Remove Traffic Noise from Video

You spent an hour filming an interview outside a cafe, nailed the conversation, got great footage — and then you listen back and all you hear is cars driving by, trucks rumbling past, and the occasional horn blast drowning out everything your subject said. Traffic noise is the bane of outdoor video production, and it's one of the trickiest noise types to deal with in post.

What Makes Traffic Noise So Difficult

Most noise removal tools are designed for constant, predictable sounds — an AC hum that stays the same, a fan that drones at a fixed pitch. Traffic noise is the opposite. It's variable, broadband, and intermittent. A passing car might generate a whoosh of tire noise centered around 500–2000 Hz. A truck rumbles through with deep low-frequency energy below 200 Hz. A motorcycle rips past with a sharp, aggressive tone between 1–4 kHz. A car horn blasts at 400–500 Hz for two seconds and then disappears.

This variability is what makes traffic noise harder to remove than something like AC hum or fan drone. There's no single spectral profile the tool can learn and subtract. Each passing vehicle creates a unique noise event with different frequency content, duration, and loudness. Traditional noise reduction tools based on static noise profiles simply can't keep up.

How AI Tackles Variable Noise

Our AI model approaches the problem differently. Instead of trying to characterize "traffic noise" as a single noise profile, it's trained to recognize and preserve human speech patterns while suppressing everything else. The model has learned what speech looks like in a spectrogram — the formant structures, pitch contours, natural pauses, and harmonics — and it extracts those elements while attenuating the non-speech content around them.

This speech-first approach is why the AI can remove traffic noise from video even when the noise is constantly changing. It doesn't need to know what traffic sounds like; it needs to know what voices sound like. Everything that isn't voice gets suppressed. A car whooshing by? Suppressed. A truck engine? Suppressed. Honking? Suppressed. Your subject speaking? Preserved.

The tradeoff is that ambient sounds you might want to keep (birds chirping, background music from the cafe) also get reduced. This is usually an acceptable tradeoff when the alternative is traffic noise smothering your dialogue. And honestly, for most video content — interviews, vlogs, B-roll narration — clean dialogue matters way more than ambient atmosphere.

Common Outdoor Recording Scenarios

Street Interviews and Vlog Content

Street-level content creators deal with traffic noise constantly. You're interviewing someone on a sidewalk, and every 30 seconds a bus rolls by. Or you're vlogging while walking, and the road right next to you is a constant wash of tire noise and engine rumble. These recordings are prime candidates for AI traffic noise removal because the speech is clearly present but competing with a noisy environment.

Outdoor Events and Festivals

Event coverage near roads — outdoor markets, street festivals, sporting events — often has a persistent traffic noise floor. Even when the event itself is the focus, nearby road traffic adds an unwanted layer of noise that reduces the production quality of your footage. When you need to remove traffic noise from video of these events, the AI isolates the voices and event sounds from the road noise backdrop.

Real Estate and Property Videos

Real estate videographers know this pain well. You're showcasing a beautiful property, and the outdoor shots near the street have traffic noise that makes the narration hard to follow. Potential buyers don't want to hear trucks going by while you're describing the charm of the front porch. Clean audio makes the difference between a professional listing and an amateur one.

Documentary and Journalism

Documentary filmmakers and journalists often shoot in uncontrolled environments where traffic is unavoidable. The story is happening where it's happening, and you can't stop traffic for an interview. Getting clean dialogue in post-production is essential, and AI noise removal has become a standard part of the documentary workflow for exactly this reason.

The Frequency Spectrum of Traffic

Understanding what traffic noise actually looks like on a frequency spectrum helps explain both why it's problematic and why AI can still handle it:

• Tire noise (road roar): Broadband, centered around 800–2000 Hz. This is the constant "shhhhh" you hear near busy roads. It overlaps heavily with speech consonants (s, f, th sounds live in this range).
• Engine noise: Low-frequency, typically 50–500 Hz. Heavier vehicles like trucks and buses produce more low-end energy. This overlaps with male vocal fundamentals.
• Horn and siren sounds: Mid-frequency, 400–4000 Hz. These are loud, tonal, and intermittent. They can completely mask speech for their duration.
• Braking and acceleration: Variable squeals and roars that sweep across the frequency spectrum. Unpredictable in timing and character.

The combined effect is noise that covers virtually the entire audible spectrum, which is why simple filtering approaches fail. You'd have to filter out so much of the frequency range that voices would sound thin and robotic. The AI approach of identifying and preserving speech patterns is fundamentally more effective for this type of noise.

Tips for Outdoor Recording Near Roads

• Use a directional (shotgun) microphone pointed at the speaker. Shotgun mics reject sound from the sides and rear, reducing traffic noise that's not directly in front of the mic. The Rode NTG series and Sennheiser MKE 600 are popular choices for outdoor work.
• Get the mic as close to the speaker as possible. A lavalier mic clipped to the speaker's collar puts the mic inches from their mouth, dramatically improving the signal-to-noise ratio. Traffic noise at 15 feet away is much quieter relative to speech at 6 inches.
• Choose your location carefully. If you have any flexibility, position yourself so the speaker is between you and the traffic, using their body as a partial sound barrier. Or find a spot set back from the road, behind a wall or parked vehicles.
• Record a "room tone" clip of just the traffic noise. Even though the AI doesn't require a separate noise sample, having a clean traffic-only reference can be useful if you need to do additional manual cleanup in a DAW.

What Results to Expect

When you remove traffic noise from video with our AI tool, the improvement depends on how severe the original noise is. For recordings where speech is clearly audible but traffic is annoying in the background — the most common scenario — expect a near-complete cleanup. The road noise fades away, voices come through clearly, and the recording sounds like it was shot in a quiet location.

For recordings where traffic periodically drowns out speech entirely (a truck passing right next to the mic, a horn blast directly overhead), the AI may not fully recover speech during those peak moments. It can reduce the traffic noise significantly, but if the original speech signal is completely masked, there's a physical limit to what any tool can recover. The speech information simply wasn't captured by the microphone in those moments.

For the best experience, also consider whether your recording has wind noise on top of the traffic — outdoor recordings often have both. The AI addresses all noise types simultaneously, so you don't need to run multiple passes. Upload your video to our noise removal tool, let the AI process it, and download a clean version. Processing takes 1–3 minutes for most recordings, and the video track stays completely untouched.