How to remove noise from AGC-leveled audio file?

[Alexey Lukin]

Give us a short sample to look at. Sometimes it's possible to automatically level the noise using a sidechain compression based on a bandlimited noise signal (with an added bonus of undoing the work of AGC).

[Rob Murray]

I know this might be a pain depending if it's a lengthy segment, and just how much/often the level is changing, but you can try several (3?) levels of Broadband noise reduction, and edit between those where appropriate. Not the fastest/simplest nor the most elegant, but might work depending on the material.

The only other (slightly crazy) idea crossing my mind at the moment is trying to key in on the noise while avoiding the content into a sidechained compressor bringing it back to a steady noisefloor, sort of try to undo the AGC, de-noise, and then apply any dynamics needed to resulting file, but the odds of this solving more than it fixes, or avoiding tons of compression artifacts are probably 0.

Are they focusing on fidelity or audibility? Is this music or speech? Broadband, focused etc. Maybe post a sample?

[Nonlinear]

Thanks guys for the suggestions.

I don't have access to post a sample right now but I can tell you that it's a combination of music and dialog - a musical play that was recorded. The noise floor sometimes rises up quite high during the lower level passages and it becomes distracting.

I'm going to try both of your suggestions - the keyed sidechain comp and some manual split points set at different thresholds.


Riccardo - perhaps this post is better suited for the "Post Production" forum? If so, please move it. Thank you.

[Adam Dempsey]

Originally Posted by Rob Murray

Are they focusing on fidelity or audibility?

That alone can be a very important distinction to make and will dictate the approach.

[Nonlinear]

Originally Posted by Adam Dempsey

That alone can be a very important distinction to make and will dictate the approach.

I suppose it's a matter of "fidelity". The issue is not that the audio is lost in the noise it's that the noise is distracting to the audio.

During musical passages the noise floor is virtually inaudible whereas during dialog passages the gain (and noise) comes way up. Some compressor literature refers to this as the "breathing" effect.

Is there any way to reduce this type of noise?

[Adam Dempsey]

Originally Posted by Nonlinear

I suppose it's a matter of "fidelity". The issue is not that the audio is lost in the noise it's that the noise is distracting to the audio.

During musical passages the noise floor is virtually inaudible whereas during dialog passages the gain (and noise) comes way up. Some compressor literature refers to this as the "breathing" effect.

Is there any way to reduce this type of noise?

Sounds like perhaps an expander/single ended noise reduction, on just the dialogue sections, might work. In the analogue domain and for cassette mastering I used to use a Symetrix 511A extensively for such things, for gentle expansion combined with its sliding HF filter. Used judiciously it worked really well, quite smooth yet fast enough for dealing with speech.

[Nonlinear]

Originally Posted by Adam Dempsey

Sounds like perhaps an expander/single ended noise reduction, on just the dialogue sections, might work. In the analogue domain and for cassette mastering I used to use a Symetrix 511A extensively for such things, for gentle expansion combined with its sliding HF filter. Used judiciously it worked really well, quite smooth yet fast enough for dealing with speech.

OK, thank you - will give it a try!