Home · Personal · Writing & Creative · Music & audio
As of 13 August 2026, AI can remove background noise from your recording.
Most people should do this themselves; the prompt is on this page.
Can you do it?
5 minutesto a draft.
15 minutesto something you’d act on.
Cost, all in£0
Skill neededchat-fluent
Who has to check ityou
What the alternative costsA purpose-built editor such as Descript is the relevant alternative; the supplied tool data gives no price.
If this goes wrong, the recording develops watery or metallic artefacts, but you can return to the untouched original and try a lighter setting.
What to actually do
Do it yourself
The route this page recommends
A chat interface, chat-fluent skill, and roughly 15 minutes until you can act on the result.
How to actually do it
- Make a copy of the original recording and rename the copy as the working file.
- Open Descript or another audio editor with AI cleanup features, then import the working file.
- Paste the prompt and upload the recording, stating whether the important material is speech, music or both.
- Ask the tool to process a short representative section first, then compare it with the untouched section for noise reduction and voice quality.
- Apply the lightest setting that reduces the unwanted noise without creating metallic, watery or pumping sounds.
- Listen through the complete processed recording with headphones, check words and music against the original, and export a new file while keeping the original unchanged.
Prompt
Use the attached recording to reduce background noise while preserving the speaker's words, natural voice, music, timing and room tone. First identify the likely noise, such as hum, hiss, fan noise or traffic. Apply the lightest effective reduction and do not remove pauses, breaths or words unless they are clearly noise. If you cannot directly edit and return the audio file, give me exact step-by-step settings for a suitable audio editor instead. Return the processed file if possible, plus a short list of what was changed and any parts that may need checking. Do not invent what you cannot hear.
Open it prefilled in ChatGPT or Claude, or copy it into Gemini, which takes no prefill link.
Use a tool built for this
Second choiceHand it to a person
The distant thirdA person who owns the outcome does this end to end, worth it when the failure is dear.
What it gets wrong
- AI cannot know how much room tone or background sound you want to keep for a natural result.
- It can mistake quiet speech, breaths, music or reverb for noise and damage them.
- It cannot reliably repair noise that overlaps the same frequencies as the voice or instrument.
- It cannot hear whether the final result suits your intended audience as well as you can.
- A severe or changing recording problem may need manual editing by an audio specialist.
Even on a YES, the friction has a name: taste and context depth.
How we scored this
Five axes, each scored nought to two by hand: ten means AI carries the task cleanly, and the thresholds that turn a total into YES, PARTLY or NO are published in the methodology. Each axis name links to its definition.
| Axis | Score (0–2) |
|---|---|
| Output | 2 |
| Inputs | 2 |
| Verification | 2 |
| Liability | 2 |
| Effort delta | 2 |
| Total | 10 / 10 |
The methodology and its thresholds are published in full.
FAQ
- Can AI remove background noise without ruining the recording?
- Often, yes, when the noise is steady and clearly separate from the speech or music. Listen for metallic, watery or pumping artefacts and keep the original so you can undo the change.
- What background noise can AI remove from audio?
- It is most useful for steady sounds such as hiss, hum, fan noise and some room noise. Traffic, overlapping conversations and noise that changes at the same time as the speaker are harder to remove cleanly.
- Can ChatGPT clean up an audio recording?
- A chatbot can help you choose a cleanup approach and settings, but it may not be able to return a properly processed audio file in your chat. Use a dedicated audio editor such as Descript when you need to import, process and export the recording.
- How do I know if AI noise reduction has damaged my audio?
- Compare the processed file with the original using headphones, checking quiet words, consonants, breaths, music and pauses. Stop and use a lighter setting if the voice sounds metallic, underwater or unnaturally cut off.
Nearby answers
Assessed by gpt-5.6-luna (gpt-5.6-luna) on 2026-08-13, second-checked by an independent model. Wrong somewhere? Email [email protected] and it gets re-checked.
The newsletter
AI news, new answers and product picks, straight to your inbox.