Automatic mic leveling? or Software-based commpression?

Hi Folks,

I’m getting varying levels when recording spoken word vocals through a mic and was wondering if there was any way for Audacity to automatically set the input signal to a consistent level. I’m hoping to avoid either doing too much work Normalizing or Compressing the tracks after the fact or having to purchase a hardware compressor like the RNC 1773 (though I really want one!). Any thoughts?

Thanks a lot!

wondering if there was any way for Audacity to automatically set the input signal to a consistent level.

Not currently, no. I also don’t think it’s likely to be added any time soon since the Compressor effect can do this after you’re done recording.

If you don’t want to spend a lot of time Normalizing, you don’t have to. It’s as simple as pressing [ctrl]+a when you’re done recording everything, then clicking effect → normalize.

If you Normalize multiple tracks at once, Audacity will deal with each track individually and normalize all of them to the same level (this is not true of the Amplify effect).

Thanks for the reply. I have problems with Normalizing a vocal track that has louder and quieter levels. Sometimes, when I Select All and Normalize, nothing happens. This is while using the Normalize plug-in by Dominic Mazzoni that ships with Audacity with the “Normalize maximum amplitude to -3 dB” unchecked. When I leave that option checked it leaves the louder and quieter levels in the same relation but makes the track quieter as a whole. This is why I have been selecting the quieter parts and using the Compressor plug-in (also by Dominic Mazzoni) with the default settings. This works fine, but it does take time. I’m hoping to spend lest time in post and more time recording new material.

Oh, ok. The normalize function works just fine, what you’re looking for is a Compressor. They aren’t the same thing.

An outboard compressor might be your best bet, but try using Audacity’s built in compressor (or use Audacity 1.3.4’s Leveler plugin, it’s a very drastic compressor meant strictly for speech).

The Normalize tool finds the loudest peak of a show and changes the loudness of the whole show until that peak goes where you told it to.

That’s not what most people mean when they say Normalize. They want all the parts of a performance to end up the same volume like an AM radio station. That need turns the tool from one-click into an art form. AM stations do it with expensive real-time limiters and multi-band dynamic compressors.

Peak Limiters, or just limiters are real time clippers. Any volume over a certain set level is reduced on the fly to a set level. That would seem to be perfect except for music. Bass notes drive limiters nuts. You get pumping if you’re talking and a bass guitar is playing behind you, every fourth word will be missing as the limiter tries to even things out. Carried to extremes, limiters sound like airline pilots.

OK, so that sucks, so we bring in the compressors. That’s the tool that gradually brings the overall volume up and down to even things out. Cool, right? That’s the one for me. Compressors aren’t easy to set up. The simplest compressor has at least four different controls. Attack, Release, Hold, and Threshold. Tune until your brain bleeds.

Audacity has a tool for this. Effect > Compressor. It has five controls counting the rubber band sliders for compression characteristic. Have a good time.


Hey guys, thanks for the info. So, basically, my previous assumptions are correct, right? Either spend time in the post working with the Compressor effect or buying some hardware? I figured as much. Guess I’m off to eBay/Craigslist to look for that RNC…

Yes, that’s correct. The outboard compressor will probably save you quite a bit of editing time.

Personally, I prefer not to use compression when recording, and use compression after I’ve recorded “dry”.
If you over-do the compression, it’s almost impossible to “uncompress” it accurately.

I will however use “limiting” (a very fast “brick wall” compression to prevent any spurious peaks from clipping), but will try to keep the levels set so that the limiter does not actually do anything. (I just have it there so that I don’t ruin a recording because of one unfortunate plosive).