00:00Hello, and welcome to the Sound On Sound Recording and Mixing Podcast channel. I am Eddie Bazil. To cover the subject of equalisation in its entirety would be a mammoth task, and I implore you to head on over to the Sound On Sound website, magazine, forums and podcasts, and search for all things EQ if you want a deeper and more detailed experience. However, in this podcast, I'll explore some equalisation techniques that I've found to be really useful, both in terms of corrective and creative processing. I won't concentrate on dynamic equalisation, as I've covered this in another podcast, but I will dip my toe in the dynamic EQ pool for the odd example.
00:48Equalisation, also referred to as EQ, is the process of adjusting the gain or volume of specific frequency bands to achieve a predetermined result, be it corrective or colouring. Traditionally, the most common forms of equalisation were shelves, parametric and graphic. However, nowadays we have all manner of equalisers which, although based on traditional topologies, offer far more precision, dynamic and colouring control of a frequency band. Examples include dynamic EQ, split EQ, overdriving bands, harmonic generation of bands and so on. It won't be long before an EQ plug-in warms my bath, makes my coffee, and generates usable memes.
01:33Apart from the various types of equalisers on offer, it helps to understand a little bit about the differences between active and passive equalisers, and linear and minimum phase designed equalisers. Not just for the sake of completeness, of course, but to help you making your life easier by selecting the right tool for the job. Active EQ circuits use powered components, op-amps, transistors, to boost or cut frequencies, allowing for precise surgical adjustments. Passive EQ circuits use only passive components, resistors, capacitors, inductors, to cut frequencies, offering a more musical or coloured sound, but they cannot add gain. That is a simplistic explanation, but the one thing it doesn't address is the fact that some passive EQs will require a makeup gain element to restore the processed volume. The Pultec EQP-1 is a good example of a passive equaliser that utilises a tube amplifier stage to restore the gain attenuated from the passive process. The iconic Manley Massive Passive also uses tube circuitry for its makeup gain.
02:45In a linear phase equaliser, all frequencies are processed through the filter by exactly the same amount. A cut or boost will yield the same phase response. Linear phase EQ preserves phase relationships, avoiding frequency-dependent phase shifts at the expense of latency and pre-ringing. Minimum phase EQ offers natural analogue-style or coloured results with no latency and no pre-ringing. You will often hear folks say that for corrective purposes that require surgical precision, linear phase is your best friend, and the same folk will state that if you want to colour a sound, then minimum phase is the way to go. Generally, this is true, but it is context-driven. I've often processed the air band of a mix using linear phase as I achieve a better controlled response, but I've also used minimum phase for the same job and achieved the same result. Everything in audio production is context-driven, so feel free to experiment.
03:46And it doesn't end there. Many DAW manufacturers place filters in the same category as equalisation. It makes sense as an equaliser utilises a collection of filters, but there are marked distinctions in design, application and behaviour. Filters are generally described as cut-only devices in that they remove certain parts of the frequency spectrum and their designs reflect this. Commonly, filters come bundled as high pass, low pass, band pass and band reject, but nowadays we have a whole gamut of varying filter types and many are also combined with pre and post EQ processes to afford more detailed and coloured control. The reason I've dropped the word filter on you is because I often make my sound design and mixing decisions based on what I feel will achieve the desired result.
04:36Let me indulge myself and give you an example of how simple synthesizer filters can create gorgeous textures. Have a listen to the following simple EDM bed I put together using two sets of filters, one static and one modulated. Here's the processed example. [music] What you're hearing is one in the bowels of hell sub bass, two sets of pads with one having its filter, a band pass, modulated with an LFO, whilst the other is using a static low pass filter and nothing else. Here is the static filter example. [music] I created this sound on the fly using a simple saw wave for the oscillator, a 12dB slope low pass filter with the cutoff at around 620Hz, and some reverb. And here is the example with the modulated band pass filter. [music] I'm using Xfer's mighty Serum for both pads. However, the real action takes place with the modulated filter adding expression and movement to the second pad sound. I've created an LFO shape that is a peak triangle and set its rate to a slow eight bar duration. This is modulating a band pass filter with resonance and drive, and with a slope of 12 decibels per octave. The movement of the filter from 190Hz to 337Hz is what allows for that sweeping texture. Thank you for allowing me this indulgence.
06:47Right, back to EQ. Although EQs are weapons for creative processes, it is corrective that they truly excel at. With an EQ, you can tame sibilance, remove or attenuate plosives, notch cut cycle hums, work the air band, balance and colour masters, control reverb responses, band pass channels for gain staging and so on. The list is endless. And with today's dynamic EQs, we have all manner of detailed band control, including frequency specific side chaining, threshold control multiple band dips and boosts, revealing mass frequencies, attenuating summed frequencies, and so on.
07:27And now developers have taken equalisation to a whole new level. Behold, the split EQ. Ah, yes, the split EQ. The process involves splitting the audio into two distinct elements, transients and body, and affording tools to manipulate both elements separately. I'm going to concentrate on split EQs for the next example. I have to say, some of these software developers are quite clever. Splitting a sound into transients and body is one thing, but to then be able to drive the body to generate harmonics takes this topology to a whole new level. A seat at the EQ Jedi Council, if you will. The three split EQs I use are Boz Digital's Transgressor 2, which I exhaust when editing drum sounds, Eventide's sublime Split EQ, which is the mother of all split EQs, and I use it for processing acoustic instruments, and finally, Sonible's insane entropy:EQ+, which is a weapon when enhancing transients and generating harmonics. I'll start with a simple kick enhancing example using Boz Digital's Transgressor 2. Transgressor 2 is my go-to plug-in for all drum-based transient processing. Its GUI is simple yet detailed, but it is what it can do under the hood that makes this one of the best transient shaping tools available. The plug-in is sensibly split into two sections. The left-hand section deals with transient processing, while the right-hand section deals with the sustained element or body of the sound. Each section employs a four-band equaliser as the primary processor. Sections can be switched on or off if the user wants to hear the processing in isolation or to completely remove a section. Shaping is confined to a simple hold and release envelope, and there is a gain knob to manage the overall output. Basic filtering and re-triggering are provided at the output stage to further shape the processed response. I'm using it to shape a kick into something a little thuddy, warm and a tad dirty. Have a listen to the dry version. [music] And now the processed version. [music] I've dialled in a massive 16 dB boost at 60 hertz in the transient section. I've done this to highlight the attack a little more, as I felt it was wanting in that department. And in the sustain section, I've cut everything up to 200 hertz by just under 10 decibels using a low shelf. The cut creates room for the transition from attack to body, and this highlights the two elements nicely without having huge smearing between attack and sustain. The kick has more thud and a nice clicky, dirty attack.
10:31In the next example, I'm going to use Eventide's SplitEQ to rebalance an acoustic guitar line. Thank you to James Perret for allowing me to use this recording. Here is the dry version. [music] And here is the processed version [music] Eventide SplitEQ is a must-have for any producer who takes the creative approach to EQing sounds and mixes. Yes, it is a full-blown, detailed, and unnervingly accurate equaliser, but it is its creative and musical weaponry that is its crowning glory. The concept is almost simple enough. The beautifully designed and presented engine provides a split function that can isolate transients from the tonal body of a sound and afford processing for both, both independently and in unison. To list all the exquisite features this beast offers would require another podcast. So my suggestion is you head on over to Sound On Sound and dig up Matt Houghton's review of this plug-in. Back to the example. I've utilised three bell-shaped bands with two bands highlighting the tonal aspect of the guitar line and one band to accent the transients. Band one, which is set to tonal, has a one and a half dB dip at 150 hertz. Band three, which is also set to tonal, has a 3dB boost at 290 hertz. Band four, which is set to transients, has a 2.4dB boost at 1.7 kilohertz. Using two tonal bands, one to cut and one to boost neighbouring frequencies, while sharing crossover frequencies, is an old school method of removing muddiness from a specific frequency range. The crossovers reign supreme here, as cutting a range and then reintroducing a boost at the tail end of that range that then fades into the neighbouring range with a 12dB per octave slope allows for a smooth transition from cut to gradually reintroduced frequencies. Band four is simply boosting the transients, and I'm using a 12dB per octave slope and a very wide Q value, i.e. the bandwidth. The result, albeit a little overemphasised, is a more open sound with highlighted finger picks and strums.
13:48We have now reached that point whereby a little blah, blah, blah, blah, blah about EQ slopes and lead ins and lead outs and crossovers could be a tad helpful Whenever you EQ a sound, it helps to match the slope or gradient of the filter to the response of the sound being equalised. It doesn't matter whether you're using bell shapes, low pass or high pass filters, shelves, and dedicated band pass and band reject filters, and that includes dynamic and split nodes, matching as close as possible to the response of the sound being equalised will yield more natural results. I've often heard clicks when people try to high pass a kick drum using too steep a slope on a high pass filter. I've also heard jarring attacks and decays on sounds that have used a bell shaped filter with very steep slopes. The lead into a sound is defined by how a sound starts, and the lead out of a sound is how the sound decays or tails off. When you're dealing with a mix, these lead ins and outs of individual sounds within the mix are indistinguishable, so we rarely use steep slopes on EQ filters when EQing the overall mix or when mastering. But when working on a single sound that is dynamic in nature, say a vocal line, the utilised bands that cover the various frequency ranges need to be managed correctly, and this is where matching slopes to responses and managing crossovers comes in handy. Frequency transitions within a sound that is being equalised requires sensible crossovers. Think of a single crossover as the response created by two slopes for two bands, one that leads out from one frequency range and then leads in onto the next frequency range. A very basic rule of thumb is that if you have two filter bands close to each other, then the crossovers need to be smooth. If they're far apart, then the simple decay of one range leading into another will be sufficient and act as a faux crossover. But if they're too close to each other, then steep slopes will make the lead ins and outs abrupt and unnatural.
15:55Let's now move on to what I feel is one of the most important roles that EQ plays at the pre-mixing stage, and that is of band passing channels, be it hardware or software, i.e. the DAW. The reasons we do this are as follows: channels will sum at the output stage. In other words, shared frequency gains will add up from channel to channel, and that can play havoc with gain staging and trying to achieve a LUFS target. Summing is common and whether it is acoustical or electrical, gain values will add up. That would normally not be such a big problem, but you also get summing with frequencies you cannot hear. We call these useless frequencies redundant frequencies. They serve no purpose if you can't hear them. But these redundant frequencies will also sum, and you can imagine the hell they'll play with any threshold dependent process like compression, expansion, and so on. They will invariably false trigger these dynamics if their sum values exceed the waveform peak value that is used to set the dynamics threshold. Removing these redundant frequencies, both high and low, by using a band pass equaliser will ensure that false triggers are managed. Additionally, these redundant frequencies will clutter a mix. The low end of a mix can often sound smeared, dark, and murky. By removing redundant frequencies, you will be amazed at how easily the low end opens up and sounds cleaner. It also helps greatly with managing the kick bass relationship and any low register instruments that battle for prominence. The same can be said of the high end. By leaving redundant frequencies to populate the high end, quite often the air band will sound brittle, harsh, and dark, and any mid-range sounds that have or generate harmonics higher up the register will also sound harsh or hard. Redundant high frequencies will often make reverb sound metallic and hard. Delays can sound dark and flat. Phasers can sound grainy. Distortion can lose clarity, and so on. In production, we have a term we like to use, work with the sweet frequencies. In other words, work with the frequencies you want to keep and can hear. By band passing channels, almost all of these problems will disappear.
18:15In the final example, I'm going to use band passing and air band processing to both highlight and rebalance a mix called Another Way, kindly donated by Rene Schneeberger, who wrote, performed, and mixed the track. The artist featured is Forester. This was all made possible thanks to that Jedi, Mike Senior, who helped me choose a mix from his wonderful website, Cambridge Music Technology. If you want to practice your mixing skills, then head on over to Mike's site and play around with a multitude of mixed audio tracks and stems kindly donated by various artists. I must stress that I'm not trying to do anything to Rene's mix other than display how band passing and working the air band can dramatically alter the perception of a mix. I've not added any effects or dynamics other than band passing and air band processing, but I have pushed the bass, vocal drop, and drums so that you can clearly hear the results of the used processing. However, I have used SSL's saturator to add some growl to one of the bass tracks, and the hook has been gently caressed with SSL's valve comp for no reason other than the fact that I'm an erratic vagabond and like to be difficult. Here is Renee's mix. [music] And the processed version. [music] I have bandpassed every channel, and it would take forever to list the values, so I'll cover the strings one bandpass, as it demonstrates how I have not only removed redundant frequencies, but matched both the high-pass and low-pass slopes to the sound's response. I tend to use FabFilter's Q3 equaliser for all bandpassing chores, as it has a great GUI which makes it easy to hone in on the frequency ranges I want to manipulate. I've set a high pass at 62 hertz with a slope of 24 dBs per octave to match the lead-in response of the string sound. The low pass is set at 7.2 kilohertz with a slope of 18 dBs per octave to match the lead out of the sound. I've used SSL's X saturator for the main bass track, and I've used a hefty amount of drive coupled with emphasis on second and third-order harmonics, albeit leaning more towards second order. And although I'm using the wet/dry to afford a parallel setup, I'm opting for more wet than dry, as I like the growl. The synth hook is being run through the mighty SSL X valve comp for some thickness and gentle second-order harmonic distortion. The hook sounds a little warmer and more pronounced. The bandpassing has beautifully highlighted the sweet frequencies and removed any clutter that would have arisen had all the redundant frequencies summed and coloured the mix. The result of this process means I can now work the air band truthfully, and my weapon of choice here is Tokyo Dawn Lab's gorgeous TDR Arbiter. The plug-in is described by the manufacturer as a frequency specific spectral balancer. Acting independently of absolute level, this audio plugin makes an ideal solution for a wide range of typical audio mixing, restoration and mastering challenges. Who am I to argue with that statement? I must say, Fabien's plug-ins are my go-to for all manner of corrective and colouring balancing tasks, and Arbiter is simply sublime. I'm using Arbiter to dynamically boost the air band. The premise is that by manipulating the high end of a sound, you can impart more transparency and presence without it sounding harsh or brittle. While we're here, let's make something very clear. The nonsensical internet advice of the air band being a specific frequency is not true. The air band is determined by what lies at the high end of the audio signal and not a predetermined value. However, many will agree that the air band should exceed the fundamental frequency content and highlight the higher register harmonics. Personally, I stick a high shelf filter on a sound or a mix, sweep until I find the range that I believe encompasses the air band, and boost gently and audition the response. If it sounds like it has more presence, more intimacy and more lift, then it's worked. That is exactly what I did here. I swept until I settled on 5.4 kilohertz as the center frequency. A gentle shelf boost of less than three decibels was enough to lift the mix. And as it was dynamic in nature, the response sounds far more natural than a static boost. These are the important settings I wanted to share with you. There are many more parameters you can play with and I encourage you to play with all the supplied parameter manipulation tools that come with this plug-in, as they can have a subtle or dramatic effect on whatever process you opt for. I've soloed the air band in the screen I have provided so you can gauge how gentle the response I've dialled in is. But listen to how it has altered the perception of the entire mix, even though I have processed a small frequency range. I have barely covered all the wonders that equalisation gifts us. My recommendation is to head on over to the manufacturer's websites I've mentioned and explore their products, as they will help you to clean and colour your mixes effortlessly. That's it for now, thanks for listening. This has been Eddie Bazil for Sound On Sound.
Thank-you for listening and be sure to check out the show notes page for this episode, where you'll find further information, along with web links and details of all the other episodes. And just before you go, let me point you to the soundonsound.com/podcasts website page, where you can explore what's playing on our other channels.
