Compression is something every audio engineer will use but it’s also something that is widely misunderstood and easily abused!
It’s hard to pinpoint the very first use of a compressor for audio. “Communications”, Dec 1937 feature on broadcast limiters mentions only 3 units, WE 110-A, RCA 96-A, and Gates 17-B. Limiters were invented to prevent overmodulation of carrier waves by the audio signal and were mostly used for Radio and communications. It was sometime in the 1960’s when recording was booming that the modern day classics like the Teletronix LA2A were born. Universal Audio gives a very good description on how this legendary compressor was conceived and it’s certainly worth a read so please click here to check this story out.
In a nutshell Audio Compression or “dynamic compression” is a process that reduces the gain of an audio signal as it passes over a preset threshold at a rate set by the ratio. This is a somewhat simplified description of what is taking place, but it’s how it was originally described to me and it stuck! To get a bit more technical, when a signal enters the compressor it is split, one copy is sent to a variable gain amplifier and the other to a side-chain where a circuit controlled by the signal level applies the required gain to an amplifier stage.
There are a few different methods of achieving gain reduction. Some compressors rely solely on a vacuumed tubes, some use a photosensitive cell, (OPTO) and others use a Field Emitting Transistors or FET as the system to apply gain control.
Regardless of the method used to create this compression effect, they all set out to achieve the same result – gain reduction. An opto compressor will sound different to a FET compressor though and different engineers will favour a Valve, OPTO or a FET compressor depending on the job at hand.
Most compressors will have the same generic controls to manipulation how the compression effects the audio. These would be:
Think of Threshold as an imaginary line, (level in db). Any audio that goes over this line gets compressed and anything that stays below doesn’t – Simple! That line can be lowered or raised by turning the Threshold knob. The lower the line the more signal gets compressed. Some compressors will call threshold “input level”.
The Ratio is the amount of compression that will take place once the signal goes above the imaginary line. A 2:1 ratio is subtle and is good for vocals while a 20:1 ratio is dramatic and can be thought of as Limiting. Limiting is useful for Mix Busses and Mastering.
I’ll often use 2 compressors for a vocal recording running one compressor into the other. (Often know as serial compression). One compressor is set to a higher ratio and high threshold so that only a little bit of signal goes over the line. The second compressor will be running a much more gentle 2:1 or 4:1 ratio but with a much lower threshold allowing more signal to be compressed. You will ultimately get a very controlled upfront vocal without it pumping and sucking and disappearing into the mix.
The last 2 controls are attack and release. Not every compressor will have these controls but they a fun to use as you can really shape your signal. Attack is the speed, or time at which the compression will kick in, and the Release is the speed at which the compression will return to zero once the level has fallen below the threshold.
How to use attack and release? My favorite, go-to compression attack and release setting when processing drums is a very slow attack and very fast release. What this effectively does is allow the fast transient of the signal to get through before clamping the signal up really fast by the release. Your drums will pump and this is the desired effect. You can use this on acoustic drums, electronic drums, drum loops or individual hits to great effect.
A nice Vocal setting might be a relatively slow attack and a slow release. When limiting use a fast attack and a slow release. These settings are pretty generic but experiment and to see how these envelope controls can change the sound.
Lastly, many compressors will have an Output control or gain makeup control. The reason for this is because a compressor is effectively "reducing gain". The gain control will bring the compressed signal back to the same level as it was before gain reduction had taken place. A common error when using a compressor is having a high output gain control, so that when the compressor is engaged, it sounds louder than when it was bypassed. Many people will note how great the signal sounds even though no compression is taking place at all, the only thing that has happened is the signal is louder. This brings me to metering.
Most compressors will offer up some form of metering. Whether it’s via an LED or a VU meter you will get a clear idea of how much compression is taking place. If the LED or VU meter is not moving in “GR” gain reduction mode, no compression is taking place. The more it’s moving the more compression. If you’re not seeing any movement at all try dropping the threshold control.
As mentioned earlier; when a signal enters the compressor it is split, one copy is sent to a variable gain amplifier and the other to a side-chain where a circuit controlled by the signal level applies the required gain to an amplifier stage. A compressor with a side-chain input controls gain from the main input to output based on the level of the signal present at the side-chain input. By injecting an EQ at the side chain input set to a frequency you want to reduce the gain of, the compressor can act as a de-esser reducing the gain of that frequency.
Recently side chaining has been used for more creative purposes in Electronic Music. By feeding a bass line to the compressors inputs and a kick drum into the side chain input you can effectively punch holes in the bass signal whenever a kick drum triggers. The bass will seem to suck or pulse to the rhythm of the kick drum.
Something I hear a lot is, what compressor would be good for "insert - Genre specific music"? In my opinion there are no genre specific compressors. A compressor is a tool for a job, think of the job at hand then choose the compressor for that job. If you want a big fluffy vintage sound, go for a valve compressor, if you want a slow gentle vocal compressor try an OPTO model, if you want a fast, clean compressor try a FET one.
There are so many compressors on the market, all at drastically varying prices. Classic valve compressors like the Teletronix LA2A or the newish Retro STA are expensive because they use an assortment of rare and delicate valves to create their distinct sound and character. The Retro STA is a remake of that amazing 1956 Gates Sta-Level compressor and sounds fantastic. Retro also make a killer mono 500 series compressor called Doublewide. If you have a 500 series rack by API you could add this and get super high quality valve compression for a fraction of the cost of a stand-alone unit.
FET compressors are the most common and the most cost effective. The Universal Audio 1176 is a monster FET compressor and is great on drums due to its fast attack speed and unique characteristics. I also love this unit on vocals for extra punch.
Some compressors are great for a stereo bus. Most notably is the SSL G-Bus compressor. This unit is super fast and super clean.
Some of these compressors I’ve mentioned are quite expensive and this is due to their quality and heritage. You can get a great compressor without spending a fortune though. Brands like DBX, Golden Age and Warm Audio make very useful, low cost units. Warm audio have made a very realistic clone of the Classic UA 1176 called the WA76, and although you can’t press all the buttons in like on the original it’s pretty dam close sonically.
Hopefully this has helped demystify audio compression and helped guide you in your quest to use compression more effectively.
We are happy to answer any of your questions so please send us a message to firstname.lastname@example.org and we’ll make sure we get back to you.
Comments will be approved before showing up.