Save Noise Profile When Saving Project

Why does Audacity lose the “noise profile” we last used when we save a project? I edit sermon audio files for my church website, and sometimes there are background noises that I am able to take out by using noise reduction. I can’t remove them from the entire track at once because doing so affects the quality of the audio. But if there is a small section of audio, I can often use noise reduction to get rid of the unwanted background noise. However, when I save the project and come back to it at another time, the noise profile is gone. I have to hunt for another otherwise silent section of audio that possibly has the same background noise so I can do another “noise profile,” but often I can’t even find the same type of background noise. (Examples: baby crying, someone coughing, even “thumps” from the pulpit vibrating or being touched) So when I save the project, the noise profile I last used also should be saved along with the project.

I’m using Audacity 2.3.1 so if this has already been added to a newer version, please let me know and forgive this suggestion.

Thank you.

Because the noise profile only exists in RAM.

You could of course export a short section of noise as a separate audio file and then use that to generate the noise profile when you need it in the future.

I think I can invoke a Friedel Cycle here. Audacity is perfectly capable of selecting its own noise profile by just OKing the Profile step.

As I understand it, it averages out the whole performance. On Average, destructive noise is present for the whole show, so it’s by far the major contributor to the profile or damage. Did you try that? Just OK the profile without selecting anything?

Anyway, to complete the cycle, what you really want is the ability to select a Preset and Auto in the Noise Reduction panel making it one-step. That’s too much work so we won’t do anything.


I just want it to remember the last saved noise profile for a particular project when I save that project. It can remember my stopping point and all of the changes I have made. So why can’t it remember the noise profile? I don’t want a “preset,” and it’s too much work to save a tiny file with a “thump” sound or some other weird sound every week.

The background noises vary from recording to recording. Even the same type of sound may be slightly different on a different recording. So I would just like Audacity to remember the noise profile within the saved project file so it will still be there when I reopen the file the next day. It’s hard to edit a 30-50 minute speech file in one sitting!

That’s not exactly what I mean.

One of the current sources for Profile is Overall Average. Don’t select anything and press “Profile.” Let’s say you wanted to use that technique and your Project already has a Profile from a previous edit. How do you know and then how do you choose?

Let’s say you drag-selected a new portion of the show for Noise Reduction. How and when does it overwrite the old Profile? Does it overwrite?

See what the problems is? You can’t add a feature in a vacuum. It has to work with everything else.

The real solution is stop making noises.


I have no idea what you’re talking about. As for “stop making noise,” I have no control over the recording. Our church has some volunteers who operate the sound board and record the CD. Even if we figured out what’s causing the “thumping” noise that I think is coming from the pulpit, there is often going to be some kind of other background noise in this type of environment.

“Noise Reduction” may not be the best tool for dealing with thumps. Noise Reduction is intended specifically for dealing with low-level constant noise, such as low-level hiss or hums. For it to work effectively, the “noise” needs to be fairly constant.

For transient sounds (such as clicks and thumps), “Spectral Delete” may be better.

1 Like

Can you add a new track to your project to hold only noise clips? Instead of selecting some audio and creating a noise profile right away, copy and paste that selected audio into the noise track as a new clip and then create the noise profile from that clip. Then if you need a piece of noise again later in the recording you can grab it from the noise track. (Of course, don’t forget to mute the noise track!)

1 Like

I guess I could copy/paste the noise into a section at the end of the file and delete it when I’m finished. That sounds simpler than creating a new track. But it would still be easier if Audacity would just save the last-used noise profile when we save the project. I’m not a pro; I’m just an amateur volunteer for my church but I want to do the best job I can do. I probably spend too much time on the editing already because I’m so picky. The more things I figure out how to fix, the more picky I get with trying to fix the sound.

If I knew how to use “spectral delete” and find the frequency for the thumps and coughs, I would try that. I do have to be careful with using noise removal because it can also change the quality of the speaker’s voice. I don’t know if “spectral delete” would be more precise, but I have no idea how to use it, anyway.

I have been taking out “clicks” manually. I’m concerned if I try to automate anything about the clicks, it might take too much away from final consonants such as Ts and Ds. Some of the recordings have “clicky consonants” (I don’t know what else to call them) and I fix a lot of them manually, but I have to be careful and not take out too much.

Search for Paul Licamelli’s videos on spectral editing in YT. You don’t have to be a sound engineer to use what he demonstrates. I can do it.