Channel dynamics – Part 1 – Compression

February 6, 2009 at 9:36 am 3 comments

One of the coolest features of the new M7 has to be the ability to vary the dynamics of each input channel.  We’re all used to having EQ on each channel, but the M7 adds separate compression, expansion, and gating, just to name a few!  Separate compression on every channel – I can’t tell you how excited I am about that (seriously – I’ve whined more than once about only having six compressors at our disposal!).  These are phenomenal tools for correcting a number of common problems we’ve all run into.  But since I’ve seen them used somewhat incorrectly over the years, I’m going to give a quick refresher on their functions.

Compression and expansion get their names from the effects their applications have on the dynamic range of a signal (dynamic range is the difference between the loudest and quietest portions of a signal).  Compressors reduce (or compress) dynamic range, whereas expanders do the opposite (i.e., they expand dynamic range).

In this post, let’s look at compression, since it’s easiest to understand (imo).  If you take a signal whose level is slowly increasing, a compressor leaves the signal alone until it reaches a certain level, called a threshold (the setting of which is configurable by you).  Once the threshold is reached by the input signal, the output is reduced by a ratio (which is also user configurable) – anywhere from around 1.5:1 to 20:1 (or even greater).  For an n:1 ratio, it takes n dB of input signal over the threshold in order to achieve 1 dB of output signal over threshold.  Make sense?  A 10:1 ratio compresses the signal more than a 2:1 ratio.

Compressor Input/Output Characteristics

Compressor Input/Output Characteristics

Now, what happens as the level of the loud signal (which is over threshold) starts to decrease?  As long as the level remains above threshold, the compressor continues doing its thing.  But once the input signal falls below threshold, the compressor lets go, and the output level returns to non-compressed levels.

Simple enough, right?  Well, there are several more parameters worth discussing.  First, attack and release times refer to the speed at which the compressor starts and stops compressing once the signal passes through threshold.  For example, if you specify a 10ms attack time, then once the input signal crosses threshold, 10ms must elapse before the signal is compressed by the selected ratio.  Similarly, for a 100ms release time, it takes 100ms after the input signal drops below threshold for the compressor to stop compression and let the signal pass through, unmodified.

Since compressors work by reducing the output level of a signal after the input crosses threshold, they often include an output gain parameter that lets you boost the signal to recoup that loss.  At first, this may seem like you’re undoing the effect of the compressor, but the output gain boost is applied to the entire input signal, independent of whether it’s being compressed (i.e., independent of whether it has crossed threshold).  You’re still achieving a compressed dynamic range, but the entire signal is boosted; in fact, it behaves much like if you bumped the fader level on the mixer (incidentally, this is why compressed music often sounds louder, and is the source of the nasty “loudness wars” among record companies these days).

Lastly, I’ll mention hard and soft knee, since the M7 allows you to configure this (btw – the dbx 166 we have refers to soft knee as “OverEasy”).  The knee refers to the point in the figure where compression kicks in (i.e., where the signal crosses threshold).  So far, as I’ve described compression (and as is shown in the figure), we’ve been assuming a hard knee.  Soft knee compression causes the compression to take effect more gradually, usually by beginning compression at a very low ratio slightly before threshold and then gradually increasing the ratio up to its set point.  If you “smooth out” the knee in the figure and replace it with a gradual curve, you’ll see the effect of a soft knee.

Now, finally, why do we use compression?  While there are many uses, the two most common ones I’ve had over the years are vocalists and instrumentalists with large dynamic ranges.  Some of our vocalists (in fact, some of our best vocalists) occasionally get really into a song, and they get quite loud.  It may only happen for a few seconds, but when it happens, I wish I could have controlled it so the audience didn’t get blown away for that brief instant (and don’t assume you can “ride the fader” – your brain isn’t as fast as a compressor).  Or I’ve mixed for bass guitarists that start plucking with such fervor that they begin peaking on the console.  In both of these situations, I reach for a compressor.  For vocalists, I like to start with around a 2.5:1 ratio, middle-of-the-road attack and release times, and set the threshold so they are being compressed 6dB or so during their loudest passages.  And then listen – if they’re still blowing you out of the water, try bumping the ratio a bit or dropping the threshold slightly.  For the bass, I like a slightly larger ratio (maybe 4:1), but make sure you set the threshold so that compression only kicks in when they’re really plucking hard, otherwise they’re likely to complain and start turning themselves up even more.  Also, since the bass has a fairly quick attack, I like a faster attack time so the compressor is doing its job before you see red lights on the console.

OK, that’s it for now.  Next week, I’ll build on compression when we get into limiters, expanders, and gates.

Rob

Advertisements

Entry filed under: Audio.

Exposure, Focus & White Balance – Part 1 Concepts – Transitions

3 Comments Add your own

  • 1. angela  |  February 7, 2009 at 10:17 am

    Just a note on compression, – I am HUGE fan of compression and thanks Rob for pushing us towards using that board to its potential.

    Just an added note on why use a compressor v. riding a fader – not only is the compressor faster than you, but it’s (a lot of times) got better judgment. Something that might sound louder to you might not actually need to be compressed at the setting that we set for a particular vocalist or instrumentalist. So if you ride a fader as opposed to using a compressor, you run a risk. Compressor settings that have been done well can help all of us sound techs have a much more uniform sound as we mix. In other words, it doesn’t sound like “Rob” week or “Ryan” week or whoever – we’re more likely to sound consistent, which will help us to sound great.

  • 2. Oz  |  February 9, 2009 at 10:49 am

    From what I’ve read, consistency in sound goes pretty far in furthering the Kingdom. For example, if you’re overall volume fluctuates, then this week some people don’t like it loud and next week, some people don’t like it soft. Now, lots of people just think the sound is “bad” (completely forgetting the week it that they liked it!). No church can be be everything to everyone. Consistency in sound lets people find the church home that right for them. (And that’s not just in volume.)

  • 3. Oz  |  February 13, 2009 at 9:55 am

    I once had compression described to me like this:
    Say your pastor gets all riled up in a sermon and starts talking really loud, almost shouting. Compression makes it so that you hear all of his “energy” (dynamics) and you can tell he’s almost shouting, but you’re not blowing out the speakers and no one’s covering their ears cause he’s too loud.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

Trackback this post  |  Subscribe to the comments via RSS Feed


February 2009
M T W T F S S
« Jan   Mar »
 1
2345678
9101112131415
16171819202122
232425262728  

Flickr Photos


%d bloggers like this: