If you want to play any game at a higher level, you have to know your field. There is nothing quite like the feeling of knowing exactly what to do, what to reach for. But that also goes in the other direction as well - it feels bad to be fumbling around and second guessing your decisions. For example, what's a VCA buss compressor and do I need one?
A big part of work in the field of audio engineering is picking the tools that you use to shape your sounds. And the master bus compressor is arguably the tool that leaves the largest imprint on everything as it works on the master bus, which constitutes the entirety of the song.
Compressing the entire master output of your mixing desk or DAW is an age-old trick of mixing engineers to achieve a more compact and lively sound. Because a compressor reacts to loud sounds, it will be triggered by different elements of the mix throughout the song, creating movement that follows the highlighted element at each moment. It's like some sort of a gentle "auto-mix" that follows the music and adapts to it. Cool huh?
The "SSL 4000 bus compressor" and various clones and derivatives are still by far the most popular devices to get this done, but there are so many to choose from! So let's up our game together and get to know as many of them as possible and shoot them out against each other. To be sure we are comparing apples to apples, we will use the same material and carefully calibrate the levels and other circumstances to get a fair and unbiased result.
About the Authors
This shootout is brought to you by team mix:analog and if you know any of our previous work, you will know we are a bunch of audio nerds that don't settle for "good enough". We are the maniacs who try 20 different incarnations of the same design, manufacturer, make, model, age or any other parameter just to find what gives a good result and how it gets there.
And with this shootout we did exactly that! But before we start listening to the processed results, let's first spare a few words for the methods used. As proper audio nerds, we like scientific approaches: developing calibration and testing procedures, level-matching, that kind of stuff.
The VCA method
Because VCA style compressors are known for their low-end punch and bringing a kind of cohesiveness or "glue" to the mix, we chose a rock song to test it out; it's called "Gravediggers" by Hollowstate - available HERE for educational purposes. It conveniently has all the elements to expose the different aspects of each compressor - lots of dynamic low frequency information, busy stereo midrange and a lead vocal in front and centre.
Different variations of the VCA bus compressor, software and hardware alike, have slightly different internal levels and settings at the same knob positions. The attack and release times are not continuously variable but discrete on most units and thus easier to reproduce, but threshold and makeup gain require a bit more attention to set the same on different units. So instead of going for the setup by the same dials (or numbers), we want to get into the same "zone" of the compression curve of each unit, regardless of the knob positions.
Why a weird calibration signal?
Using a 1kHz sine wave as a calibration source would be tempting, but we have probably all experienced pretty obvious differences between different compressors' response on the low-end. And then there's the sibilance that some compressors bring out more than others. So to take all that into account, the test signal used to calibrate the units before passing the mix, is a sum of three -20dB tone generators' outputs at 80Hz, 1kHz and 8kHz. Now that's a better approximation for a kick drum, mid range and sibilance!
That signal was used to first set unity gain without compression (Yes, some of the units add some gain without any apparent processing. Naughty.)
Next step was setting the threshold to obtain a level drop of 4dB on an external meter, not on the GUI, at the settings of 4:1 ratio, no side-chain filter, 3ms attack and 100ms release, and then using the make up gain to get back to the original level.
In this way we achieved two goals of fair comparison - we are probably at a very similar point on the compression curve and we are level-matched! If the software unit had any "analog modelling" features, they were turned on. For the hardware units, the ADDA converters were calibrated to 0dBFS=+18dBu, which for most hardware units meant operating the threshold at around noon, not starving nor pushing the input and output stages.
Even if most of the units set out to do more or less the same thing, the differences turned out to be quite dramatic in a few cases! Some react to the low-end like a bomb-drop, others fare smoothly. Some are clean and transparent, some are plain boring, and then some have some serious bite. Some make the mix flat, others pull out the detail and "3D space image" that you never knew existed in the mix!
But all in all, the point is not to proclaim a winner and call it a day. The most valuable lesson, at least for me, was to get to know them all and I now have a better idea of which one to use to get an aggressive sound, which one for transparent and deep, and so on. Give the samples a listen on a good pair of speakers and broaden your own sonic palette as well!
The results are available HERE!
The list will be updated with more samples if and when we find other units of that sort and owners willing to participate.
To date, we tested those units (in no particular order):
- Mix:analog GoldCan VCA
- Native Instruments Solid Bus Comp
- Acustica Audio TAN
- Acustica Audio Sand
- Slate Digital FG-Gray
- Waves G-Master Buss Compressor
- Ableton Glue
- IK Multimedia Bus Compressor
- Cytomic Glue
- Brainworx bx_townhouse
- AirWindows Logical4
- UAD SSL G Bus Compressor (Legacy)
- UAD SSL 4000 G
- Sonimus Satson
- DMG TrackComp VCA
- Apple Logic Pro X Vintage VCA
- PSP BussPressor
- GSSL Clone /w Turbo, single THAT2180 VCA per main channel GR element
- Klanghelm DC8C3
- SSL 9000J Master Bus Compressor @Studio RSL
- SSL 1U G Series Bus Compressor
- SSL 4000G Master Bus Compressor @Embrio Studio