Now that you know how to sample and arrange your material into mashups, it’s time to talk about levelling and mastering, so that your mashups feel cohesive, volume and equalisation-wise, helping the sounds fit together a little bit better. I’m going to share my personal approach to volumes as well as outline my compression set-up and explain the theory behind it, so you can understand how I do it in order to aim for your own goals. My setup is basically some traditionally used plug-ins for live acts with some custom adjustments I made.
You should from this point, start following along with the fifth video of my tutorial.
The first thing I want to share is a tip about envelopes. I find that it’s way easier and less confusing to change any clip parameter envelopes in the clips themselves. Use the clip envelopes for changing volumes and volume automations for example, instead of using the channel envelopes. Only use channel envelopes to automate plug-ins and channel-only parameters. That said, let’s now see my approach to levelling.
I always level my samples in order to achieve a final mix where everything feels balanced, with nothing buried or too loud. But I also make sure that my beats are punchy and not overpowered. Beats should hit at somewhat of a constant “hardness” – it’s dance music after all. Also, vocals should be clearly perceptible and not buried. If you have trouble hearing the majority of the lyrics, it’s probably because the vocal is too quiet. With this in mind, I also try to have my riffs and melodies the most audible possible.
The best way to have this kind of balance throughout your set is to make one mix that you feel is very balanced and reference any other to this one. If everything in this reference mix is clearly audible, it really doesn’t matter about its volume. If you make every single mix sound similar level-wise to this one, you’re on your way to having a balanced set overall.
This way, if you feel the overall set volume is too low or to high throughout, you just need to increase or decrease the gain once for everything. Just make sure not to mix with samples levelled too loud, or it will over compress.
Mastering is basically EQ and compression. There’s an unwritten rule in mastering that states “garbage in, garbage out”, this means that no matter how hard you try to make a mix shine, if it is a bad mix to start with, it will always be bad. On the other hand, if it is a good mix, mastering can make it a great one. I mention this because you should always strive to have mixes sounding good before any EQing and compression. If you have a poor combination of samples – for example, vocals on top of melodies in which the same frequency range is completely full – you won’t be able to find space in the mix in order to have everything properly balanced.
In popular songs, if you notice, the instrumentation is very sparse in the verses, in order to accommodate for the vocal and for it to be perfectly perceptible. On the other hand, the chorus generally has a lot of instrumentation and a simpler vocal hook. That’s another of the reasons why you should mix verse vocal parts with verse melody parts and chorus vocal parts with chorus melody parts, like I mentioned in part two of this series. I’m also constantly trying new settings myself, sometimes I think it gets better but then end up reverting to what I had before.
I’m going to share my mastering setup with you, but more importantly, the theory behind it. This is important so that you can create the set-up that best fits your needs, and so that you won’t be afraid to tweak anything. I’m also constantly trying new settings myself, sometimes I think it gets better but then end up reverting to what I had before. Let’s just say it isn’t an exact science!
The objective of this set-up is to have everything fitting together as well as possible, while keeping a reasonable level of simplicity. It employs equalisation in the form of EQ plug-ins and multi-band compression, as well as final master compression and limiting.
Compression is a process that affects the dynamics of music, generally narrowing the gap between the higher and the lower amplitudes by lowering the higher ones and raising the overall volume. In other words, it will make the quieter parts sound louder.
It works by setting a volume level (“threshold”). Every time the volume exceeds this, it is lowered by a certain ratio. A ratio of 2:1, for example, will mean that for every two dB exceeding the threshold, only one will pass. “Attack”, “release” and “knee” are also common compressor parameters. Attack represents the time it takes for the compressor to start after the signal exceeds the threshold. Release means the time it takes for the compressor to stop after the signal is below the threshold again. Finally, knee controls how gradually or suddenly the compressor kicks in.
Multi-band compression is a form of compression that only affects certain frequencies of the spectrum. As such it can be used for “smart EQing”.
One possible set-up
I have 13 channels on my live act, divided into six groups, namely: Beats, Pacekeepers, Riffs, Melodies, Vocals and Transitions. Each one of these groups acts as a bus. This means that if you apply an effect to it, it will apply to every channel in the group. It will also combine the sound from all the channels in the group into one.
In each of these 13 channels I have an Ableton Live EQ8 plug-in, which I only use for removing problematic frequencies, when needed (for example, overly loud leftover hi-hats on DIY acappellas).
I also have an EQ8 plug-in in all the groups. This plug-in rolls out any sub-bass below 40Hz, helping to avoid muddy lower ends. I use a very small curve as you can see in the video; and you shouldn’t worry too much either if you’re using already mastered material, like commercial music.
The only exception is the Pacekeepers group, in which I roll out everything below 300Hz. This is because I only want hi-hats, snares, cymbals, and other higher pitch percussion sounds coming from there, since the kicks are coming from the Beats group.
At the end of the chain in every group I have a limiter. A limiter is basically a compressor set to an infinite ratio, not letting any sound pass the threshold volume. In this case the threshold is set to 0dB and this helps to prevent distortion.
I send both my Beats and Pacekeepers groups into a third bus, called percussion. This bus is where I compress together my beats with my pacekeepers. I only have a limiter in there which is not a very elegant solution, compared to a properly set compressor, but it works since I have my samples fairly loud to start with. So when they’re combined, they’ll pass 0dB and get compressed together.
You should follow along now with video six of the tutorial.
In my Riffs and Melodies channels I also have a multi-band compression set-up. This works by listening to the beats and vocals that I want to mix with the riffs/melodies, and compressing (lowering the volumes) of the appropriate frequencies in the riffs/melodies. This uses a technique called side-chaining that helps to “carve” space in the frequency spectrum for the new material.
Side-chaining is a popular compression technique that is often used in hip-hop mixtapes, where a rapper raps on top of a loop of some other song. Side-chain compression helps in getting space in the mix for the vocal. It means that the signal being listened to, in order to see if it crosses the threshold or not, comes from a different sound source.
In these two groups, I have an effects rack with four chains in each one, namely: Low, Mid, High, and Pass. Each of these chains will filter and compress the different ranges of the frequency spectrum.
The Low chain will compress the low end on the riffs/melodies (below 100Hz). The Mid will compress the part of the middle range where the vocal clarity lies (roughly 3.5 to 6kHz). The High will compress the higher frequencies in order to accommodate for the high end parts in vocals, as well as for the high end parts in beats (above 6kHz). Finally, the Pass chain will let pass the majority of the middle range unchanged.
This chain starts with a low pass EQ, in order to let only low frequencies through. I’m using the Linplug EQ from Waves because I like its curves, but any EQ will do. You can use one or more of Ableton Live’s EQ8, with one or more EQ points, in order to get the curves you want as well.
This chain will listen to the bass part of the beats coming from the percussion bus, and compress any bass or beats in the riffs/melodies in order to fit in the new percussion. The side-chaining is being EQed so that it only listens to the range below 100Hz on the percussion bus, which is the same range that will get compressed in the riffs/melodies.
This chain starts with a band filter EQ, in order to only let past the frequencies associated with vocal clarity. Vocals are present in a wide rage of frequencies, but I choose only to compress this certain range and above, so that the main range of the riffs/melodies doesn’t get changed too much.
This chain will listen to part of the middle frequencies in the percussion (around the 3.5 – 6kHz frequency range), as well as for frequencies up to this point in the vocals. It will then compress this range in the riffs/melodies in order to get some space for the vocals. The side-chaining to the percussion is not that necessary, it is only to compensate in case the percussion also has this range too busy.
This chain starts with a high pass EQ, in order to let only high frequencies through.
This chain will listen to the high frequencies (6kHz and above) coming from the percussion bus, and the same range coming from the vocals group. It will open some space for the higher frequencies in vocals as well as high frequency percussion sounds.
This chain only has an EQ letting the rest of the frequencies past unchanged: 100Hz to 3.5kHz, most of the riffs/melodies.
The above chains make up what can be called “smart EQing”, meaning that they will lower the frequency ranges according to what’s coming into the mix. This way you won’t have to manually EQ everything for every combination of samples you make. You just need to find a balanced setting that works overall and jam ahead.
If you have no idea how to use Ableton Live’s EQ, compressor, limiter, as well as creating chains, there are some good basic videos about them on YouTube.
My compressor settings
My settings for the compressors vary with what’s being compressed and what’s being listened for. But generally I have “harsh” settings to open space for beat, which are loud, fast attack, short duration sounds; and softer settings for vocals, which are quieter, slower attack, longer duration sounds.
To open up space for beats: low thresholds, high compression ratios, fast attack and release speeds, and no knee.
To open space for vocals: higher thresholds, lower compression ratios, slower attack and release speeds, and some knee to soften the starting of the compression.
Finally, on the master chain, I have another sub-bass roll off, a gain stage (to increase or decrease the overall loudness of the set), and final compression.
For this compression I’m using two more third-party plug-ins. BBE Sonic Maximizer and PSP Vintage Warmer.
- BBE Sonic Maximizer gives a little more punch to the beat. It helps compensate for the smearing effect of Ableton Live’s warp modes, recovering some of the transients
- PSP Vintage Warmer is a limiter. Limiters should be the last process, in order to avoid signal passing over 0dBFS (0 decibels Full Scale) and distorting. This last limiter sets the overall loudness of the final product, and this limiter in particular also helps compensate for the metallic/electronic sound of Ableton Live and digital processing in general. It gives some “warmth” to the bass and cleans the highs a little as well. In other words, it colours the sound to resemble a more analogue recording
I also have a spectrum analyzer that helps me monitor levels in general, as well as helping me find troubling frequencies, and the notes played in riffs/melodies.
Having a set-up like this one in Ableton Live will help your mixes sound balanced overall, without buried vocals or faint beats, provided you picked good sample combinations to start with.
It’s fairly capable for live performance; however, if you wish to make a complete “song”, you might need to dig in deeper and tweak more precisely in order to meet the needs of the “song”. Also don’t forget that original songs have all the parts written in order to fit in with each other from the start.
Mastering is a skill that takes a lifetime to master (no pun intended). I believe that bringing such techniques to the mix can help you differentiate from other DJs. Get these concepts down and you’ll be on your way to setting up your “smart EQs” so you can DJ using Ableton with immediacy and still sound great.
We’ve covered the third part of making loop-based mashups, and also the third of four parts about building a sample-based live act. Next week we’ll cover how to set up a performance around loops and how you could incorporate these techniques into traditional DJ sets as well.
I will show you how I have my live act set up and how you could continue to add new material to it. I will also share a few more general tips and insights on this type of mixing.
Finally, if you feel you can perfectly follow along, you can go ahead and watch the rest of the video tutorials. If not, I will be back next week with some background theory about what’s happening in the videos.
• Fred Dancekowski is a mashup artist/DJ and music student who lives in Porto, in the north of Portugal. You can hear a live mashup DJ performance by Fred Dancekowski on Soundcloud.
Check out the other parts in this series:
- How To Make And Perform Mashups, Part 1 – Sampling
- How To Make And Perform Mashups, Part 2 – Arranging
Are you making mashups using Ableton Live? Do you have any problems or tips you’d like to share? Please let us know in the comments.