What the result says.
When the try finds them, the result reads: “Other voices seem to sit behind the speaker, so speech cleanup is used.”
SOLIPSIUM Super resolution audio engine
Background voices
Lower the voices behind the speaker.
Drop a recording made in a café, an office or a room with the television on. Solipsium measures it first, and lowers what sits behind the main voice.
Drop a recording here
Solipsium checks it in your browser for clipping, hum and noise, without uploading it. It can then repair the first 20 seconds for you to hear, with no account.
WAV, MP3, M4A, MP4 and more, up to five minutes for the check.
Your first 20 seconds
After signing up, drop the same file into the editor to repair all of it.
What Solipsium hears
Make a free account to repair the whole recording
Then drop the same file into the editor. Scans and short repairs are free.
How it works
The main voice stays in front.
When the try finds other voices sitting behind the speaker, it uses speech cleanup, a speech model that keeps the voice's full range and lowers what surrounds it.
Which voice is kept.
The model keeps the voice in front, the loudest and nearest one. A second person at the same distance is not behind the speaker, and stays.
Steady noise too.
A café's chatter often comes with a steady hum of machines. A steady floor within 26 dB of the voice gets the same cleanup: “A steady noise floor was found, so it is removed.”
What it cannot do
It lowers; it does not pick people out.
There is no step that separates one person from another.
No choosing a voice.
You cannot pick which person to keep. The repair keeps the voice in front and lowers the rest.
A television.
A television brings voices and music together. Solipsium's free try is built for speech, and its cleanup may remove the music along with the voices.
Words that were drowned.
Where the chatter was louder than the speaker, the speaker's words were never clearly recorded, and lowering the chatter cannot bring them back.
Questions
About background voices.
Will it remove laughter and applause?
Sounds behind the voice can come down with the chatter. Switch between Original and Repaired as they play, on one clock, to hear what was taken.
Two people talking at once?
Both stay when both are in front. A recording of each person on their own track keeps them apart.
Does the check in my browser find other voices?
No. It names clipping, dropouts, mains hum, background noise, muffled sound and a very quiet recording. Voices are measured on Solipsium's server when you press Repair the first 20 seconds.
Will my own voice sound different?
It can. The editor's note on Full-band speech cleanup reads: “Targets hiss in speech. Check quiet or distant voices; may remove music and ambience.”
Background music rather than voices?
In the editor, Rebalance finds the voice with a speech model and turns the rest down. In an October 5 test, the music usually went down by less than the setting asked.
Is my recording uploaded?
Not for the check, which runs in your browser. Pressing Repair the first 20 seconds sends those 20 seconds; when your browser could not open the file, or it runs over five minutes, the file itself goes to Solipsium's server, so the try is offered for it only up to 25 MB. Solipsium keeps only the first 20 seconds and deletes them within an hour. If you are signed in, a file you drop or choose is uploaded to your account and opens in the editor instead. The privacy notice has the rest.
Free audio tools
Related free tools.
For the room around the voice.
- Remove background noise
Noise, hum and clipping, checked together.
- Remove echo from audio
When room echo is reduced, and when it is not.
- Clean up a meeting recording
Voices behind the speaker on a call.
- Clean up podcast audio
A guest's café, cleaned before the mix.
- All free audio tools
Every free tool on one page.
The whole recording.
Make a free account and drop the same file into the editor to repair all of it, up to 30 minutes at a time.
Repair the whole recording
Free account, no card.