press Start Sweep. Web Audio automation schedules the changing frequency, while the progress display reports the calculated current point.
This frequency sweep generator includes broad presets for 20 Hz–20 kHz, bass, midrange, treble, and a 20–200 Hz subwoofer range. Looping is off by default, and Stop cancels scheduled automation and fades the signal. The result is useful for subjective observations and demonstrations. It does not measure frequency response because the page has no calibrated microphone or analyzer connected to the acoustic output.
Three simple steps
Use this frequency sweep generator deliberately: set the signal, begin quietly, and interpret only what the workflow can actually show.
Enter start and end frequencies or choose a full-range, bass, midrange, treble, or subwoofer preset.
Set duration, logarithmic or linear behavior, direction, waveform, channel, and a conservative level.
Run the sweep, follow current frequency and progress, and stop at discomfort, rattling, or unexpected strain.
A linear sweep adds the same number of hertz in each equal unit of time. During a 20 Hz to 20 kHz linear sweep, much of the duration is therefore spent at numerically high frequencies. A logarithmic sweep multiplies frequency by a consistent ratio over time, allocating similar time to each octave. It often sounds more evenly paced when the goal is to move across musical or perceptual ranges.
Neither curve is universally better. Linear motion is easy to relate to absolute hertz per second, while logarithmic motion is convenient for octave-oriented listening. The frequency sweep generator schedules a linear AudioParam ramp for the first case and an exponential-in-frequency ramp for the second. Logarithmic endpoints must be above zero, which the tool enforces through positive frequency limits.
The full 20 Hz–20 kHz preset is an audible-focused reference, not a claim that every listener or system covers that entire span. The 20–250 Hz bass preset concentrates on low-frequency behavior. A 250 Hz–4 kHz midrange sweep covers much speech and instrumental information, while 4–20 kHz explores treble. The subwoofer preset narrows the run to 20–200 Hz.
Shorter custom ranges make observations easier to repeat. If a rattle appears near one point, stop, lower the level, and use the exact frequency generator to check nearby steady values. Avoid repeatedly sweeping a broad range at high volume. Frequencies near the limits may be inaudible even though the digital automation is still running.
A cautious sweep can reveal obvious changes such as a channel dropout, strong rattle, sudden gap, or region that sounds unusually prominent in one listening position. Begin with a moderate range at low level and keep processing settings consistent. Listen once, write down the approximate frequency shown on screen, and repeat only if necessary. Avoid treating a smooth subjective impression as a specification.
Headphones remove room reflections but retain their own response, fit, seal, electronics, and listener variation. Speakers add placement and room interactions. A frequency sweep generator supplies the changing source but cannot separate those influences. The guided speaker test is a better first choice when you mainly need left/right routing and a few broad checkpoints.
Rattling can come from a speaker enclosure, loose object, furniture, vent, or another item in the room. A dropout can result from the source, processing, connection, amplifier, transducer, room cancellation, or hearing. A resonance may make a narrow region sound louder at one position. The live frequency readout helps locate where a symptom appeared, but it does not identify the cause.
When something unexpected occurs, stop the frequency sweep generator rather than increasing level. Inspect the environment and repeat at a lower setting. Change one factor at a time, such as listening position or a loose nearby object. Equipment that sounds distressed should not be driven further. Qualified service or proper measurement may be appropriate for persistent problems.
Frequency response is a quantitative relationship between input and output across frequency, normally measured with a defined reference, calibrated capture system, known geometry, and appropriate analysis. This page generates only the input side. It does not record sound, know the acoustic level, compensate for a microphone, or remove room effects. The progress graph is timing information, not a response curve.
Human loudness sensitivity also changes with frequency and level, so equal digital amplitude does not mean equal perceived loudness. A frequency sweep generator can support preliminary listening and help reproduce a symptom, but a measurement microphone and suitable software are required for defensible response data. Do not turn subjective variation into a numerical score.
Audio hobbyists use sweeps to explore a playback system or locate an obvious buzz. Teachers use them to demonstrate continuous frequency change and compare linear with logarithmic motion. Musicians and studio users may check a monitoring path before more formal work. Developers can verify that Web Audio automation, channel selection, and output routing remain active across changing values.
Engineers also use sweeps in professional measurement, but those procedures define excitation, level, capture, timing, processing, and uncertainty. This browser frequency sweep generator should not be confused with a complete measurement system. It offers transparent parameters and deterministic scheduling for everyday checks, education, and investigation.
Start below your normal listening level and keep the first sweep short. A wide sweep can pass through regions where the playback system or human ear is more sensitive, causing an abrupt change in perceived loudness even though digital gain stays constant. Stop immediately if any part feels uncomfortable. Looping remains off unless you explicitly enable it.
Low frequencies can demand substantial driver movement, while high frequencies may not seem loud enough to discourage unsafe level increases. Never raise volume sharply to chase an inaudible tone. The frequency sweep generator is not a hearing diagnostic device, and the output percentage is not dB SPL. You control the downstream equipment, exposure duration, and environment.
An upward sweep runs from the lower endpoint to the higher endpoint; a downward sweep reverses them. Direction can change how a short-lived symptom is noticed, but it does not change the selected range. Longer durations make it easier to read the current frequency and locate an event, while short sweeps are convenient for a broad confirmation.
For comparison, keep endpoints, curve, duration, waveform, level, channel, device volume, and position unchanged. The browser can repeat its scheduled parameters, but background throttling and the wider playback system still matter. Stop starts the next run from the beginning rather than pretending to provide a perfectly resumable pause.
Clear answers
Use logarithmic when you want comparable time per octave and linear when equal hertz-per-second movement fits the task. Both are correctly scheduled but emphasize the range differently.
Because frequency changes by a consistent ratio rather than a consistent number of hertz. Each octave therefore receives a similar share of the duration.
Not by itself. Quantitative response needs calibrated capture hardware, known placement, suitable analysis, and a defined method.
Speaker response, placement, room modes, reflections, processing, and human sensitivity can all contribute. The sweep does not isolate those variables.
Scheduled frequency automation is canceled, gain ramps down briefly to reduce clicks, and oscillator and progress resources are cleaned up.