A powerful interview, a great performance, or an important recording shouldn't be undermined by background noise. Whether it's the hum of an air conditioner, the echo of an untreated room, outdoor wind, or distant traffic, SuperCool can strip the noise and return a clean, professional-sounding file.
Here is how to prompt for perfect audio restoration.
1. Identifying Your Noise for SuperCool
Different types of noise respond differently to cleanup. Before you write your prompt, identify what you are working with so you can describe it accurately to the SuperCool AI:
Consistent background noise: This is the easiest to remove. It includes air conditioning, fan hum, electrical ground hum, computer fan noise, or consistent room tone.
Erratic or environmental noise: This is more complex and requires a careful approach. It includes wind, outdoor ambient sound, intermittent traffic, crowd noise, echo/reverb from hard-walled spaces, or overlapping background conversations.
Voice quality issues: This requires a different type of correction entirely. It includes harsh sibilance (overly sharp S and T sounds), inconsistent volume levels (loud vs. quiet words), or muffled speech caused by poor microphone placement.
Naming the specific type of noise gives SuperCool a much clearer brief and produces a significantly better result.
2. Functional SuperCool Prompt Examples
Here are the most common audio restoration scenarios and the exact prompts you should use in SuperCool:
The Podcast Polish: (For home studio recordings with consistent room hum)
"Clean up this voiceover recording. Remove the consistent background hum and the slight room echo. Enhance the vocal clarity and presence so it sounds like it was recorded in a treated professional studio. Do not over-process: the voice should remain warm and natural after cleanup."
The Outdoor Rescue: (For interviews captured outside with wind or traffic)
"This recording was captured outdoors. Reduce the wind noise and distant traffic sounds as much as possible. Prioritize preserving the natural clarity and warmth of the speaker's voice. Do not apply heavy processing that makes the voice sound robotic or hollow."
The De-Esser and Level Corrector: (For recordings with harsh consonants or uneven volume)
"Remove the harsh sibilance from this recording: the sharp S and T sounds are too prominent and fatiguing to listen to. Also level the volume so the speaker's quietest and loudest moments are consistent throughout. Keep the natural tone of the voice intact."
The Echo Removal: (For recordings made in rooms with hard floors, bare walls, or poor acoustics)
"Remove the room reverb and echo from this recording. The speaker sounds like they are in a large untreated space. I need the voice to sound close, dry, and intimate as if recorded in a properly treated room. Keep everything else the same."
The Archival Restoration: (For old or degraded recordings with accumulated noise and distortion)
"Clean up this archival recording. Remove the background hiss, crackle, and tape noise while preserving as much of the original voice quality as possible. Prioritize clarity and intelligibility over a perfectly clean signal."
3. Reviewing and Refining Your SuperCool Result
Listen to the cleaned audio in full before using it. The three things most likely to need a follow-up correction in SuperCool are:
Over-processing: If the voice sounds robotic, hollow, or like it's coming through a phone, the noise removal was too aggressive. Prompt a lighter pass:
"The cleaned version sounds over-processed and slightly robotic. Apply a lighter noise removal pass and prioritize preserving the natural warmth and texture of the voice. Keep everything else the same."
Remaining noise: If the background sound is still audible, request a second targeted pass:
"There is still a faint low-frequency hum audible during the quiet pauses between sentences. Apply an additional pass targeting specifically that low-frequency hum. Keep everything else the same."
Lost presence: If the voice sounds thinner or less present after cleanup, ask SuperCool to restore it:
"The cleanup removed some of the warmth and body from the voice. Boost the mid-range frequencies slightly to restore the natural presence and richness of the speaker's voice. Keep everything else the same."
4. Getting the Best Source Audio for SuperCool
The cleaner your original recording, the better your SuperCool cleanup result will be. Here are the most impactful things you can do to improve your source quality before you even upload it:
Record in a soft space: Rooms with carpets, curtains, bookshelves, and soft furnishings absorb echo. Hard-walled bare rooms are the worst environment for voice recording.
Use a dedicated microphone: A USB microphone or a wired headset dramatically outperforms a built-in laptop microphone. Built-in mics are the hardest source material to clean up.
Use high-quality formats: WAV files clean up much better than compressed MP3s. Always use the original uncompressed file if you have it.
Send the raw file: Do not apply EQ, compression, or any effects before cleanup. Noise removal works on the original noise profile of the recording. Pre-processed files make accurate cleanup much harder for the AI.
5. Pro Tip: Cleaning Video Audio in SuperCool
If you are cleaning up the audio track of a video recording, do not extract the audio first! Upload the full video file directly to SuperCool. The AI will clean the audio track natively within the video and return a new version of the full video file with the cleaned audio perfectly in place. This saves you an extra step and keeps your video and audio perfectly in sync!