Tech Tutorial: Range Compression, Ups and Downs.
Oct 7, 2017 5:05:02 GMT -8
Brittany Ann Phillips, h0m3st4r, and 4 more like this
Post by duffyweber on Oct 7, 2017 5:05:02 GMT -8
Tech Tutorial: Range compression and mitigating its effects on low volume areas/sentence ends
Compression! There's a word you've heard and may quite possibly be terrified of.
It's got its ups and downs. And brings them closer together. Heh. ; )
Seriously though: Compression is short for "Dynamic range compression" and what it basically does is it ensures all sound ranges of your audio recording fall within a certain volume level. 1500Hz range too loud? It drops it down a bit. 600Hz range too low? It's gonna beef it up a bit, due to the volume levels being closer together. With caveats.
It makes the sound "richer" because it squeezes it all into a certain volume range where all the ranges are more equally audible, without making them entirely uniform volume (which would sound terrible. Imagine if the 80Hz and 2000Hz ranges where all cranked up to the same level as your voice. The wave would be one big square block. "BWWWWWWWOOOOOOOOMMMMMMM.ZZZZZZZZZZZZZZZTTT. FWEEEEEEEEEEEEEEEEEEEPT!")
Note: The above is sort of like a PBS science-show explaining nuclear physics. It's a rather stripped-down and simplified explanation, but it makes the premise easier to understand. And it's enough to go forward on.
So basically, running a light compression on your audio will give more "presence" to your voice and recording.
That said, sometimes there are issues.
Most commonly, the compressor tool will amplify sounds you don't WANT aplified. All of a sudden there's swoops and swooshes, and bits of noise floor and paper shuffling and clothing scrapes you didn't know were in there before the compressor was run, and these are usually unwelcome.
Proper mic/booth technique and comportment will help you here. Also, background noise can be identified and eliminated this way. That will help you tweak your recording environment for better overall acoustics. There's not a lot in the DAW* you can do to help, but if you want to copy some small parts of the audio BEFORE applying compression and pasting them in afterwards, and working to richen them up individually, you can, but that's an advanced and time-consuming technique, and it would be much, much better AND faster to simply re-record the line.
*Your recording software.
There's also another problem: quieter sounds, whispers, and some other parts of the human voice run afoul of dynamic range compression and there's a little trick that I, personally, like to use to get around it.
You'll notice this problem in that it sometimes makes the ends of your sentences trail off. Quieter parts, whispers, sibilants, and consonants like F's suffer because they're faded and warped sounding.
Here's a really quick and dirty way to overcome this, and still get some of the richness compression can offer, if you want to put your best foot forward for an audition.
(Since most of you seem to use Audacity, we'll go over the process in that DAW and you can scale it to the process for your own DAW.)
First Step: Clean and prep your audio. Get the final, noise-reduced, trimmed track.
Second step: Normalize the track to -0.1dB
Third Step: Clone the track. Start by selecting the whole track. CTRL+a will do it if you're only using one track. HOME, then SHIFT+End will do it if you're using multple tracks and only want to select the one. Then, press CTRL+SHIFT+M (which actually mixes and renders multiple tracks to a new one, but if you've only selected one track, it has the effect of cloning the track.)
NOTE: Cloning is better than copy/pasting to a new track. IF you copy/paste, you run the risk of adding/removing a leading space and these tracks CANNOT BE OUT OF SYNC. Not even by 1/1000th of a second
FOURTH STEP: Apply dynamic range compression to one of the tracks. EFFECT > Compressor. (My usual values are Threshold -2dB, Noise Floor -40, Ratio 2:1, Attack 0.2, decay 1.0. In other words, the defaults.) ; )
Fifth step: Normalize the compressed track to -0.1dB
Sixth Step: Select both tracks and go to "Tracks > Mix and render"
Seventh Step: WOW! That looks loud! Let's normalize it to -2dB
THERE! You have audio compression mixed into your track, richening it up a bit, but you also have the ORIGINAL track mixed in, leaving you with the parts that compression seems to cut out. LISTEN CAREFULLY to make sure the audio sound right and that you haven't amplified unwanted sounds, and that everything sounds like it should.
This particular "trick" isn't what you'd call proper engineering, and it halves the effect of the compressor on the entire wave, but you get a lot of the benefits without anything suffering. It's rather a "compromise" technique. And one that serves me well from time to time.
Happy recording!