We all have a basic idea of what a compressor does and how to use it, right?
I’ve googled „what does a compressor do?“ and the „top“-results are pretty much all similar but still all wrong.
Something along these lines:
„Compression controls the dynamics of a sound, it raises low volumes, and lowers high volumes“
If you’ve read a few of my articles, you know that I’m questioning some things – and if necessary hit you with raw data to destroy popular myths 😉
Check our FREE Compression Harmonics Test
A free product that includes the Compression / Harmonics DAW-Session from YOUR MIX SUCKS for 5 DAWs: Logic Pro X, Pro Tools 2018, Cubase 10 Studio One 4 and Ableton Live 10.
TEST SETUP & PROCEDURE
Let’s run some popular compressor plug-ins through a test setup and procedure, then look at what the results tell us!
(BTW, I’ll give you the test setup as a Logic-Template for download so you can test your own plug-ins – just subscribe to the newsletter!)
The Test Oscillator in Logic Pro X feeds a compressor with a test-tone:
• The test tone is a sine-wave (as you know, a sine-wave has no added harmonics).
• We cycle through 55 Hz, 110 Hz, 220 Hz, 440 Hz tones
• then a sweep from 20 Hz to 20.000 Hz
• ending the cycle with a 100 Hz tone
We cycle through this 3 times, with rising levels:
• 1st Cycle:
Oscillator hits Compressor with – 18 dB of level
Compressor Threshold is set JUST BEFORE compression
for the compressor NOT to compress (unity gain)
• 2nd Cycle
Oscillator hits Compressor with – 12dB of level (6 dB more than on the previous cycle)
Compressor settings stay the same, but of course compression now kicks in!!
• 3rd Cycle
Oscillator hits Compressor with – 2 dB of level (another 10dB added on top of the previous cycle)
Compressors settings remain the same, but now hitting compression quite hard!
The upper track you can see in the videos is the automation curve for the Test Oscillator’s frequencies and levels, the lower track shows a huge analyzer plug-in after the output of the compressor (using Logic Pro X’s Channel EQ), and that shows as the frequency spectrum in realtime.
BTW: – 18 dB in your software is a GREAT average level for your recordings and signals in ALL situations. It assures clean and pristine sound and compatibility with all plug-ins. More on this another time when I hit your head with a bat called GAIN STAGING.
Summary: on the first cycle, the compressor doesn’t actually change the level of the signal, on the 2nd cycle there is some compression, and on the 3rd cycle: a lot.
Every cycle ends with a 100 Hz tone – that makes it easy to read the added harmonics on the analyzer.
• 2nd harmonic = 200 Hz
• 3rd harmonic = 300 Hz
• 4th harmonic = 400 Hz
• 5th harmonic = 500 Hz
• n th harmonic = 100 Hz x n
Just for reference, here’s what the test procedure looks like with NO COMPRESSOR inserted in the signal path.
As you can see, the analyzer just shows basic sine-tones, with no added harmonics. Music theory and physics calls this is the 1st “harmonic” – but don’t be confused, that is the term for the original frequency of the sine-tone.
Fairchild 670 Compressor (1959) – The Royal Harmonics Orchestra
To give you a proper contrast – here’s what this looks like with a plug-in clone of the legendary Fairchild 670 compressor, as some of you might know, the most expensive and sought after vintage tube compressor on the market.
I bet you that the designers of this plug-in looked at a spectrum like that forever, and did endless coding and testing until the plug-in matched the original hardware closely.
You can already see some harmonics even when the compressor doesn’t compress, but they really kick in the more you compress.
Note: the lower the sine-tone, the more harmonics show up – I can count 13 added harmonics in the 3rd cycle on top of the 50 Hz sine tone. When the oscillator reaches 2000 Hz, the Fairchild doesn’t add more harmonics on top, at least not visible anymore in the spectrum.
If you look at the rich harmonics added by the Fairchild, you start understanding how it gives a dull bass-sound or a 808 subsonic kick a richer frequency spectrum.
This is very useful as it helps low-end to subsonic sounds to translate better on smaller systems (think laptop, tablet, smartphone, kitchen radio).
On the other hand, a Fairchild might not be the ideal compressor to purely control volume, because the more you compresses, the more it changes the sound of the source. This is not typically what you need if your goal is to level something that is dynamically uneven.
On the contrary, you want to make sure that your source is already in control dynamically BEFORE you even hit the Fairchild.
There are other ways to achieve a consistent loudness in a performance.
Look at waveforms of your recording and just bring the lower parts up in level, reduce loud sections, automate, and a lot of times, a consistent level is what a great performer brings to the table!
If your signal is well-levelled and even, you can alter its tone by the amount of drive the signal into compression.
I’m almost ready to go into detail on parallel compression at this point – imagine a setup where you’re bring the same low 808 Kick into two mixer channels.
The first channel kept unprocessed, the second channel pushed hard into a tube compressor like the Fairchild – the added tube channel will make the 808 come through on smaller systems and adds a nice texture to a sound thats pretty close to a sine-tone.
On the parallel channel you can even cut off the low end and just add the harmonics (of course cut off POST compression) – more about that in Part 2 of this article.
Essentially, what the added harmonics do is adding frequencies to the original sound that weren’t there before.
PUT THE COMPRESSOR IN THE MIDDLE OF THE SIGNAL-CHAIN
A typical signal chain around a compressor like the Fairchild would look like this:
Source Signal
-> Surgical EQ (to remove unwanted frequencies)
-> Compressor (adding harmonics)
-> EQ (for color and tonal balance)
PRE COMPRESSION: Surgical EQ – to remove unwanted frequencies
It’s very important to put an EQ BEFORE the compressor. Use this EQ to remove unwanted frequencies. Typical example would be a High Pass Filter that removes rumbling impact noise on vocal recordings. Imagine how the Compressor would add harmonics to a rumbling noise at 30 Hz and really bring it out – you don’t want that.
Same goes for unpleasant room resonances – find them using a narrow EQ boost then set a small notch to remove them.
This so-called „surgical EQing“ works best with „Linear Phase EQs“ – many plug-in manufacturers make them, they don’t add coloring resonances. I like the one that comes with Logic Pro X a lot.
POST COMPRESSION: EQ for color and tonal balance
Going back to the example of a low 808 subkick, which is a sound that is very similar to the sine-tone used in our test, there would be no point in EQing a pure sine-tone, right?
You can’t add frequencies that are not there – in contrast to a tube compressor, an EQ does NOT generate any frequencies, it can only adjust the tonal balance of the given frequency-content.
With that, we have once again turned common audio-knowledge upside down:
• Compressors can add color to any frequency
• EQs are static, all they do is adjust the tonal volume
That rule is of course not totally holding its own once we look at a few more types of compressors. What we’re interpreting in this article focusses on frequencies and harmonics, which is just one aspect of compressors. The other one is the actual ability of a compressor to level, limit or „grab“ a signal, attributes that all refer to the volume of the sound.
Universal Audio/UREI LA-3A (1969)
The Waves CLA-3A is a plug-in clone of the original Universal Audio LA-3A Compressor/Limiter.
In contrast to the Fairchild, it’s a lot better suited for levelling a signal.
The LA-3A adds only one harmonic (the 3rd one). The Fairchild and the LA-3A can co-exist in a signal chain. Use the LA-3A to even out levels, then hit the Fairchild.
The LA-3A is typically used as leveller for bass, guitars and even vocals.
Less suitable for percussive sounds – it’s not following fast enough to control a drum sound.
Teletronix LA-2A (1965)
The Waves CLA-2A is a clone of the Teletronix LA-2A. The design is a few years older than the LA-3A.
It adds more harmonics than the LA-3A, but still a lot less than the Fairchild.
Typically used to control bass, backing vocals or laid-back lead vocals. A fairly slow and laid-back tube compressor.
Universal Audio 1176 Rev. A Blue Stripe (1967)
The Waves CLA-76 is a clone of the Universal Audio/UREI 1176.
Various revisions were made, the „blue stripe“-version being the first one ever created.
Nearly every plug-in manufacturer offers a clone of the 1176. I like the ones made by Waves, and they got their name because Waves developed them with Mr. CLA aka Chris Lord-Alge.
The 1176 displays extremely rich harmonics. In comparison to the Fairchild 670, it sounds a lot more agressive and levels superfast.
That makes the 1176 very flexible – it can be used on almost any source.
Like the Fairchild, the 1176 is a true studio classic and it would be worth writing a dedicated article about.
If you have a bunch of those in your rack (or a great plug-in clone in your collection), you could mix an entire project exclusively with those.
One of the things it works very well for is making vocals agressive and bring them upfront.
You can drive it hard into a lot of compression, and as a result will see a lot of harmonics.
It would not be the only compressor in the chain, I’m usually running another compressor for levelling before the 1176.
Solid State Logic SSL E/G-Series Bus Compressor (1977)
This is of course one of the most famous compressors ever build.
Waves teamed up with SSL to create one of the first true emulations of an original hardware, and this plug-in (as part of the Waves SSL-bundle) is now a classic, just like the original SSL 4000E and G-Series consoles.
It does – of course – a great job levelling a signal, and adds more harmonics the harder you hit it.
The trick with the SSL Bus Compressor is hit compression with your peaks in your finished mix, e.g. the kick drum.
What happens is that the SSL „grabs“ and reduces the peaks in a very clever way, while adding harmonics to them.
The SSL bus compressor controls the dynamics and makes the bits it compresses more punchy by enriching it with harmonics (almost like compensating for the lost level). This effect has widely been described as mixbus-“glue” and the reason why everybody loves SSL Bus Compressors.
Solid State Logic SSL E-Series Channel Compressor (1977)
The SSL Channel Compressor, also a part of the Waves SSL-bundle, adds a healthy portion of bright and agressive harmonics, and is capable of controlling and “grabbing” percussive signals like no other compressor.
Widely used by famous mix engineers on Kick, Snare and any type of percussion – the SSL gives drum sounds a prominent place in the mix, makes drums punchy and cut through.
Summit Audio TLA-100A (1984)
The Summit TLA-100A is a very subtle tube compressor. The original analogue hardware has been used by engineer Al Schmidt as a tracking compressor on many of his Grammy-winning projects, for example on Diana Krall’s vocals, to catch some peaks with light compression during vocal recording.
The Summit adds some harmonics when you drive it, and works well on a wide selection of sources with 3 easy settings for both attack and release-time.
A very subtle leveller for tracking and mixing.
Logic Pro X Compressor (Platinum Mode, 1996)
This is the original compressor plug-in that came with the first version of Logic, so the design goes back to the mid-90s.
It’s actually the ONLY plug-in in our test that does not colour the signal AT ALL. Contrary to all other compressors tested, this is a compressor suitable for applications where you want to iron out the dynamics of a track without adding harmonics.
The later versions of this plug-ins (like the current one in Logic Pro X) added a few more modes you can select, and when you switch from the “Platinum Mode” in any of the others (like “VCA”), the plug-in starts adding harmonics.
[convertkit form=5236308]
