Fixed gain multiplies the whole signal. It raises the wanted voice and the unwanted background together, so the voice does not become cleaner relative to the noise. Check the source before adding more gain.
The quiet parts contain sound too
Hiss, room tone and distant traffic can be hard to notice at the original listening level. A gain increase makes them more audible along with the speaker. That is expected behavior for a fixed-gain tool.
For a simple example, speech near −18 dB and background near −48 dB retain their approximately 30 dB separation after a +6 dB adjustment. Both levels rise; the gap between them does not improve. This is an illustrative level calculation, not a measurement of your recording.
First check how you are listening
Check the device volume and output selection before modifying the file. If everything is quiet through one pair of headphones, a system-level adjustment may be more appropriate than making a new audio file.
If only this recording is quiet, try a modest gain change and listen to both a quiet passage and a loud one. Watch the predicted peak; an acceptable average listening level does not prevent a short loud peak from clipping.
Choose a different fix for persistent noise
Noise under speech requires a workflow that can address that noise. Volume Changer does not separate a speaker from the background, and Silence Remover only shortens selected quiet time.
If you can record again, move closer to the microphone, reduce the unwanted sound at its source and make a short test recording. If you cannot, keep the original and use an appropriate noise-treatment workflow before deciding on the final level.
- Noise only in an unwanted passage: manual trimming may remove that passage.
- Noise throughout the voice: fixed gain will not isolate the voice.
- Already distorted source: lowering the level does not undo the damage.
Questions about this task
Would negative gain remove the noise?
It lowers the voice and noise together. It does not selectively remove the background.
Can I delete the quiet gaps instead?
You can shorten unwanted gaps, but noise inside the speech you keep will remain.