Blind Mice mixing

At Blind Mice Productions, we do mixing and mastering for some of our releases. We split the mixing & mastering duties: I (Pete) do the mixing and Ben does the mastering. This post will look at the first part: mixing, and go through some of the processes and tricks involved in how to achieve the best sound with as much headroom, clarity and separation before handing it over to Ben for mastering.

Step one: Housekeeping

Setting up all the folder tracks, and keeping an original stereo mixdown in a separate channel so I can solo/mute it as an A/B test.

The first thing I do after importing the multitracks into a new Cubase project is create Folder tracks for each of the tracks, and then get my synaesthesia on by colour-coding them. Black for the kick group, red for the bass group, blue for the synth melodies and purple for the vocals (because vocals are the most important part deserving of the coolest colour), etc. The use of Folder tracks in Cubase also allows easy muting/soloing groups with a single click which is invaluable in the mixing process.

Step two: Gainstaging

Sample-rate or "slices-per-second" and bit-rate or "steps-per-slice"

Hit play on the mix and take a look at the master fader. If every individual channel is nearing 0db then the master channel will be clipping. Don't drop the master channel to compensate (and for the love of god, don't just throw a limiter on the master channel to rein it in!). Instead, set every fader in the DAW to -6db. The VU meters on the individual tracks should never go near 0db and in fact I like every channel to hit an absolute maximum of -6db at its loudest point. There is some science to be aware of here: dropping the faders of the individual channels will effectively reduce the bit-rate unfortunately, but nowhere near as damagingly as clipping the output or throwing a hard limiter on it. For those who don't know, your music quality is defined by its bit-rate and sample-rate. The sample-rate is the amount of "slices"-per-second your waveform has, while the bit-rate is the amount of steps-per-sample. When you smash your track with limiters, you're reducing the amount of possible steps-per-sample and damaging the overall resolution; instant lo-fi. Incidentally, this is also a good time to mention that you should always work in at least 24-bit audio, but all the bits in the world won't protect you from bad gainstaging!

Step three: Subtractive EQ

See this mess below 200hz on this synth lead? Chop it!

This is the single most important step of mixing. I use Fabfilter Pro-Q and getting this plugin has got to be one of the single most important steps forward in my production advancement. Stick it on a lead synth-line and turn on the post-EQ visualiser and prepare to be appalled at how much low-rumble there is that you can see but might not necessarily perceive by listening to the sound. This low-end rumble contributes to the volume of that channel which all gets summed to the master channel and results in less overall headroom. Chop the unneeded frequencies out of each of your channels to win back some energy and clarity in your mix.

I have the following general rules in mind when mixing:

  • Give the kick and bassline a rumble cut at around 18-22hz
  • Nothing but the kick and bass really need any frequencies below 125hz
  • Synth melodies can have an EQ cut as high as 400hz
  • Cymbals can be cut as high as 800 or even 1200hz
  • Guitars should be quite thin on their own; chop out their low and high ends so that they fit into a mix instead of eating it all up
  • Notch out any shrill frequencies by sweeping a band around the spectrum at +12db; when it peaks at a particularly undesirable frequency, that’s the frequency you’ve got to chop (double check on headphones to make sure you’re not just detecting a standing wave in your room though)

Step four: Envelope shaping

I use Xfer LFOTool to add shape to almost anything, especially kicks, snares, hats, percussion and basslines. Kicks and snares should retain plenty of shape and a clear transient; brickwall kicks and snares aren’t going to translate well on big sound systems. Using an envelope shaper at the end of your plugin chain, you can crush and saturate the fxxk out of your kicks to design a nice beefy sound, and then give it back a nice shape with plenty of punch by shaping the amplitude.

Step four b: Easy on the Transient Designer

Throwing a Transient Designer plugin on your kicks and snares will give them an incredibly satisfying spikey punch that really cuts through a mix. But beware: this is actually a fallacy. What you’re doing is creating a spike in amplitude for that layer. Now what do you think is going to happen to this when it comes time to master the track? The mastering engineer is either going to have to squash your spikey kick with a compressor (thereby undoing it’s transient bite anyway), or in order to maintain the dynamic range the overall perceived volume of the master will have to be much lower. Try and achieve the effect of a nice transient with other methods such as compression, envelope shaping and layering instead.

Step five: Sub bass

Double your bassline with a single sinewave a couple of octaves down. Any synth can oscillate this, but the Rob Papen SubBoomBass VST is particularly well suited.  This will give the super low end a superlatively stable foundation that will hold its own even if your bassline changes pitch. Don’t forget to give your main bassline a low-cut (80-120hz should do) to make room for the sub layer. If your monitoring situation is lacking, this is the point where you’ll start to really feel the need to put some money into some new speakers.

Step six: Sidechaining

If the kick isn't 4/4 I'll set the sidechain input trigger to MIDI and program a MIDI channel to feed into it.

Benny Benassi didn’t invent sidechaining. It’s been in use since the early days of recording and there are plenty of ways of using it as a mixing utility without necessarily sounding like an electro house try-hard (though I still personally love the sound of ducked basslines). I use the Vengeance Multiband Sidechain plugin and use it on every bassline. It’s invaluable in reducing the low-end of a bassline in an extreme way but having a lighter effect on the high-end, to get the kick/bassline mix just right. Of course you can use it as a single-band pump-machine but that’s going to really date your production when electro house gets out of fashion. 

Step seven: Stereo placement

An overly “mono” mix can sound amateurish. I do things like sending a bassline to a group with a heavy fuzz distortion and a heavy-panning doubler effect (and a low-cut EQ), to give basslines a sense of stereo width. This also works on other layers e.g. vocals and percussion. Just make sure you flick your master channel to mono mode and check on headphones as well as speakers to make sure you don’t produce any whack phasing effects.

Step eight: Automation

If you chopped the hell out of the bass-end of a layer to get it to fit into a mix when the track’s in full-swing, but there’s another section of the song where that layer is solo’d, you could automate the EQ so that that layer’s not so castrated when it’s on its ownsome.  

There is actually a vaguely scientific reason DJs can't get their hands off the bass knob.

There are also other things you can do like lift up the low-end just before a drop so that when the beat kicks back in it has a fuller effect. This is the same effect of a DJ fxxking with the low-EQ knob like they always do. The science behind this is that speakers are constantly pushing and pulling/vibrating to produce the sound, and if you chop out the low end of a track you’re giving the main speaker driver (the one that produces the bass frequencies) a change to rest and “reset” into its default position, so that when your track jumps back in with a heavy kick drum the speaker is ready to respond with full force instead of being in the middle of a push/pull-cycle.

Step nine: Reverb

I like to use reverb sends, rather than sticking reverb plugins on individual channels, to give a greater sense of “glue”ing the mix together. All percussion can be sent to the same send. Vocals to another as it’ll probably require different settings. There are exceptions to this: I’ll use reverb as a channel insert when I want to automate an effect where the layer gets pushed back and “gets further away”, which is super useful in breakdowns to do things like fade the bassline or percussion out without being so lo-tech as to simply pull down the volume fader. Remember: panning places sounds left and right, but reverb places it between the foreground and background; use this 3rd dimension to place elements in particularly busy mixes.

Step ten: Overall check

How does the mix sound overall? Can you hear the kick presently and feel it in your chest at high volume, while also picking out each element of the bassline? How does it sound if you go stand in the kitchen, any weird bass effects from this distance? Most importantly: how does it sound tomorrow? Never mix a track and call it “done” on the same day.  

This should really kicks your mix’s ass by the time you’ve done all this. Hope it’s useful, and feel free to comment below if there’s anything you want to add or contest!