Mix Buss compression is usually done in the end of the mixing process at the stereo mix down. Some engineers do this, while others are not. The purpose of mix buss compression is to “glue” the entire mix elements. After the mix is cohesively balanced with all the EQ, effects and individual compression applied for each track, a mix buss compression can further improve the quality of the mix if it’s done right.
This is tutorial will use Rcomp Renaissance compressor in Waves. The software used is Adobe Audition although the technique can still be applied with any DAW that uses Rcomp. Below are some very important guidelines when you decide to apply a compression to your stereo mix buss:
1.) Less is more. Apply ONLY conservative amounts of compression. Remember that the objective of mix buss compression is to “glue” all elements in the mix together and not to make it loud. Making it loud is best left to the mastering engineer.
IMPORTANT: Compression reduces the dynamic range of the recording. I highly recommend to measure the dynamic range after mix buss compression to make sure the dynamic range would be preserved.
To get an idea how to do this, read this tutorial on how to make or create a radio friendly mix to learn how to measure the dynamic range.
As a rule, you need to aim for a dynamic range significantly greater than DR 13 after mix buss compression. Recommended values would be DR 15, DR16, etc. The higher the DR values, the better the dynamic range that is preserved.
It would be smart to measure the dynamic range before mix buss compression. If the dynamic range is already small; further manipulation of the dynamic range is best left to the mastering engineer.
2.) Aim for -1dBFS to -3dBFS reduction in gain (average RMS power) after compression. More than this amount, you increase the risk of over compressing your mix which is not appropriate for mastering.
3.) Apply a make up gain to bring back at the comparable level before compression. Make sure that after applying make-up gain the maximum peak of your mix down should not exceed -3dBFS. If its exceeds this amount or even more, applying further processing is best left to the mastering engineer.
Rcomp Settings for Mix Buss Compression
One of my personal favorite mix buss compressor is Rcomp (Renaissance compressor) from Waves. There are other more expensive compressors and you can still apply the mix buss compression principles as discussed in this tutorial. Below is the complete procedure how you can use Rcomp in stereo mix buss compression:
Step1.) Load up the high resolution stereo mix down in your DAW software (32-bit float/96KHz). For example this is the mix down without the mix buss compression:
Step2.) Get some statistics out of your mix. In Adobe Audition, you can go to “Analyze” – “Statistics”. Take note of the peak amplitude and average RMS. For example, this is the result:
RMS is a quantitative measure of loudness. If you are not familiar with this term, you can read this tutorial on understanding what does RMS stands for in audio.
According to the correct audio mixing levels and headroom in preparation for mastering, the recommended peak amplitude should be no greater than -3dBFS to give more headroom for mastering. The waveform statistics of the stereo mix down complies with these guidelines. So proceed to the next step.
Step3.) Select the entire wave (Control-A) and load the Rcomp plug-in from Waves. This is how the full reset/default settings would look like:
You need to deal with the following settings:
a.) Release mode – ARC or Manual (ARC stands for “Automatic Release”)
b.) Compression behavior – Electro or Opto
c.) Character – Warm or Smooth
d.) Compressor Threshold
e.) Compressor Attack
f.) Compressor Ratio
g.) Compressor Make up gain
Below are the recommended settings to start for Rcomp for mix bus compression:
a.) Release mode = ARC (automatic release is recommended because the song beats per minute can vary across the entire mix)
b.) Compression behavior = Electro (recommended with moderate compression applications such as in mix buss)
c.) Character = Smooth (recommended because you will need the sound to be as close as the original)
d.) Threshold = Average RMS power (higher figure in either LEFT or RIGHT channel) + 6dBFS
Since the mix down example has around -22.36dB (higher figure) average RMS power. The starting threshold is = -22.36dBFS + 6dBFS= -16.36 dBFS
e.) Compressor attack = 35ms (This is medium-set attack time. Small attack time can affect well-shaped transients that are important to the mix.)
f.) Compressor Ratio = 1.5:1
g.) Compressor make up gain = ~dBFS reduction after compression
h.) Release time (ms) = 30000/bpm
The tricky part is setting the release time. First, you need to read this tutorial on how to measure the song tempo in BPM. Take note that you will need to convert your mix down to an MP3 file first so that the BPM software can measure it. Using that tool, the BPM of the mix down is 120.02 beats per minute. So the starting release time of the compressor is:
Release time = 30,000/120.02 = 249ms ~ 250ms
Now press Control A, to select the entire waveform and implement the above compressor settings to Rcomp. Do not yet apply a make-up gain.
Step4.) Click “Preview” of Rcomp to listen to the results. And make sure that the attenuation in real time should not exceed to somewhere around -5dBFS (this is maximum not average) after song. Make sure most of the time it hits around -1dBFS to -3dBFS which should be the average/targeted dB reduction. See screenshot:
It shows that at one time the maximum attenuation is -5dBFS; however most of the time the attenuation meter is hitting somewhere around -1dBFS to -3dBFs which is OK. If it does not hit the target, lower down the threshold from -16.4dBFS to -15dBFS until the dB reduction shown in attenuation meter is acceptable. Use your ears to listen to any pumping presence, distortion, bad results, etc and tweak your attack/release time conservatively for best results.
Step5.) If everything is OK, apply a +2dB make up gain on the compressor. Listen to it again and finally hit OK to implement the settings.
Step6.) Confirm if the dB reduction, peak amplitude and average power SPL after compression is acceptable by getting the waveform statistics again. This is the result:
Although the statistics are not exactly the same as before compression, the subjective volume is somewhat comparable. Also the peak amplitude is less than -3dFS. This is the mix down after implementing the mix buss compression:
The volume is obviously low because this not yet mastered. Comparing before, the mix cohesiveness and punch has improved significantly as compared without a mix bus compression.
Content last updated on June 20, 2012