Speech recording - What Compressor parameters should I use?

[Mekratrig]

I'm doing speech recording from a very basic setup - a single Shure SM58 clone mic, Kustom KMP4080 amp, and a couple of Kustom KSC10 speakers. The amp has two outs, a pair of RCA jacks and a 1/4" mono jack. I run the mono jack into the mic input jack on a old rebuilt HP/Compaq notebook PC running Windows XP, and record the audio using Audacity 1.2. (I know, I should install the 1.3.13 beta, and will, as soon as I gain enough familiarity using it on my desktop.)

For months, I've been taking the recorded file to my desktop to do manual editing of the file, typically consisting of: Trim the start; trim and fadeout the end; run minimal noise reduction; run the high pass filter with a cutoff of 300. Then I spend what can turn out to be hours going thru and manually selecting the spots where clipping occurs, as well as areas that get close to clipping, and dropping the amplitude down on them, in order to run the amplify effect on the whole file to maximize the sound levels.

Then, it finally dawns on me that most of the above effort can be done a lot faster and probably a lot better, by running the Compressor effect. Great! But can anyone suggest what I should use for the parameter settings when recording speech? Threshold? Noise Floor? Ratio? Attack & Delay times? Despite pouring over the description of the effect in the documentation, I have no clue.

Also, one of the basic difficulties I'm having is understanding the graph used in the Compressor. Forgive a ignorant newbie's remarks, but it seems backwards to me. Why are all the decibel values negative? Why do the axes cross at -60dB and the zero point appear in the upper right corner?

Any and all comments or suggestions appreciated!

[DVDdoug]

There are no easy answers, because every recording is different. If you are doing the same kinds of recordings over-and-over, once you have an idea of what works with your set-up and your recordings, you can use that as your default compression set-up and make small changes when necessary.

Threshold? Noise Floor? Ratio? Attack & Delay times?

Do you understand what any of these these terms mean? I think you first need an understand of what they mean, and then you can experiment with the settings to see how they actually affect the sound. Maybe try over-doing the compression settings to the point where it sounds bad...

And for speech, you can try The Levelator (FREE!!!), which doesn't have any options.

[Mekratrig]

There are no easy answers, because every recording is different. If you are doing the same kinds of recordings over-and-over, once you have an idea of what works with your set-up and your recordings, you can use that as your default compression set-up and make small changes when necessary.

Threshold? Noise Floor? Ratio? Attack & Delay times?

Do you understand what any of these these terms mean? I think you first need an understand of what they mean, and then you can experiment with the settings to see how they actually affect the sound. Maybe try over-doing the compression settings to the point where it sounds bad...

And for speech, you can try The Levelator (FREE!!!), which doesn't have any options.

Just tried a test of the Levelator, it appears to be EXACTLY what I needed. Does a better job of eliminating clipping and maximizing volume than either the manual edits I was doing, OR the Compressor. Many thanks for the suggestion!