The mix is close to being finished and it’s time to add the final touches. Time then to talk mix-buss or mix-buss processing – the plugins that are often used to finish off your mix.
The overall approaches to the master fader are of course varied, but typically most mix engineers use plugins sparingly and with the intention of leaving room for the mastering stage of processing. There are no real rules in terms of order but one sensible approach is to start with a little saturation or analogue emulation.
There are several popular choices for this type of effect now, with Slate Digital VCC or McDSP Analogue Channel being two of the earliest examples. These types of processors add a subtle character to the mix that suits many modern styles of music. You might notice a slight thickness due to the effects of a specially tuned light distortion and compression. The idea here is to generally add a light touch, nothing too extreme. One thing you might look for is a little crosstalk and individuality with regard to dynamics between the channels (L and R) to simulate a real mixing desk or tape machine.
Secondly, we might add a compressor. Some reckon this is best left for mastering and some argue that mix compressors often feature on mixing desks – so why not use them? If you fall into the ‘compress’ camp, then there are several great options to choose from. One of my personal favourites is the SSL style VCA compressor set to a ratio of 2:1, with the slowest attack (30ms) and the fastest release (100ms). Then dial in the threshold so that you are only getting a couple of dB gain reduction in the chorus (2-4dB would give you a moderate effect).
This works for many styles of music however the use is obviously tied to how much compression has already been applied to the individual tracks and even to any samples / instruments used on those tracks. The key to a good mix compressor is a low ratio, this ensures that the effects aren’t too extreme and louder elements aren’t likely to influence the compressor as much. When to add the compressor is another interesting choice. It adds an interesting dynamic to the mix (pun intended) so that when you turn a channel up, the compressor may indeed turn it back down. Therefore, if unfamiliar with this phenomenon, then adding the compressor near the end of mix completion is probably the right answer. The mix should already sound well balanced within itself and the frequency spectrum should be nice and smooth.
Now, should you use EQ? This one is also a little contentious in some circles. The mix stage is the last time you will have true freedom to fix any frequency issues at the source, so if you feel the mix is a little dull, boost the high frequency tracks in level (hi-hat, cymbals, percussion etc.) and add a little treble to them as well if needed. This is a good time to calibrate your ears to some commercially released material – listen to the brightness of the hi-hat on something popular and use that as a gauge. If you do decide on EQ many engineers opt for a high quality linear phase EQ processor similarly to why you might use one in the mastering stage – but equally there are engineers that opt for a colourful EQ like the Pultec EQP-1A.
To begin with, I generally like to have the master fader at unity (0dBFS). Ultimately, the only really hard rule for the level of your mix is don’t allow clipping. (Arguably exporting in floating point can alleviate this, but that’s a whole other story.) If you are getting the track mastered professionally you may want to check what levels they expect, but as a good hard and fast rule – don’t allow clipping and don’t use a limiter.
The last plugin should be dither… if you are exporting as a 16-bit wav/aif file, and your DAW doesn’t allow for this in the bounce dialogue (eg. Logic Pro). If you add dither, don’t lose sleep over it, just add a 16-bit option with some noise-shaping and forget about it.
So in conclusion, keep the master fader at unity. Apply minimal EQ if it can’t be avoided. Use light compression with a fairly low ratio.
Some sort of analogue emulation goodness is always great, and if dither is used, it should be the final plugin.
For further experimentation, stereo expansion and psychoacoustic processors such as exciters could be employed, but this definitely overlaps with the mastering stage and should probably be treated as a once-in-a-while type of processor.
Reuben Rowntree is a lecturer at SAE Institute. In his spare time, he spends way too much money on building hardware compressors. He can be contacted at firstname.lastname@example.org.