Well, this week let's try to figure out in working-mans-terms just what the deal is with compression.
The thing to remember about an audio signal is that there are three ways to affect it:
- Change the spectral content (frequency)
- Change the time relative to other sounds (phase)
- Change the amplitude (volume)
You change the spectral content with an equalizer, and the phase with delay lines. A compressor goes after the amplitude.
So, why would you want to affect the volume other than turning things up and down? You have to consider the dynamic range of a sound, and how well it works with what's going on around it.
OK...so what is dynamic range? Simple...it's the difference between the softest sound and the loudest sound, whether you're talking about a recording or a live performance. The greater that difference, then the greater the dynamic range.
So, more dynamic range is better, right? Well, if you're in a perfect listening environment with an orchestra, then yes. But as soon as you introduce electronics into the picture, you have a limiting factor...and that's the signal-to-noise (or s/n) ratio of the equipment. If the dynamic range of the program material is greater than the signal-to-noise ratio of the equipment, then either one or both of these things happen:
- The really quiet parts get lost in the background noise of the electronics
- The really loud parts exceed the available power and distort
Neither is good.
So, enter the compressor. On a very basic level, a compressor turns up the really quiet parts and turns down the really loud parts. It just happens to be able to do it really fast. So what you wind up with is less of a difference between the softest and loudest sounds....or a reduced dynamic range.
Compressors started out being used ahead of the cutting lathes when record masters were being cut (remember those round black things with a hole in the middle). If the levels were too low...the sound got lost in the inherent "hiss" of the vinyl. Too hot of a signal, and the lathe could overcut and get into and adjacent groove.
Next, they found their way into the recording process to optimize levels to tape. Then, radio stations started using them to make themselves seem louder (we'll talk about how that works in a second). Now, they're used all the way through the recording process to make the records sound louder (like the radio stations did), and they're now all over live sound rigs (so they sound like the record...and louder!).
So...why does a compressor make things sound louder? There are a lot of psychoacoustics at work here (how your brain processes sound), as well as electrical reasons.
The electrical reasons are fairly straight forward. Since you've reduced the loudest parts of a signal...you can now turn things up more without worrying about overloading something. So, if you look at the average of the energy over time, you're able to raise that average. And our brain interprets that higher average as an overall increase in loudness even though the absolute sound pressure level at the speakers did not increase.
Of course...it's possible to take this too far. If you suck all of the dynamic range out of the sound, it actually starts to fatigue you. Case in point, there have been several albums come out that were over compressed to the point of making them unlistenable. I love Rush, but Vapor Trails sounds like it was mastered with a "loudness at all costs" mentality. (A remaster of this would be very welcome). Remember Los Lonely Boys? Their first album was very open and dynamic. The second album had the life choked out of it at mastering...and the sales were dismal.
Yeah, compression is definitely one of those things that a little goes a long way. It helps individual instruments "sit" in a mix better.
When you're dealing with compressors (and limiters...we'll get to them next), you're dealing with four major parameters:
- Threshold: The level at which the compressor starts working
- Ratio: Basically the rate at which the output changes in relationship to the input.
- Attack Time: How fast the compressor reacts to a signal above the threshold
- Release (or Decay) Time: How fast the signal is allowed to come out of compression
The threshold is just what it sounds like...at what point does the compressor start turning things down. The lower the threshold, the more you're affecting the signal. This illustrates what the threshold and ratio are doing to the levels:
The compression ratio determines how tightly you're controlling the sound. The higher the ratio, the more compression you're applying. If you have a 2:1 ratio, then for every 2dB the input changes, the output changes 1dB. Likewise, and 10:1 ratio would require an 10dB input change to move the output 1dB. You can also think of this control as how much dynamic range you're removing.
When you get above a 10:1 ratio, you've now crossed into the world of limiting. You can almost think of a limiter as a wall. In other words, once you've hit the threshold point, you're not going to get any louder.
So, where is a limiter useful? Easy...that's what you use to protect amps and speakers. Whether you've got a $100 PA cabinet or a $10,000 one...there's a finite amount of energy that you can pump into them before you damage 'em. Limiters are useful for when unexpected things happen...like someone drops a live mic (boom!!!).
The attack and release times...you're actually affecting how fast things are getting turned up and down. How to set these is an installment on it's own, but in general slow attack and release times are less audibly intrusive, while faster times give you tighter control and a denser sound. But, take it too far and you start getting artifacts like pumping and breathing (listen to any Steve Miller album from the 70's for an example of this). A lot of modern compressors have an "auto" button that takes care of the attack and release times based on the signal coming in. I figure most of the time that button is smarter than me so I'll just go with that...especially on vocals.
There's usually a meter on a compressor marked "G/R"....that's "gain reduction". In other words, how much is the signal getting turned down. For me, that's what I go by. I don't get hung up on the absolute value of the threshold, because it's going to depend on your signal level coming in. I'll just turn the threshold until I see the amount of gain reduction I want (and that depends on what I'm compressing).
You'll usually see a control marked either "output", "gain", or "gain make up". As you apply compression, you're turning things down. This control lets you put the output back to where it would be without the compressor. To set this, bypass the compressor and listen (or watch the level meters) to see where the uncompressed signal level is. Then, engage the compressor and adjust the output until it matches what you had before. This way, if you need to disengage the compressor you're not going to have a radical change in levels. It also helps keep your gain structure correct.
That's a lot of tech for today. We may come back to compressors at a later date and talk about specific applications on specific instruments.
Next week...I thought we would get away from the "tech" and get back more into "show". We'll talk about set lists. We'll discuss different philosophies on making and managing them.
Until then, use compression and limiting to keep the meters out of the red!
Ken
Ken Carver has been a musician and performer since the early 70's, and involved with live music production since the mid 70's. He worked for 15 years as a broadcast engineer, building numerous studios and transmitter sites around Texas. He's also worked in Critical Care Communications for the medical industry, R&D for an automated lighting manufacturer, and owned Project Lighting & Sound in the 80's. He currently heads up an R&D Hardware Technician Team at National Instruments in Austin, and still performs on the weekends in the Central Texas area. You can reach Ken at itsjustlogistics@gmail.com
No comments:
Post a Comment