Eliminate the drone and whir of computer fans, ceiling fans, and HVAC fans from your recordings with AI.
Remove Fan Noise NowThat whirring sound. You know the one. Your laptop fan spins up during a recording session, or the ceiling fan is on because it's 90 degrees, or the HVAC system in your office has a fan that won't quit. You don't think much of it in the moment — you're focused on what you're saying, on the content, on the interview. And then you play it back and the fan noise is right there, constant, distracting, and making your recording sound like it was made inside a wind tunnel.
Fan noise is probably the most universal recording problem for home studios, home offices, and anyone recording outside a professional booth. Think about how many fans are running in a typical recording space: your computer has at least one (often two or three — CPU, GPU, and case fans), your office might have a ceiling fan, and the HVAC system has a blower fan pushing air through the ducts. In summer, you might also have a desk fan or portable fan running. That's a lot of spinning blades pushing air around within a few feet of your microphone.
Computer fans are especially problematic for podcasters and screencast creators because the computer is right there on the desk, often inches from the microphone. When you're recording a video tutorial, rendering anything, or running CPU-intensive software, the fans ramp up and the noise increases. A typical laptop fan at full speed generates 35–45 dB at close range — that's enough to be clearly audible in any recording.
Desktop computer fans can be even louder, especially gaming PCs with multiple high-RPM fans. Some configurations produce 40–55 dB of fan noise, which is comparable to a normal conversation volume. If your microphone is sitting on the same desk, it's capturing a significant amount of that noise.
Understanding what fan noise sounds like to a machine helps explain why AI is so effective at removing it. Fans produce a tonal noise — a dominant fundamental frequency determined by the rotation speed, plus a series of harmonics above it. A typical computer fan spinning at 1500 RPM produces a fundamental tone around 25 Hz (inaudible to most people) with audible harmonics at 50, 75, 100, 150 Hz and above. The most annoying audible components usually sit between 100–500 Hz.
On top of this tonal component, there's a broadband whooshing from the air being moved. This sounds like a low-level "shhhh" and is concentrated in the 200–2000 Hz range. Ceiling fans and HVAC blower fans have more of this broadband character because they move much more air than small computer fans.
The important thing is: fan noise is consistent. It stays at roughly the same pitch and volume for extended periods. Even when a computer fan ramps up, it does so gradually over seconds, not instantaneously. This consistency is exactly what makes fan noise an ideal target for AI removal. The model can identify the fan's spectral signature and subtract it continuously from the recording.
When you upload a recording to remove fan noise from recording, the AI does several things. First, it analyzes the spectral content of the audio to identify the steady-state components — the frequencies that stay constant while speech comes and goes. Fan noise shows up as bright horizontal lines on a spectrogram (constant frequencies over time), while speech shows up as complex, changing patterns. The AI learns to recognize the fixed components as noise and the varying components as speech.
Then it applies spectral subtraction, removing the fan noise energy from each frequency band while preserving the speech energy. Because the fan noise is tonal (concentrated at specific frequencies), the AI can be very precise — it doesn't need to apply broad cuts that might damage vocal quality. It targets just the frequencies where the fan lives.
For HVAC and ceiling fan noise, which has more broadband content, the approach is slightly different. The AI uses the quiet moments between speech to build a complete noise profile, then applies adaptive reduction across the full spectrum. The result is the same: the fan sound disappears, voices stay clean and natural.
This is the classic setup and the classic problem. Your laptop is running your recording software, maybe a browser with research notes, and the fan kicks in. If you're using a USB mic connected to that laptop, the mic is probably 12–18 inches from the fan exhaust. The AI can remove fan noise from recording sessions like this effectively because the noise is consistent and the speech is typically much louder than the fan.
Even purpose-built home studios in spare bedrooms often have computers in the room (for the DAW) and HVAC that can't be turned off in shared houses. Vocalists and voice actors need pristine audio, and fan noise undermines the clean, intimate sound they're going for.
Zoom, Teams, and other video call recordings capture fan noise from every participant. When you have four people on a call, each with their own laptop fans running, the combined noise floor is significant. If you're saving the recording for later distribution, cleaning up that fan noise makes a noticeable difference in professionalism.
Gamers and streamers run their GPUs hard, and GPU fans can hit 50+ dB under load. The gaming PC is usually on or under the desk, right next to the streaming microphone. A stream or video with constant fan drone in the background feels amateurish. Many streamers now use AI noise removal either in real-time or in post as a standard part of their workflow.
Fan noise shares some characteristics with other common recording problems:
Even with precautions, some fan noise usually makes it into the recording. That's where our AI tool comes in. Upload your video, let it remove fan noise from recording, and download a clean version in minutes. The video stays untouched — only the audio gets processed. For recordings that also have other issues like keyboard clicks or room echo, the AI handles everything in one pass.
Less CPU and GPU load means lower fan speeds. Quit browser tabs, background apps, and anything not essential to your recording. This alone can drop fan noise significantly.
Because fan noise has a consistent pitch and harmonics, the AI can target it with precision without affecting vocal frequencies. Expect excellent results for typical computer and ceiling fan noise.
Every doubling of distance between fan and microphone reduces fan noise by about 6 dB. Put the computer on the floor or use a longer cable to create separation.
If fan noise is a recurring problem, switch from a condenser to a dynamic microphone. Dynamic mics are less sensitive to ambient sounds and naturally capture less fan drone.
Eliminate the drone and whir of computer fans, ceiling fans, and HVAC fans from your recordings with AI.
Remove Fan Noise Now