Noise reduction: removing unwanted sound, such as hiss

Noise reduction is removing unwanted sound, such as hiss, hum or background chatter, from a recording. The engine's noise removal is switched off.

Noise reduction is removing unwanted sound, such as hiss, hum or background chatter, from a recording. It always involves a trade-off: the more aggressively sound is treated as noise, the more the wanted signal can be distorted. For speech recognition that means cleaning too hard can damage the words themselves.

Also searched as "noise removal", "denoise" and "background noise removal".

What the sources say about noise reduction

  • Wikipedia, Noise reduction. Noise reduction is the process of removing noise from a signal; for audio it covers hiss and background noise; it may distort the signal to some degree; spectral methods filter content below a threshold. Read on 22 October 2026.

Noise reduction in Autrans

The engine's noise removal is switched off. What it does do is set the overall volume to a standard level and filter low rumble such as air conditioners, handling noise and phone-line hum. It deliberately does not remove background noise from the speech, because that can damage the words along with the noise. Speech it cannot make out is marked inaudible rather than guessed. Autrans' public releases page says background noise is left in.

A worked example of noise reduction

A recording made in a market: Autrans levels the volume and removes the low rumble, but the chatter of nearby voices stays in the audio. Where a nearby voice drowns the speaker, the transcript shows that stretch as inaudible rather than inventing words.

Terms next to noise reduction

  • Voice activity detection: Voice activity detection decides which parts of an audio signal contain human speech and which are silence or noise.
  • Sample rate: Sample rate is how many times per second a digital recording measures the sound wave, written in hertz or kilohertz.
  • Word error rate: Word error rate is the standard score for a speech recognition transcript.
  • Automatic speech recognition: Automatic speech recognition (ASR) is the field and the technology that converts spoken language into text a computer can use.

Questions people ask

What does noise reduction mean?

Noise reduction is removing unwanted sound, such as hiss, hum or background chatter, from a recording. It always involves a trade-off: the more aggressively sound is treated as noise, the more the wanted signal can be distorted.

How does noise reduction show up in Autrans?

The engine's noise removal is switched off. What it does do is set the overall volume to a standard level and filter low rumble such as air conditioners, handling noise and phone-line hum.

Where does this definition of noise reduction come from?

It rests on Wikipedia, Noise reduction, read on 22 October 2026. What each source supports is listed above, next to its link.

Can you give an example of noise reduction?

A recording made in a market: Autrans levels the volume and removes the low rumble, but the chatter of nearby voices stays in the audio. Where a nearby voice drowns the speaker, the transcript shows that stretch as inaudible rather than inventing words.

Related

Start transcribing free

Send Autrans into your next meeting, use it in the room, or upload a recording. Your first 30 minutes are free, given once, and no credit card is needed to start.

Last checked 22 October 2026. How we measure accuracy. Spotted a mistake? Write to hello@autrans.online.