Home · Personal · Writing & Creative · Music & audio
As of 13 August 2026, AI can improve the sound of an old recording.
This still needs a person who signs their name to it.
Can you do it?
15 minutesto a draft.
30 minutesto something you’d act on.
Cost, all in£0
Skill neededchat-fluent
Who has to check ityou
What the alternative costsThe supplied tool list does not give a price for a human audio-restoration service.
If this goes wrong: the recording sounds watery, harsh or over-processed, but you can return to the original and try a lighter cleanup.
What to actually do
Hand it to a person
The route this page recommends
A person who owns the outcome does this end to end, worth it when the failure is dear.
Do it yourself
Second choiceA chat interface, chat-fluent skill, and roughly 30 minutes until you can act on the result.
How to actually do it
- Make a duplicate of the original recording and keep the untouched file in a separate folder.
- Open Descript or another audio editor, import the duplicate, and listen from start to finish with headphones while noting the main problems.
- Paste the prompt into the tool, upload the duplicate, and name the specific problems you heard, such as hiss, hum, clicks, muffled speech or uneven volume.
- Apply the suggested cleanup to a short representative section first, then compare it with the same section of the original before processing the whole recording.
- Listen to the full processed file for watery voices, missing consonants, metallic music, pumping volume or clipped peaks, and reduce the processing if any appear.
- Export the improved recording as a new file, then play both files on headphones and ordinary speakers before keeping the processed version.
Prompt
I want to improve the sound of an old recording without changing its content, timing or character. I will upload the audio file. First identify the problems you can hear, such as steady hiss, mains hum, clicks, crackle, muffled speech, uneven volume, clipping or room noise. Apply the lightest useful cleanup, preserve voices and music, and do not invent, replace or regenerate any words or musical notes. If you can edit and export the audio, provide the cleaned file and state what processing you applied. If you cannot edit or export audio directly, give me exact settings and an ordered workflow for a suitable audio editor. Also give me a short comparison checklist for checking the cleaned file against the original, including signs that the processing is too strong.
Open it prefilled in ChatGPT or Claude, or copy it into Gemini, which takes no prefill link.
Use a tool built for this
The distant third
What it gets wrong
- AI cannot restore speech or music detail that was never recorded or has been permanently clipped.
- AI cannot know whether a faint sound is unwanted noise or an important part of the original performance without your judgement.
- Heavy cleanup can create watery voices, metallic music and other new artefacts that the tool may not identify reliably.
- AI cannot guarantee that the result will match the character of the original equipment, room or performance.
Even on a YES, the friction has a name: verification cost 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 | 1 |
| Liability | 2 |
| Effort delta | 2 |
| Total | 9 / 10 |
The methodology and its thresholds are published in full.
FAQ
- Can AI remove hiss from an old recording?
- Usually, yes. AI can reduce steady hiss, but aggressive settings can make voices or music sound watery, so compare the result with the original and use the lightest setting that helps.
- Can AI repair a damaged cassette recording?
- It can improve some problems, including steady noise, hum, clicks and uneven volume. It cannot reliably restore detail that was never captured, and severe clipping or dropouts may remain.
- What is the best AI tool for cleaning up an old recording?
- Descript is a practical candidate because it supports audio editing and AI cleanup workflows. For an old music recording, test a short section first because a tool designed around spoken audio may treat instruments differently from speech.
- Will AI change the words in my old recording?
- Basic cleanup should not change the words, but some enhancement tools can remove consonants or create synthetic-sounding speech when pushed too far. Keep the original, compare both files, and reject any version where the words or musical notes have changed.
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.