Remove changing audio noise from laptop fan

Hi community,

just made my first steps in DR yesterday and already have a problem

Removing steady, unchanging background noise from my (quite noisy) laptop works perfectly fine within a few clicks.

But during a 20 minute video (educational stuff, sitting at a desk in front of a green screen) this noise changes more or less constantly. My laptop is quite powerful, but every few minutes the cooling fan „acts up and excelerates“ and gets quite loud.

For 20 minutes of final video I shoot around 60 minutes of footage. After putting this 60 minute video on the DR timeline I want to begin by first reducing this laptop background noise for the whole footage and only then start to cut it down into pieces and into the final 20 minutes.

The problem is: If I take the room sound from a more quiet laptop moment it only works well for these more quiet parts of the video. During louder laptop parts of the footage the noise of course isn’t completely reduced which in the final video sounds like „there suddenly is background noise“.

If I take the room sound from a louder segment then the loud segments sound good but in the more quiet segments my voice sounds kinda off like I‘m recording in a phone booth or something.

How can I fix this?

Buying a new laptop is no option btw because this is the new laptop It also has to stay in the room where I‘m recording so the noise itself won‘t go away.

Thanks

Use a lav and connect an external screen and keyboard to the laptop to put it further away or under the table.
Sorry, no good fix for this in post!

Can you slice up the clips in Fairlight and apply noise reduction separately to the clips depending on their noise profile? You could try copying and pasting clip attributes within each type of clip (low noise and high noise) to save time.

Otherwise this sounds like a job for Izotope RX, which can do adaptive noise reduction over an entire sound file. You can set up a round-trip process to Izotope from within Resolve so you don't have to export the audio and reimport it.

Sit further away from the laptop, use a longer mic cable, and record that way. A more directional microphone like a cardioid or a supercardioid will tend to ignore background noise around you a little more than an omni microphone, so it's room-dependent. It also helps to put the person speaking in an acoustically-controlled environment. There are acoustic screens you can buy, and in a pinch, you can use blankets and pillows to surround the person to dull the sound outside them in the room.

I don't think it's a wise idea to record bad sound and then try to fix it with NR after the fact. It's better to minimize the noise first so that you have to do a minimum amount of processing in post.