How can I keyframe the amount of noise reduction applied to audio?

I don't see any way to keyframe noise reduction, and the effect doesn't apear in the audio aniimation editor. Is keyframing the amount of noise reduction possible? What am I missing?

I'm not sure that's even possible in any audio noise reduction tool. The reason is that most of these plug-ins require a certain amount of "lookahead" time in order that the signal can be analyzed for noise and the processing to kick in. But what is the noise like in you track, such that you would want to keyframe it? Because if there's noise that fades in and out in intensity, a typical NR filter might not even work at all. Have you tried applying it and if so, what sort of results are you getting?

Blade up the clip. I'm sorry, that's so wrong. SCISSORS up the clip, place short cross dissolve transitions, apply NR to only the sections needing it.

Oliver Peters wrote:

I'm not sure that's even possible in any audio noise reduction tool. The reason is that most of these plug-ins require a certain amount of "lookahead" time in order that the signal can be analyzed for noise and the processing to kick in. But what is the noise like in you track, such that you would want to keyframe it? Because if there's noise that fades in and out in intensity, a typical NR filter might not even work at all. Have you tried applying it and if so, what sort of results are you getting?

Thank you for your response. I am familiar with the issues you raise regarding (a) lookahead time, and (b) the ineffectiveness of broadband noise reduction against intermittent unwanted sounds.

In this case, a second audio track is introduced about halfway into the video. As it fades in, I want my audience to hear the pre-noise-reduction version and have it slowly – at my preferred rate – change into the fully-processed version of the same track. I get that I might be able to approximate that by duplicating the track and crossfading a pre- and post- noise reduction version, or by exporting the noise only and doing something similar, but keyframes exist to solve this type of problem.

Good news: I had only been looking at keyframing FCPX’s built-in noise reduction, which seems to not be possible, but I use iZotope RX8 for all audio cleanup that is beyond FCPX. Meaning there were unexplored options on the table. RX ships with a collection of audio plugins, one of which is “Spectral Noise Reduction.” In ‘Adaptive’ mode, Spectral Noise Reduction works similarly to the built-in noise reduction… and I was just reminded that all of its parameters (even lookahead time) can be keyframed.

If you have RX, then you should also have the standalone application. Maybe use it instead of the plug-in for better processing. But to do what you want, I would suggest creating a fully processed track. Export that and re-import it and line it up in sync with the unprocessed track. Then mix between the two.