Philth Tips #11 – mixing week 6 (dynamics – part 2)

Last week we looked at dynamics processors, in particular compressors. This week we will add some new tools, and consider how we can apply dynamics processing to groups of sounds.

Dynamics processors

Transient shaping – shape the Attack (the transient) and Release (the tail) of a sound. Particularly effective with drums to make them more pokey or tighter.

Noise gate – used to cut background sounds. The gate will only open when the signal goes over the threshold, so only the loud parts of a track pass through the gate. Very effective on live drums to remove the noisy background in between hits.

Saturation – using Overdrive and similar plugins will limit your sounds and also create harmonic distortion. This effectively makes things sounds thicker and fatter.

Limiting – stops signal going over 0db and clipping the channel. However be careful as running into a Limiter too hot will still cause distortion.

Clipping – some compressors have built-in Distortion. This enables you to clip the signal and squeeze it up to appear louder without leaving spiky transients.

Applications of compression

Buss compression
Using a compressor on a buss group will help to hold all the sounds together by reducing their dynamics and shaping their transients as a group rather than individually. For example – compressing layers of vocals to avoid sudden peaks; compressing all of your drums as a group to avoid peaks and also potentially make your drums pump. This is why people refer to buss compression as ‘glue’, sticking groups of sounds together.

To use this technique you simply need to route all your drums/vox/etc directly into a Bus track instead of going straight to the stereo output. Then when you apply plugins to this Bus they affect all of the incoming audio at the same time. Now a compressor will respond to your entire drum mix. You can setup the compressor to emphasise transients, to reduce dynamic range, to cut off any peaks…. all of the normal uses of compression but applied to your whole drum mix, processing it all together.

This also means you can apply Saturation or Limiting (or even both) at the end of your chain, which will cut off any peaks and recover some headroom. Then you can push your drums up so they appear louder in the mix without using up all of your headroom. My absolute favourite tool for this is Camel Phat at the end of the drum chain to give the drums an extra squeeze.

Parallel compression
This technique involves mixing a heavily compressed signal back into the original dry signal. Most commonly used on drums (for extra impact and energy) and vocals (solid signal with the impression of dynamic range).
Set up an auxiliary channel and use aux sends to choose which signals are routed to the compressor. You can then apply aggressive compression to your signal (no dynamics on vocals/exaggerated pumping drums) and use the aux fader to mix this signal with your dry signal. I normally set up a final bus to group the dry and parallel signals together for a final bit of control.

When writing DnB I use parallel compression on my drums in a variety of ways, depending on the style of track.
For minimal or techy tracks I will push the compressor on the parallel chain to emphasise transients (hard compression, Attack/Release set to let the drums poke before they are compressed), I will then double the effect by using an Enveloper to cut the tails and push the initial attack. When you mix this back into your original drum mix you end with a very tight and snappy sound, spiky drums.
For liquid I will often go in a completely different direction, and use the parallel compression to bring out the shuffles of my breakbeats by having a very short attack then timing the release to let the shuffles jump out. Using the Soft Clipping feature in the Logic compressor will squeeze the drums more and result in the shuffles appearing to pop out of the drum mix. This can then be mixed with the dry signal to achieve a swinging or pumping effect.

Dynamics - parallel compression

This first screenshot shows how the two drum busses are mixed into a final drum bus. This Drum Master channel makes it very easy to adjust the volume of the drums in relation to the rest of your mixdown.

The Parallel is compressed aggressively, with a fast attack and release to ensure all transients have a POP. The Enveloper is pushing the first 10-20ms of the transients to bring out more poke, which is then clipped with the Overdrive.

The Drum Mix is compressed gently, a little squeeze after letting the first 50-60ms punch. Then the UAD plugins are a Tube Amp for saturation, a Neve EQ which is bringing up the bass of the kick with a low-shelf, and finally, Overdrive to catch peaks and squeeze a bit more.

On the Drum Master there is EQ on the whole drum mix, in this instance to push the mids, and finally Limiting to push up the perceived volume while avoiding any peaks.The result is the brick wall of drums you will see below.

In this second screenshot, you see the effect that each stage of processing has on the audio. The end result is thick fat drums that are controlled by the Limiter.

Dynamics - processing drums.png


Philth Tips #4 – Mastering a DJ mix

Screen Shot 2017-01-03 at 21.18.24.png

I’ve spent the last few weeks working on my 2016 End Of Year Mix; first choosing the tracks, working out the running order, exporting to USB’s, then actually mixing it (it took me 3 shifts because I had to stop for food a few times). The final step is giving the mix a rough master before it goes online. And I do mean rough – this should not be interpreted as a guide to mastering your own music, pay a professional to do that. But you want your DJ mixes to sound good right? The following steps will help you to balance the levels and squeeze the mix up a bit louder.

At the risk of being patronising – watch the levels when you record your mix!! If you are smashing tracks in at wildly different volumes then it is impossible to level your mix later without major surgery. Give yourself some headroom when you record, and watch the input level as you record into Logic, Audacity, whatever you use, to make sure your gains aren’t creeping up.

You also need to bring down the volume faders if you have got two tunes dropped together or you are going to make peaks and troughs every time you double-drop, and end up with one track sounding quiet when you play it solo. Finally, don’t play two basslines on top of each other because that will: a) double the volume; b) sound like shit when the subs clash.

Now your mix is recorded and ready for mastering. Do the basic trimming so you haven’t got any silence at the start and finish. Now it’s time to do some processing. My examples will all be in Logic, but the principles are the same, you need a compressor, EQ, a limiter and a spectrum analyser.

Compression will enable you to level out the mix, bringing down the level of the loudest parts. I use the standard Logic compressor. As a very very basic guide to the functions of a compressor:the Threshold is the volume level that will trigger the compressor, the Ratio is how hard it will compress (reduce) the signal, the Attack is how quickly the compressor will react, and the Release is how quickly it lets go once the level goes below the Threshold.

If you want to stop any peaks completely then set the Attack as low as it can go. and then carefully bring down the Threshold so the compressor only responds to the loudest tracks in your set. You don’t need to compress the tracks that are playing quietly, you just want to catch the bits where you got excited and went too loud. You can now use the Ratio to determine how heavily the signal is compressed. A higher ratio will mean that the level of the louder parts of your mix is reduced more. This comes back to your recording itself, if you have been consistent with your gains then the compressor won’t need to do so much work.

Depending on your compressor of choice this quick attack time can completely remove any spikes, and now you can set the Release of your compressor to determine how quickly it lets go. Too short and you are going to hear the compressor pump in and out. A longer release time will mean that once the input goes too loud, the volume level will be held down for long enough that you don’t really hear it come back up, it just happens naturally when the music lets it. When you are mixing down your music the release time is vital to making your compressor dance with the music. For a DJ set it is better that you don’t hear the compressor moving, so err towards a long release.

I hope this makes sense so far. The idea is simple, set your compressor so it only reacts to the bits of your mix that have gone too loud, and then it pulls these loud parts down by the appropriate amount to keep your DJ mix at a consistent level. You don’t actually want to heat the compressor working so use the Attack and Release until the effect is transparent.

Next is EQ. And to be honest I find myself doing less and less EQ to my DJ sets – the tracks have already been mixed and mastered and my job as a DJ is to use the controls of my mixer to keep the sounds balanced. A good mix doesn’t need loads of EQ. If you are working from a live set sometimes it needs more work, especially if the monitoring wasn’t great or if you have a vocalist that is too loud. But for a standard studio mix I will often just use something like Slate’s Revival to give a little bit of extra roundness to the bottom end, and if it is a vinyl mix then a healthy boost in the top end.

Once upon a time I would obsess over getting all my tracks to be level in terms of their frequency balance. This is where your Spectral Analyser comes in, set it up so you are metering the mastering output, and then you can see how much bass, mids, tops is playing at any one time. You can then automate EQ for tracks that really need it, and in this case I would just use a high or low shelf EQ to bring up a wide area while one particular track played. But as I said, nowadays I mainly avoid doing this. I try to stay on top of the Eq while mixing and then if there are slight differences that’s just how the songs sound!

At this point you should have a mix that is balanced in terms of volume and also frequency. The final step is to make it LOUD so get a limiter and push the threshold/gain up until your mix is hitting 0db consistently. If you were mastering a song then you need to think carefully about the Attack and Release of your limiter, but we already established that we will pay a professional for that.

I’ll be honest, I smash my DJ mixes until they are silly loud, there will be no headroom at all on any of the drops, but then I ease off enough that I’m not crushing the breakdowns and there is no audible clipping. If you push the limiter too hard you will start to hear things like vocals and strings distorting. Calm down mate. Ease back.

Bounce your mix to mp3, 320kbps for best quality, 192 0r 128 if it is going straight onto the web to stream. With Logic you can also add ID3 tags, so add a link to your website, Facebook page etc. A top tip is to add some jingles at the start because that seems to stop the Soundcloud police from taking your mix down. Ask a mate who is an MC to do a short jingle/ident for your mixes and layer it over your set in Logic.

The final step for me is to add some artwork, and with iTunes it is easy to embed the artwork onto the file. If you want a guide to making artwork I’m definitely not your man. I have got about a hundred photos of me playing out wearing checked shirts (thanks Apps!) and I have just about worked out how to paste a logo on top of a photo using Preview in OSX. Skillz.

My 2016 End of Year Mix is done, mastered, tracklisted, and has been sent off for an upcoming feature. Watch this space.