Compression, I guess I've yet to understand it...

[ajcarson1]

Hey everyone, back to question the world of recording...

I thought I knew about compression and its uses. I've always considered compression to bring the highs down- and bring the lows up. To almost "smash" the sound into a square-looking waveform. Is this not correct?

The reason I'm asking, is because I hear that a lot of people use compression for various reasons, and I've yet to use it much at all in my mixes. What can I use compression for? What Is it commonly used for? When should I avoid using it?

Things like that.
I would greatly appreciate any responses, as I'm really very new to a lot of the lingo around here, and it'd be nice to get things cleared up on a compression note.

Thanks!

[Bigduggieface]

Hoo boy, this is a big can-o'-worms.

Compression is one of the engineer's most useful tools. It is used to control the dynamic range of a particular audio track, and often across the entire mix.

It is most commonly used to help to automatically level out the differences in dynamics. For instance, a singer's performance might have some words that are quiet, and other words are SUPERLOUD and tear your head off. You can set the compressor to clamp down on these loud passages, thereby making the quieter words as easily discerned as the loud ones, and providing a track that sits solidly and clearly in the mix.

You can also use your compressor as a distinct audio effect. A good example are many modern rock and country drums. Run the kick and snare through compressors and (if it is set correctly) you can get those larger-than-life, thick, phat drums that we all love so much.

Many mastering engineers use compression on entire mixes as well. Most often, they are set with a very high ratio (compressors set at very high ratios are known as limiters) in order to really clamp down on any very brief transients. You can get a track to sound much louder, and yet still keep the track sounding natural.

I'm sure we'll have many forum members chiming in with more details on this subject, because it is a vast one.

[ajcarson1]

Nicely said bigduggie, I understand quite a bit more, just from your post alone.

Why would placing a compressor on the snare make it more "thick" though? wouldn't it just make the snare hits more evenly loud?

[Bigduggieface]

Originally Posted by ajcarson1

Why would placing a compressor on the snare make it more "thick" though? wouldn't it just make the snare hits more evenly loud?

A good observation.

Drum hits have very short but dramatic lives. When you hit a snare drum hard, you get an immediate and profound spike in level, and it decays relatively quickly. If viewed as a waveform, you'll see the almost straight vertical initial transient, and it dies away rather quickly.

If one sets up the attack and release controls on your compressor correctly, it will soften the initial transient, and emphasize the decay of the drum.

Like you said in your initial post, it makes the loud part quieter, and the quieter parts louder. Only all of this happens within a few milliseconds with a drum hit.

Your ears don't react very quickly to sharp, short transients, so when you use the compressor to squash the first couple of milliseconds, but leave the remaining few milliseconds uncompressed, your brain is tricked into hearing a thicker "thud" rather than a thin "snap".

A classic example is AC/DC's "Back In Black" snare. That's a "thud" if I've ever heard one. It's not an unprocessed jazz recording. It's a compressed one.

You are correct that it will even out the snare hits over the course of the song, but I'm talking about using the compressor at VERY short attack and release times.

If compressing a voice, or a bass guitar, the transients and decay rates are usually much slower, so you would set your compressor's attack and decay rates to react slower.

[cakewalkgg]

Im not an expert but Ive found it useful just to loop a section of snares and really spend time with the compressor Im working with. Its pretty easy to understand the threshold and ratio part.. at xdb the volume (above threshold) will be reduced by y ratio

The musical part is the attack and release.. and thats harder to get a sandard understanding of.. in my opinion that where you have to solo the track loop a sample of what your trying to compress and really listen carefully to what its doing.. especially with something as sharp as a snare... you can really dramatically change the sound with the attack/release parameters...

Im sure with lots of experience one would pretty much know what values they are likey to use and would do it quickly.. but for me, I still have to spend some time really listening closely when Im setting up the compressors

[ajcarson1]

Thanks for the responses guys, I'm learning quite a bit now about compressing.

So I've heard you can "hear" when a track is compressed too much. Can anyone provide samples of a "too much" compressed track, vs a track that isn't compressed (as much)?

I wonder if I can spot the difference.

[BillyShears]

For a practical exercise in what a compressor does try this link. It is someones take on an article written in a Australian mag called Audio Technology which is easily the best article on compression I've seen. This link pretty much covers what was said in the original article.

The best way to try this is to have a drum beat going and send only the snare to the compressor as this will make it obvious what is going on. Basically just set your compressor to the suggested initial settings. What this does is try to squash the body of the snare so the attack will stand out more. After the attack you set the release, ratio and then threshold. One thing I noticed is that this worked better with an external compressor than a plug in but that would probably be due to me not the software.

Have a play and I guarantee you will hear what a compressor does with this method.

www.Goingto11.com » Compression: Yes, You Can!

[ajcarson1]

Hey thanks for your information Billy, that link worked wonders for my understanding of compression.
So from what i've gathered so far (correct me if im wrong):

1. Attack - How quickly the compressor kicks in when it reaches something beyond the chosen threshold.
2. Release - How quickly the compressor stops compressing when the wavelength goes back above the threshold
3. Ratio - presented in a number such as 3:1 (3 db of compression for every 1 db over the limit) - Ratio determines how much compression is being applied and thus how "big" the waveform will look after it has been compressed.
4. Threshold - determines at what level to kick the compressor on. If the threshold is set to -10DB, anything louder than -10DB would get compressed.

Is this all correct?

Can someone explain release times? what is so important about them?

[sparqee]

Originally Posted by ajcarson1

Hey thanks for your information Billy, that link worked wonders for my understanding of compression.
So from what i've gathered so far (correct me if im wrong):

1. Attack - How quickly the compressor kicks in when it reaches something beyond the chosen threshold.
2. Release - How quickly the compressor stops compressing when the wavelength goes back above the threshold
3. Ratio - presented in a number such as 3:1 (3 db of compression for every 1 db over the limit) - Ratio determines how much compression is being applied and thus how "big" the waveform will look after it has been compressed.
4. Threshold - determines at what level to kick the compressor on. If the threshold is set to -10DB, anything louder than -10DB would get compressed.

Is this all correct?

Can someone explain release times? what is so important about them?

If your compressor has a soft knee then compression will begin *before* the threshold is hit. It's confusing which is why compression is such a tough art to master.

Attack, threshold, ratio (and other parameters) can be very important but it's the release that can really have a very noticeable affect on a rhythmic track. Tweaking release will give you control over "pumping". Sometimes you want to avoid pumping. Sometimes you want to use it. Tweak the release to sync up the pumping of your drum bus compressor with the tempo of your song. Adjust the release of your master bus comp to suit the rhythmic elements (and reduce noticeable pumping). On something like a snare or kick it's a good idea to make sure your comp is releasing fully (i.e. gain reduction is back to zero) before the next hit. Otherwise the first hit will receive less compression than successive hits as demonstrated in this chart:


Check out: Dyn-Tech_Detail for some great graphs and info about compression and how it affects the envelope of sounds.

[Vulconizer]

I'm liking this thread. There have always been a few murky areas in my understanding of compression and this is helping.

This site rocks, as usual.