removing *really* bad hum

I just made a 3 hour recording today, and about half way through, one of the lapel mics suddenly “went bad” and started generating a terrible hum which I didn’t hear until it was all over.

I really don’t want to have to re-record over an hour of this Q&A podcast, and am willing to sacrifice some quality to avoid that. I don’t really know anything about editing audio other than copy & paste, and would love it if anyone could give some suggestions for trying to clean up this audio.

Here’s a sample to give you an idea of the damage:

I used Audacity’s noise reduction, which worked, but made it sound almost as bad, just in a different way.

Is there anything that can be done better? Thanks in advance for any help you can give! :slight_smile:

How about this:

Here’s the plug-in that I used (it’s not on the Audacity wiki yet, but will be)
de-hum.ny (2.11 KB)
Installation instructions:

Settings that I used: Europe 50 Hz, 8, 8, 35

OMG that’s fantastic!!! Thank you so much, I was really pretty demoralized, but now I’m riding high. You’ve made a difference today sir! :slight_smile:

A tip, in case you’re not aware: Avoid working with MP3 (or other compressed formats) whenever possible. MP3 compression reduces the sound quality each time it is encoded (a cumulative effect, like taking a photocopy of a photocopy) and the damage, though usually small, is not recoverable. MP3 encoding should only be done after the editing / processing job is complete (if you need to use the format at all, which generally you do for podcasts).

Backups and production can be Exported in WAV (Microsoft). It’s the default Audacity “perfect” sound format and it’s acceptable on all three major computer systems. Intermediate steps can be Saved (not Exported) as an Audacity Project.

Projects can be handy because you can save all your show tracks, desktop and screen layout. It allows you to stop work and pick it up tomorrow, but fair warning, Projects only open in Audacity, they can be brittle and they do not save UNDO.


Thanks for the tips… I actually work with the raw audio audacity pulls out of the MTS files my camera generates when recording, and I don’t save to MP3 until I’m all done.

As an aside, anyone have any suggestions for what standard setting should be for that MP3 to get it as small as possible without compromising too much on audio quality?

The fuzzy rule is mono at 32 and stereo at 64. Much lower than that and the show falls apart pretty quickly and a majority of people can tell there’s something wrong.

Even if you have a show at those values, you won’t be able to compare it to the original. That kind of quality is usually much higher. Audacity default MP3 used to be 128. I think it’s one of the variable compression settings now.

MTS files my camera generates

These MP3 numbers are only valid if the show has never been compressed or processed before


The BBC’s recommendations for podcast bit rates are:
MP3 Mono Speech: 64 kbps , 44.1 kHz, constant bit rate.
MP3 Stereo Music: 128 kbps , 44.1 kHz, constant bit rate.

Absolute minimum for mono MP3 is 32, so 64 is about right, and constant bit rate 128 used to be the Audacity stereo default.

That’s for listening only. Not archive or pre-production. If you have to edit or do production on your work it gets more complicated.

There are reasons you might want it higher. Recommended Mono MP3 ACX Audiobook Submission is constant bitrate 192. They do that because they may need to change and recompress your work. MP3 quality degenerates every time you make a new one from an old one.


There really are some great solutions now.

I remember when I was recording about 10 years ago and picked up some hum on an electric organ (dodgy electrics), I spent so long trying to EQ out parts of the 50/60hz and it’s harmonics, in the end I just high-passed the thing!

I expect that the original poster has either solved the problem or given up after 2 years.