Background noise is the enemy of clear audio. A humming air conditioner, a rumbling fan, or a persistent hiss can make an otherwise good recording hard to listen to. The Noise Remover reduces such noise using basic high-pass and low-pass filters that cut the lowest and highest frequencies, where rumble and hiss live. It is an honest, simple tool, and this page explains exactly what it can and cannot do.
The tool carries a notice stating its scope plainly: it applies filters to reduce rumble and hiss, and it cannot separate voices from background noise the way AI noise-removal services do. If you need to remove a dog barking, a conversation in the next room, or music playing in the background, filters alone will not do it. If you need to reduce low rumble and high hiss, this tool is the right starting point.
What Is the Noise Remover?
The Noise Remover is a free browser-based tool that applies high-pass and low-pass filters to an audio file. A high-pass filter lets high frequencies pass and reduces low frequencies below a chosen cutoff, which removes rumble. A low-pass filter lets low frequencies pass and reduces high frequencies above a chosen cutoff, which removes hiss. You choose one or both filters, press the Remove Noise button, and receive a filtered WAV file with a player.
The high-pass choices are no filter, 80 Hz (labeled as removing low rumble), 120 Hz, and 200 Hz (labeled for voice only). The low-pass choices are no filter, 12000 Hz, 8000 Hz (labeled as removing hiss), and 4000 Hz (labeled for voice only). Choosing at least one filter is required; if both are set to no filter, the tool shows an error asking you to choose at least one. The result note states which filters were applied and their cutoff frequencies, so the processing is fully transparent.
The output is a WAV file in 16-bit PCM format, with “-cleaned” added to the file name. Duration, sample rate, and channel count are unchanged, since filtering affects frequency content, not timing. The original file is not modified, and all processing happens in the browser.
How to Use the Noise Remover
Filtering is a balance: cut enough to reduce noise, but not so much that the wanted sound becomes thin or dull. Start gently and increase only if needed.
- Read the scope notice. At the top of the tool, a notice explains that this is basic filtering, not AI noise removal. Keep this in mind so your expectations match what filters can achieve.
- Choose an audio file. Use the Audio file field to select the recording with rumble or hiss. The tool decodes it in the browser.
- Choose a high-pass filter. Use the High-pass filter dropdown. Start with 80 Hz for rumble removal. Use 120 Hz or 200 Hz only for voice recordings where low-frequency content below those points is mostly noise.
- Choose a low-pass filter. Use the Low-pass filter dropdown. Start with 12000 Hz or 8000 Hz for hiss reduction. Use 4000 Hz only for voice-only material, since it will dull music considerably.
- Press Remove Noise. Click the Remove Noise button. If both filters are set to no filter, the tool will ask you to choose at least one. Otherwise, it renders the filtered audio.
- Listen and compare. Play the result and judge two things: is the noise reduced, and does the wanted sound still sound natural? If the voice sounds thin or muffled, try gentler cutoffs from the original file.
After listening, download the filtered WAV file if it sounds better. If the noise is still present, remember that filters only cut frequency ranges; noise in the same range as the voice cannot be removed this way.
Key Features of the Noise Remover
The tool’s strength is its transparency. Every choice is labeled with its effect, and the result reports what was applied.
- Four high-pass options. From no filter to 200 Hz for voice only, the choices cover gentle rumble removal to aggressive low-frequency cutting.
- Four low-pass options. From no filter to 4000 Hz for voice only, the choices cover gentle hiss reduction to strong high-frequency cutting.
- Combined filtering. Both filters can be applied together, narrowing the audio to the frequency range where speech lives, which is effective for voice recordings in noisy rooms.
- Validation for at least one filter. The tool refuses to run with no filter selected, preventing a pointless render that changes nothing.
- Honest scope notice. The notice explains that this is basic filtering, not AI separation, before you use the tool, so there is no false expectation.
- Built-in preview player and clear labeling. Listen before downloading, and read the result note to confirm exactly which filters were applied.
Practical Use Cases for the Noise Remover
Filters work best on steady, frequency-specific noise. These examples show where they help most.
Removing air conditioner and fan rumble from voice
Air conditioners and fans produce low-frequency rumble that sits below most speech. A high-pass filter at 80 Hz or 120 Hz reduces that rumble while leaving the voice largely intact. This is the single most effective use of the tool for home recordings.
Reducing hiss from old recordings and cheap microphones
Tape hiss, microphone self-noise, and electronic hiss live in the high frequencies. A low-pass filter at 8000 Hz trims the top end where hiss is strongest. For voice recordings, this trade-off is usually worthwhile; for music, use 12000 Hz to preserve more sparkle.
Cleaning up a phone recording made outdoors
Wind and traffic noise include strong low-frequency components. A high-pass filter at 120 Hz or 200 Hz can reduce wind rumble in a voice memo recorded outside, making the speech easier to follow, though it will not remove voices, horns, or other distinct sounds.
Preparing voice audio for transcription
Transcription, whether by a person or by software, works better when rumble and hiss are reduced. Filtering a voice recording before transcription can improve accuracy, especially for quiet speakers in noisy rooms.
Tips and Best Practices
Filtering is a trade-off between noise and natural sound. These practices help you find the best balance.
- Start with the gentlest useful filter. Try 80 Hz high-pass or 12000 Hz low-pass first. Move to stronger cutoffs only if noise remains and the voice still sounds natural.
- Use headphones to judge. Rumble and hiss are easier to hear on headphones than on laptop speakers. Judge the filtered result on headphones before deciding.
- Filter from the original each time. If a setting is too strong, return to the original file and try a gentler one. Filtering an already filtered file compounds the loss of frequencies.
- Do not expect miracles with overlapping noise. If the noise is a voice, music, or a barking dog in the same frequency range as your voice, filters cannot separate them. Accept the limitation or seek AI-based separation elsewhere.
- Filter before boosting or normalizing. Remove noise first, then raise the level. Boosting before filtering raises the noise along with the voice and makes it harder to judge the filter.
- Keep the original. Filtering removes frequency content permanently. Keep the unfiltered file so you can try different settings or use a different method later.
Limitations of the Noise Remover
This tool performs basic frequency filtering only. It cannot identify and remove specific sounds, separate a voice from background music, remove echo, or repair clipped and distorted audio. Noise that shares frequencies with the wanted sound will remain, and aggressive filtering will make the wanted sound thin or muffled. The tool offers only the listed cutoff frequencies, with no custom values, no noise profile learning, and no preview of the original and filtered versions side by side.
The output is always a WAV file in 16-bit PCM format. Files the browser cannot decode will produce an error. For recordings where the noise is the main problem and filters are not enough, a dedicated noise-removal service with AI separation, or a re-recording in a quieter environment, will be needed. Within its honest scope, however, the tool is a fast and effective first step for rumble and hiss.
Frequently Asked Questions
Can this tool remove background voices or music?
No. Filters only reduce frequency ranges. Background voices and music overlap with the frequencies of the wanted sound, so they cannot be separated by filtering. The tool’s notice states this clearly.
Which filter removes humming and rumble?
The high-pass filter. Start at 80 Hz, labeled for removing low rumble, and try 120 Hz or 200 Hz for voice-only recordings if rumble remains.
Which filter removes hiss?
The low-pass filter. The 8000 Hz option is labeled for removing hiss. Use 12000 Hz for a gentler effect that preserves more high-frequency detail.
Why does my voice sound thin after filtering?
The cutoff was too aggressive for the recording. A 200 Hz high-pass or 4000 Hz low-pass removes a large part of the natural voice range. Try gentler settings from the original file.
Can I apply both filters at once?
Yes. Choose a high-pass and a low-pass setting together to narrow the audio to the middle frequency range, which works well for voice recordings. At least one filter must be selected.
How Filters Work, and How to Choose Cutoff Frequencies
A filter is best understood as a gate that is open to some frequencies and closed to others, with a gradual transition in between. A high-pass filter with a cutoff of 80 Hz is mostly open above 80 Hz and mostly closed below it, so it removes the deep rumble of traffic, air conditioning, and handling noise while passing the voice. A low-pass filter with a cutoff of 8000 Hz is mostly open below 8000 Hz and mostly closed above it, so it removes hiss while passing most of the voice. The cutoff frequency is the point where the transition happens, and choosing it is a matter of placing the gate between the noise you want to remove and the sound you want to keep.
For voice recordings, the most important frequencies for intelligibility lie roughly between 300 Hz and 3400 Hz, which is why telephone systems historically used that range. Frequencies below 300 Hz add warmth and body to the voice but are not essential for understanding words, and frequencies above 3400 Hz add clarity, brightness, and the distinction between similar consonants. A high-pass cutoff of 80 Hz or 120 Hz removes rumble while preserving nearly all of the voice, including its warmth. A cutoff of 200 Hz begins to thin the voice, removing some of its body, but remains acceptable for speech-only material where clarity matters more than richness. A low-pass cutoff of 12000 Hz removes only the very top end, where hiss lives but little speech content does. A cutoff of 8000 Hz removes more hiss at the cost of some brightness, and 4000 Hz removes a great deal of brightness, leaving a voice that is understandable but dull.
Music has a wider frequency range than speech, and aggressive filtering damages it more noticeably. The low notes of a bass guitar and the kick drum live below 200 Hz, so a high-pass filter at that cutoff will thin the music significantly. The sparkle of cymbals, the air in a vocal recording, and the detail of acoustic instruments live above 8000 Hz, so a low-pass filter there will dull the music. For music, use the gentlest filters that achieve the goal: 80 Hz high-pass for rumble and 12000 Hz low-pass for hiss, and accept that some noise will remain. The alternative, a heavily filtered track, may be quieter but will sound worse than the noisy original.
The best way to choose cutoffs is to experiment systematically. Start with the gentlest settings, listen, and move one step stronger at a time, always returning to the original file for each attempt. Keep notes on which combination sounded best for each type of recording you work with, since the same settings will often suit similar recordings made in the same room with the same equipment. Over time, you will develop a standard recipe for your own setup, such as 80 Hz high-pass and 12000 Hz low-pass for music, or 120 Hz high-pass and 8000 Hz low-pass for voice, and filtering will become a quick, confident step rather than a guessing game.
- Place the gate between noise and signal. The cutoff should sit above the rumble and below the voice for a high-pass filter, and vice versa for a low-pass filter.
- Protect the speech range. Frequencies between 300 Hz and 3400 Hz carry intelligibility, so avoid cutoffs that encroach on that range.
- Be gentler with music. Music uses a wider range than speech, so use the mildest filters that achieve the goal.
- Experiment one step at a time. Move from gentle to strong settings systematically, always starting from the original file.
- Keep a recipe. Note the settings that work for your room and equipment, and reuse them for similar recordings.
Related Tools
These tools prepare audio for filtering, adjust levels afterward, or shape tone more precisely.