Interview subjects talk too loudly!

There are three voices on my recording - one woman and two loud men! When the men are speaking, there’s a crackle on the recording. They both have resonant voices and were quite close to the digital recorder.

Is there any way of reducing their volume (while keeping the female voice as is) or removing the crackle? So far, I’ve tried normalization, click removal and lowering the bass without much success.

Many thanks!

You may have some degree of success reducing the distortion by using the “ClipFix” plug-in. There is not much you can do about the relative levels of the voices if it is all recorded on the same track apart from perhaps a little Eq and a bit of dynamic compression. The real solution is to move the recorder further away from the men during the recording session.

After you apply clipfix and repair the show just as it is, you might be able to use this compressor to even things out a bit. This is in response to a level problem I had…

The studio feed from a radio show has the same problems you do. The compressors at the radio station even things out so you get a nice controlled show in your car.

I think the tool only works in Audacity 1.3, but that’s no big deal since you can have both versions in your machine as long as you only use one at a time.

A more severe tool also available in 1.3 is Effect > Leveler.

The big kids get around these problems by providing a lavalier microphone for each person and a small sound mixer to adjust each one. One of the complaints of the Zoom series of field recorders is the inability to change the volume once the show starts.


With the Zoom H2, you can. You can also have AGC (Automatic Gain Control) switched on if you wish, but I use mine for music, so I always have AGC switched off.


But as a practical case, you can’t have the machine smashed into somebody’s face for an interview and do that.

Do you know anybody who uses the AGC?


I’ve come across lots of people that do - and they all wonder why their recordings sound rubbish :smiley:

I’m sure there must be some situation where AGC is desirable, but I’m yet to find it.

Long ago I used to have a tape deck that had AGC - and similarly it produced rubbish recordings, so I never saw the point of it either …


<<<Do you know anybody who uses the AGC?>>>

Forgive me. Know anybody who’s used it more than once?

The end of this week’s NPR - This American Life has an interesting problem. It’s an interview in a car or bus. There is an odd variation in the interview voices which turns out to be a result of sub-bass sound I’m sure goes right past people who don’t have killer sound systems. I do and I almost missed it. Every time the rumble increases, the voice dips. I deleted everything under 100Hz (the classic rumble filter) and then ran the show through what’s-his-name’s broadcast compressor-limiter. Poof. Perfectly clear interview with minimal swooshing and variations in volume. That probably drove them nuts and nobody could hear what was really going on, except the performance wouldn’t behave while they were editing.


Things that can happen without AGC:
Ooh that’s weird… time for a rumble filter.

Thanks for the tips! I will give them a try.

I hate working in a vacuum. It makes my ears pop.

<<<Ooh that’s weird>>>

Not if you’re recording the choir and organ at St. Pats on Fifth Avenue, it’s not. That’s what, sixteen foot flute stop about E or F pedal to the left of your left knee.

Perfectly normal.