Newbie noise reduction help

Hi all, my first post.

I recently discovered Audacity and I’ve used it a few times to do very simple things and it has worked like a charm.
Now I’m in a bit of a situation, I’ve tried searching but my problem is my english, since the audio field is a bit new to me there are a lot of words and commands in the program that I simply don’t understand, among phrases that I’m not familiar with.
So in simple terms I don’t know what I’m looking for.

I do a little video editing but the audio has always been in another hands. But now it isn’t.
I did some interviews that were filmed on a bad spot, very noisie and I also had equipment that I wasn’t familiar with.
It’s a DSLR project and the audio was recorded on a Zoom H4n device. It is very sensitive and it picks up everything (I in fact knew the audio would need a little cleaning but wow I had no idea).

I can clean up the audio so and so but it’s still way to messy and if I go further the subjects start to sound like Robocop.
I don’t know what it is, something like going to a rock concert but the base player isn’t playing. The music is there but there’s something missing - you know?
It’s like the subjects voice are missing their depth.

Is there any way for Audacity geniuses like you guys to help me out here?
Any tricks I should try other than the youtubes tutorials: Get noise profile → select all → OK ?

I do a little video editing but the audio has always been in another hands.

You had the easy job.

(I in fact knew the audio would need a little cleaning but wow I had no idea).

Sound kills more shows than pictures. You can’t just zoom in a microphone like you can a camera.

We can’t tell which version of Audacity you have from your posting, but you should upgrade to the latest Audacity 1.3.

The noise reduction tools in Audacity 1.3 are far better than 1.2.

Most cases of bad audio are forever. The tools will not rescue bad audio. They were intended to clear minor show problems that do not happen often or are very bad. They will not turn a bad show into a good one.

Noise reduction is for a sound that is constant for a long time like air conditioning, refrigerator, or fan sound. Maybe small microphone buzz. That’s noise. Noise Reduction will not remove sounds that change like road traffic, an airplane in the sky, or another voice. We can not remove a TV set sound from an actor’s voice.

if I go further the subjects start to sound like Robocop.

I think you are using the Audacity 1.2 noise reduction. it’s easy to get damage like that in 1.2.

What kind of noise is it? Describe the show and your shots.

This is a thing written on how to record voices. I thought we wrote one.


The audacity’s “noise removal” can only be successful where the noise is constant throughout the recording, like a hiss artefact. So it can’t remove (say) other voices speaking at same time as your interviewee.

If you read the page on noise removal then you’ll know all there is to know on the subject … Missing features - Audacity Support

If by “depth” you mean the recordings lack bass there is “bassboost” in the “Effect” menu you could try.
Or if by “depth” you mean stereo image, you could try Steve’s pseudostero plugin, which makes mono sound like stereo.

Thanks for your reply. I’m going to read it and test it in a few.
The noise is constant and the same, this is due to a huge air ventilation fans within the space the interviews are taking place.
I don’t mind noise and background noises due to people working in the space nor the fan noises but I do need to clean it up a bit more.
The fan noise is still too loud, but like I said once I reduce it to acceptable level the subjects sound like Robocop or the bass is gone from their vocals.

It completly slipped my mind saying that I am using the 1.3 Beta version.

To descripe the shoots were the audio is bad:
It’s shot in a big garage - like a warehouse - but it stores helicopters and airplane.
I’m interviewing people from the Search and Rescue helicopter team. Off course I wanted the choppers in the background. Realizing the audio would need a bit of cleaning I took my chances and shot the interview there. I don’t mind at all the background noises and some fan noises but the audio sounds like it was recorded with an old and cheap MiniDV camera with a bad mic (loud bzzzz sound) - I want the audio to sound like it was recorded in a loud space and that people are working but with a propper recording equipment.
Hence I’m not looking for a audio that sounds like it was recorded in a studio. I am aware that it won’t be great but that’s fine. In fact I want the background noises, the working men in the background to be heard but that damn fan…it really needs a decent muffler.

Sorry for not explaining that right away.
This is my first DSLR project and I’m very excited about it. I took the cameras to their very limits and learned a lot off of this project. But now I’m trying to clean up the mistakes I made along the way.
I never took care of the audio bits before but now it interests me a lot as well. I have a lot to learn and most likely more equipment to buy but I’m trying to do things as cheep as possible, to see what you can get away with without losing to much quality (visual and audio).

So I will be using Audacity a lot since I’m taking care of my own audio now, I have a lot to learn, both on the audio recording and editing but I can’t wait to learn.

This does not help for the footage that you have already shot, but for future reference, the way this is done professionally is to cheat!
Record the interview with as little background noise as possible. In some cases this can even be done in a different room, or at a time when the room is empty and quiet, but failing that, use a microphone as close as possible to the person speaking - that could be a lapel mic or a directional microphone that is just out of camera shot.

Then record the background noise (without the interview). Again you can cheat by setting the microphone away from the noisy fan and closer to the more interesting “work noises” that you want.

When you come to edit your video the two audio tracks (clean voice track and noise track) can be mixed together in a controlled way.

(I used this technique recently for an interview in a busy and noisy commercial kitchen. If the voice had been recorded with an ambient microphone, some of it would have been totally lost behind the noise. Some of the kitchen noise was extremely loud, but on the recording the overall sense was of much less noise because the recording level for the noise had to be set very low to avoid clipping when there were really loud sounds, so I ended up using 4 noise tracks mixed together to (artificially) create a realistic ambiance without burying the interview voice. It worked very well.)