This is an overview of Part Six of the My Musical Mouth project in which I write a complete piece of music using a microphone and my mouth to write all my midi parts and some automation. In this section I look at how to boost all the levels of my audio to make it sound professional. If you haven’t read the previous parts then you probably should, but this is not essential if you are simply a Cubase user who wants to make things loud as these techniques could relate to any project. This part is really an appendix as I will not be using my mouth here, I will just be looking at how to use Cubases’s dynamics processors and mixer routings to create a finished track that will stand up against modern loud productions.
If you would like to read about how to use your mouth to write music then check out the other articles using the links below:
Many people complain that the music industry has gone a bit over the top with the obsession to be the loudest. However, there is a lot to be said for cranking up the levels and it’s not all about making peoples ears bleed. Good compression and limiting adds warmth and makes your tracks sound fuller by pumping the volumes and filling little gaps with sound; this also makes them sound better at lower volumes. Running a compressor and limiter from the beginning of producing your track can also save you a lot of time, as it means you don’t have to worry much about levels – as more synths and sounds are introduced into the mix, the compressor and limiter pulls the volumes and peaks down , giving a nice even mix without you doing anything.
It is a personal preference, how far you want to smash your audio signals to achieve loudness, but in my project I’ve been fairly rough with my treatment to get that live, blaring sound.
So far I have written a 4 bar segment which contains a drum loop, a bass-line, a melody, a couple of synth-lines and a pad. I have not arranged these components yet, because every note I write, effect I use, or EQ I introduce will be affected by the compressor and limiter, and any automation may have to be redone, so it makes sense to insert the compressors and limiters first so that I am working with the correct levels. Also, by effectively mastering the track as I write/arrange, it means that upon writing the last bit, my project will be finished and immediately ready to bounce down.
What I am looking to achieve here ultimately is the track to sound as loud as possible with the final output at dead on 0dB (i.e. the maximum level for a CD). Whenever a channel on the mixer or a level within a plug-in exceeds 0dB things start turning red to tell you that there is a potential problem. Cubase 6 can mix at much higher levels than 0dB due to its 32bit floating point architecture – this is an interesting topic, but I’m not going into it here. So, when writing, I set my channel levels high and then use compression and limiting to bring the final output back down to zero. To get an idea of where to start, I normally pick a tune that is similar to mine and listen to it, then try and get my levels to sound as loud.
Our ears aren’t designed for low frequencies and we can barely hear them, so you will find your low frequencies are cranked right up. If you don’t like bass, then you are in dreamland and can write the loudest track ever; if you love bass then your going to be doing battle with physics and biology and I wish you every luck.
My track is a dance track, and as with all dance and pop tracks, the kick drum should be nice and loud and proud, which will use up most of my “headroom” (the available loudness). Because the kick drum is so bass heavy it requires a lot of volume to be heard – the bassier you want you kick drum to be the louder you will need it to be to hear it. It is for this reason, you will notice in most pieces of pop and dance music, the audio wave clearly shows the kick drum as the loudest and biggest peak, even though it isn’t necessarily the loudest sounding instrument. Ultimately the kick drum will determine how loud my track will be, so I start by isolating it and crunching it to make room for everything else. The more I can squeeze it down, the more headroom I can recover and the louder all my other instruments can be.
My kick drum is in Groove Agent One which I loaded into the VST Instruments rack. Grove Agent One can have a massive 16 different stereo audio outputs, so the first thing I do is open the VST Instrument rack and click on the little output button immediately to the left of the rack slot (where the instrument name is displayed) and activate a second audio output, as you can see below:
I now have two stereo channels for Groove Agent One in the master mixer, allowing me to have completely different routings, levels and effects. By default, all the sounds in Groove Agent One are playing out of GAOne 1 (channel one), so I to set my kick drum to go to GAOne 2. To do this, I open Groove Agent One and click on the kick pad; now the display shows all the settings for that individual pad and using the knob furthermost to the right, beneath the instrument’s screen, I change the kick’s Output from ST1 (stereo output 1) to ST2. You can see this in the picture below:
I then open the channel view for GAOne 2 and set a steep sloping EQ with the frequency set to 30Hz and the gain pulled right down; this will remove all the frequencies that can’t be heard or reproduced by speakers and allow me to boost the audible frequencies to a higher level.
I am then ready to compress and limit my kick drum to boost it, so I click in an Insert slot and from the pop-up list I go down Dynamics folder and then click on VSTDynamics; you can see this below (I have a number of additional 3rd party plug-ins installed, so don’t expect to see the same list as I have):
The VSTDynamics plug-in then pops up. It is important to have the kick playing back whilst you are changing the settings so that you can look at the levels and hear the results. The VSTDynamics plug-in an amalgamation of three different processors called GATE, COMPRESSOR, and LIMITER, which you can see listed along the bottom of the plug-in screen. The small circles next to these names are the power buttons, so I click on the one next to COMPRESSOR and the one next to LIMITER to activate them, as these are the processors I want to use. The COMPRESSOR settings take up much of the middle part of the screen, and the controls are as follows:
- Threshold: This is the level in decibels at which the compressor will start affecting the audio, so it is effectively a dynamic on/off switch. As soon as the audio signal hits this level the compressor will start compressing.
- Ratio: The harshness with which the audio will be compressed, the higher the number, the more severe the compression.
- Make-Up: Once the audio has been compressed it will be quieter, so you use the Make-Up to get it back up to the level you want. The “Auto” button will return it to the same decibels it was to start with before compression, but to trigger the LIMITER you need to send the audio out to a level above the limiter’s threshold which is 0db. So you should turn the “Auto” off and then use the Make-Up to get the limiter working.
- Attack: The speed with which the compressor will start attacking the audio signal and compress it once the level is over the threshold.
- Release: The speed at which the compressor will stop compressing the audio after it falls below the threshold.
- IN (level meter): Shows the level of the audio as it enters the compressor, which is useful for setting the threshold.
- GR (level meter): The amount the compressor is impacting on the audio (if this isn’t doing much then lower the threshold, if this is going berserk then you may want to raise the threshold
You can see all these settings in the picture below:
For my kick, I have chosen a Threshold of -8.9dB against an input of -4.5dB (the volume works in reverse, from 0dB downward). I’ve then selected the maximum ratio of 8, the fastest attack and “auto” release. All of this means that any audio coming in over -8.9dB will be immediately crushed by 8:1 and as soon as the audio falls back below -8.9dB it will be left at it’s original level. This creates a pumping loudness – which is perfect for my tune.
I put the Make-Up gain right up to +10dB, which means when I get to the LIMITER, the “IN” level is reporting a red warning that my audio is +3.1dB. This is what I want, as the LIMITER will then bring this down to 0dB by attenuating the peaks (like getting a pair of scissors and just snipping off the top of the peaks). I then set the Output volume of the LIMITER to -6.0dB as that sounds about right and leave some headroom for other instruments.
To make mixing easier and to compress the all the percussion together for a squelchier sound, I create a group channel to send my two Groove Agent One channels to. The group channel is exactly the same as a regular channel with all the same functionality (EQ, Inserts, Send, Volume, etc.). To create a group, I simply right-click in the empty Track Header area and select “Add Group Channel Track…”, as you can see below:
I name the group “Percussion” and a new channel appears in the mixer with a corresponding track in the Project Page (for automation purposes). I want to route my kick drum channel (GAOne 2) and my old percussion channel (GAOne 1) to the new group, so I open the mixer and along the top it shows the output bus for each channel. For the two channels I want to route to the “Percussion” group, I click in the Output bus slot and from the drop-down list select Groups > Percussion, as you can see below.
I then go to my group channel, “Percussion”, and insert an instance of the VSTDynamics plug-in effect and active the COMPRESSOR and LIMITER sections. The IN level on the compressor is showing that the level is peaking at +0.2dB – this is because the kick and the clap, when playing at the same time, are very loud. I set the Threshold to roughly where the kick level on its own, -6.8dB, so that when the kick is playing on its own the compressor has no effect. I then set the maximum Ratio for a harsh compression, and the Attack and Release times to the minimum for a very fast response. I crank up the output to +4dB to hit the LIMITER but then drop the Output Volume on the LIMITER to -4dB to fit into my mix and leave room for my instruments. You can see all these setting in the picture below:
With my percussion levels set, I then look at the next headroom muncher, bass. I have two bass sounds to start with; the one which comes in during the intro section and the rather nasty sounding one that begins during the first break. I create a group called “Basses” and route my two bass sounds to this new group. I then insert another copy of VSTDynamics into that group channel with a fairly mild set of compression and limiter settings (as they aren’t particularly bassy and all I need to do is take the peaks off and leave a nice even signal level). This group highlighted another benefit of groups later on when I decided I wanted to run a 4 bar loop of the harsh bass though a long ping-pong delay to create a doubled-up, panning effect, which I needed to bounce down so that I could just the second two bars as an audio loop (this became the bass-line that runs between the two breaks). I was able to just route this audio track into the “Basses” group, avoiding having to re-do the compressor or levels.
Finally, I create a “Synth” group for all my other synths and pads and insert another instance of VSTDynamics with very light compression and limiter settings. All my audio signals are now running through one of three groups before going into the master bus which means when I mix these three “groups” of audio back together I will be totally free of unwanted peaks or surges in volume which might cause unexpected results when I compress and limit the master output. With compression and limiting working on each of these groups, the instruments within the groups are only competing with each other for headroom. So if four percussion sounds come together at once, they will only compete with each other rather than causing my bass or synth parts affected. Also, when two or three synths play at the same time through the “Synth” group, they will be compressed to have the same apparent loudness as when just one of them was playing – so the compressor and limiter act as a kind of auto-mixer.
With all my levels going into the master output channel at a controlled level, I then insert one final instance of VSTDynamics into the output bus to ensure my final mixdown will be brought down to exactly 0dB ready to be burned onto CD. I set the VSTDynamics on this channel to be far more gentle, as all I really want to do is pull the levels down to 0. You can see these setting in the picture below:
You can see in the picture above that the volume coming into the compressor is at +4.4dB, which will cause hard clipping on my final mixdown, and it is for this reason I am inserting this master set of VSTDynamics. I have set the compressor to a gentle ratio of 5:1 and the threshold to -8dB which looks like quite a heavy setting (a difference of 12dB), but from listening and looking, it is clear that the audio doesn’t often reach +4.4dB and this setting sounds fine. I then play with the Make-Up setting to push the levels over 0dB again and start the limiter working. The higher I can push this without and audible problem, the louder my final piece of music will be. As it turns out, I can push it up by another 4dB and it still sounds great, so this is where I leave it. I leave the output of the Limiter set to 0dB and now when I export my tune the levels will be at exactly 0dB.
This way of working may sound completely wrong to some people, but this is the story of My Musical Mouth project and this is how I wrote my track. You can hear the results for yourself, by clicking on the Soundcloud link below, and decide if it worked or not.
The beauty of heavy compressing and limiting is that it becomes a sound effect as much as a dynamic process; an effect that fills the empty spaces of sound and adds a pumping warmth and liveliness that brings a piece of dance music to life. However, it doesn’t necessarily work with everything and some sounds are easily distorted; resonant sounds such as vocals and sounds that have few peaks, such as sub-basses, are especially troublesome.
This is officially the end of My Musical Mouth – officially this article shouldn’t have been part of the series as it has nothing to do with using your mouth, but hopefully it was a useful insight into how I got my levels right up – and I could go on! So, I’m going to leave it to a public vote. If ten people leave comments at the bottom of this article asking for more, then I will cover the arrangement, the effects I used and how I automated them.
Don’t forget, if you are a customer of ours or want to buy a copy of Cubase, then you can email me on firstname.lastname@example.org for full tutorials in PDF. For now though, goodbye and happy mouthing.