Mastering With Acustica 1.3

Second Edition

This page intentionally left blank

Foreword

W

elcome to the First Edition of Mastering with Acustica Audio, An Essential Guide. It took us about three months to get the whole work done. We split this guide up into two sections: the first section was written by Andrea Zanini and translated into English by Acustica; then we added a second section of interviews with people who have been working with us through the years and are now part of our “team”. We hope this guide will give you practical, straight-forward advice on coming to terms with mastering.

“I feel humbled to have been given the opportunity to write this brief guide about mastering, an enormous subject that would require a hundred books only to start scratching its surface. Legendary Indian artist Zakir Hussain says that “The best thing you can do, is try to be the best student you can ever be”. I think this approach applies to any field in music, from studying an instrument for the first time to mastering a whole record. Being given the privilege to “shape” someone else’s music and to listen to it before anyone else, is something that should be accepted with respect, joy and simplicity. This is why this guide’s purpose is just to get the reader started on this matter, giving him or her the basic knowledge upon which to build a creative, critical and personal thought. Music is no competition. It is a lifelong personal evolution”.

Andrea Zanini

III

Acknowledgments

I

would like to express my deep gratitude to the Acustica Audio family for their patient guidance, enthusiastic encouragement and useful critiques of this work, as well as their advice and assistance in keeping my progress on schedule. I would also like to deeply thank my friends Earle Holder, for being such a mentor to me, Martin Walker, Eddie Bazil and Simon Barden for their invaluable teachings, trust and support and Paavo Jummpanen, for his precious tips and knowledge. My grateful thanks are also extended to Steve Turnidge, a real source of inspiration to many engineers and artists out there, including me. Finally, I wish to thank my wife and my parents for their invaluable support throughout this work. Always Keeping Learning,

Andrea Zanini

IV

About the Author

A

ndrea “Andy” Zanini is a Record Producer & Mastering Engineer based in Italy. He started his musical education in 1988, at age two, studying drums (his main instrument) and piano. He received an intensive education in Indian Classical music as a tabla player, production and audio engineering. Founder of Owl Mastering studio, he is highly regarded for his musicianship, fresh enthusiasm, and meticulously innovative approach to mastering and audio production in general.

V

Mastering with Acustica An Essential Guide

Table of contents

-Introduction...................................................................................................................................................................0007 -Analog or Digital? -The Mastering Engineer’s Mindset -Mastering Tools -Equalization....................................................................................................................................................................0014 -Frequency Ranges -Resonances -M/S (Mid-Side) Tecnique -Stem Mastering -Magenta -Amber AC-55 -Green -Compression..................................................................................................................................................................0041 -Dynamics Pocessor -Types of Compressors -RMS vs Peack -Compressor Controls -Release Time -Side-Chain Control -Look-Ahead Function -Parallel Compression -Multi-Band Compression -Limiting

VI

-Titanium -Stereo Processing.......................................................................................................................................................0063 -Sample Rate-Dith Depth Dithering.......................................................................................................................0064 -DDP and Track Sequencing........................................................................................................................................0071 -DC Offset.........................................................................................................................................................................0074 -Gain Staging. What is it and is why is it important?..........................................................................................0075 -The Interviews...............................................................................................................................................................0078 -Michael Angel -Ali Zendaki -Bob Davodian -Mark Drezzler -Gabriel Schwarz -Hubertus Dahlem -Ken Suen -Franz Mikorey

Mastering your music is like mastering your life. It’s amazing what happens when you clean up the noise, maximize your good work, and have your music sparkle and shine as you really want it to.

Steve Turnidge (Mastering Engineer/Author)

VII

Chapter 1

Introduction

A

lthough it should be called pre-mastering, mastering is often seen by many as a ‘dark art’ or ‘science’ that is difficult to master and is limited to only a select lucky few. Indeed, this discipline requires specific knowledge and no-compromise equipment in order to achieve a professional result. The role of the Mastering Engineer has changed and evolved over the years. They not only need to have ‘big ears’, but also have the delicate task of enhancing any audio material entering the studio so that it sounds optimal for the widest number of reproduction systems in the world. This Quality Control process ensures that specific required technical and artistic standards are achieved. The tools available to a Mastering Engineer are not so different from those used by a Mixing Engineer: compressors, equalizers, harmonic exciters, filters, and even reverbs. What distinguishes these devices is their high accuracy and internal architecture, specifically designed and built for this purpose. Especially when analog, they can be very expensive, or even prohibitive, if we are talking about ‘professional’ mastering. You can spend up to several thousand euros on a stereo tube EQ. A high-quality vintage compressor could easily run into thirty thousand euros, not to mention the fact that these ‘components’ are worthless if they are not interfaced together by a high-level console and esoteric converters. Then, on top of that, you have to add the wiring costs and the expenses for accurate monitors and an acoustically treated listening environment, so that the Mastering Engineer’s ears can ‘make’ precise decisions. TIP: Read Acustica’s Manifest on the website.

Mastering with Acustica / 008

Analog or Digital? How come that many professionals in the field seem to regard digital as a second-best option? Much has been said and written on this subject and a number of myths have developed due to a lack of information and the production of great works discrediting all those means which, although efficient, are within everyone’s reach. Pure digital plug-ins, especially those designed for professional use, are irreplaceable tools in a mastering studio when it comes to high-precision balancing corrections, which would be inconceivable in the analog world. “Never turn your back on digital” - Bob Ludwig - Gateway Mastering Beyond the arguable views and reasons lying behind this debate, there is a kernel of truth to it. The analog world is characterized by a series of factors (choice and quality of components, circuit architecture, and so on) and by ‘imprecisions’ giving the sound a character that can hardly be reproduced in the digital domain. Whether it is called ‘warmth’, ‘vibe’ or ‘three-dimensionality’, it is not always easy to describe the euphonic effects that, unlike static digital creations, seem to make the sound fuller. Tube or tape saturation is the most striking and ‘emulated’ example, although not always with satisfactory results. Whereas sound is ‘developed’ in the digital domain, it is ‘bent and shaped’ in the analog domain.


The Mastering Engineer’s Mindset A Mastering Engineer is not a wizard or a magician with a one-size-fits-all magic formula. This idea has been perpetuated in recent years and has led to many false beliefs, fallacies, and legends that can be easily debunked by simply using well-trained ears. “Think Holistically” - Bob Katz - Mastering Engineer - Author - Digital Domain If you want to become good at mastering, you need to learn to listen carefully, with a critical and perceptive ear. You need to learn to ask yourself some questions that can help you in the process, such as: “What is wrong with the balance of this piece?”, “What can I do to make this music more alive and vibrant?”, “Sibilants in this voice are terrible, am I able to intervene without compromising the rest of the instruments or would a remix be better?”, “Does the sound I’m getting fit to the genre?” With a bit of time and experience, anyone can learn how to master professionally. It’s all about intuition, knowledge and the desire to patiently learn the concepts underpinning this discipline.

First, you become a proficient mix engineer. Then the creativity kicks in and the label of producer takes over. It is only then that you understand the subtleties of what it takes to achieve the next stage: Mastering.

Eddie Bazil (Sound Designer/ Author - SampleCraze/Stretch That Note)


Mastering Tools Equalizers EQs allow to intervene on the frequency content of the audio material. The human ear can hear sounds whose intensity and dynamics are in the decibel range of about 0 to 120 dB (the threshold for causing permanent hearing damage), within the ideal 20Hz-20000Hz frequency range. Mastering EQs usually have an extremely linear and much wider frequency response. Thanks to these devices, you can ‘balance’ the frequencies so that your track (and tracks between them) sounds in the best possible way.

Compressors Compressors control the dynamics of a track. They can be used to smooth out signal peaks and give the sound more body or reduce the difference in volume between the loudest and softest bits in the same piece of music. They are perhaps the most ‘characteristic’ tools used in mastering, because they are often creatively used to give a static and lifeless piece a specific colour. There are several types of compressors and each can be used for different purposes (see page TBD).

Limiters Limiters can be considered as extreme compressors with very fast attack and release times. Their purpose is to prevent signal levels from exceeding a set threshold. They are usually used at the end of the mastering chain (before the SRC and dithering) to increase the final volume of a track, adapting it to the required tastes and standards, without ever exceeding the dangerous threshold level of 0 dBFS, over which digital distortion will ruin the overall work. One of the features of a well-designed limiter is that its use may seem transparent to the ear, even when a peak reduction of several decibels occurs. As a result of the excessive use of these devices, side-effects such as distortion and loss of detail will be inevitable. TIP: Use a compressor to add some character or alter the dynamics, a limiter to provide a more transparent result.


Reverbs Reverbs are more rarely used in mastering, especially convolution reverbs (able to recreate the natural ambience of real spaces such as halls, small or large rooms, etc.), in order to give some dry and sterile recordings additional width and depth. These tools are also very useful to soften the sound or ‘restore’ a sense of environmental coherence of all the elements of a track and widen the stereo image subtly. Reverbs should be used with extreme caution or the ‘cure’ will be worse than the ‘disease’. Moreover, not all Mastering Engineers agree on their effective use.

Stereo Processors Stereo processors intervene on the stereo image of the audio material, by widening and opening it. Avoid overusing them because significant phase problems and loss of detail may occur. They are the sworn enemies of mono-compatibility. Middle ways are always to be preferred. Using the ears and experimenting are normally the winning formula.

Exciters Exciters are processors using different processes (such as harmonic synthesis, dynamic equalization and harmonic distortion) to add emphasis to sounds, making them richer and brighter. Originally used to restore the high-frequency content of tapes without increasing the background noise level (unlike equalization), today they are used in mastering as a creative effect, in order to increase the sense of loudness and presence of the sound and add brightness to the highs or body to the lows. They give excellent results when used in small doses, or they will cause fatigue to the listener who, instead of enjoying the music, will only have a severe headache.

Monitors and quality headphones In addition to a listening environment that is free from acoustic phenomena such as resonance, flutter echoes and standing waves, it is important to monitor the sound on speakers (active or passive) and quality headphones. Contrary to what many claim (that headphones are to be avoided in mastering), needless to say that this concept is misleading, if not totally wrong.


Although it is true that sound reproduction by headphones is different from loudspeakers, especially stereophony (appearing more natural on speakers), it is also true that a lot of important details, such as clicks, extraneous background noises, some subtle dynamics and even serious phasing issues are heard more clearly on headphones. Moreover, if the mastering room is not optimized across the entire frequency spectrum, headphones help ‘bypass’ some deficiencies in the working environment, which can be ‘mentally compensated for’ when getting back to the monitors. So there is no reason why you should remove this valuable tool from your workflow. You just need to learn how to listen ‘through’ both systems and ‘take note’ of the information brought to your attention.

Meters Meters are very useful tools because they allow an engineer to both listen and ‘see’ what happens at the level of the sound. There are several types of meters with different purposes, among which meeting some technical standards required on a worldwide basis. Among the different types of meters, there are: VU Meter: The analog (or digital) meter used to measure the ‘volume’ of a track, meaning by this its RMS (Root Mean Square) value. The higher this value, the higher the sound perceived by the ear, compared to another whose RMS value is lower, with the volume settings being the same in each case. The VU Meter and Level Meter are equivalent, except that the latter measures peaks and the RMS value placing them on a digital scale with a maximum level of 0 dBFS. BBC Meter: Very similar to the VU Meter, but it is mainly used in the broadcast sector. Spectrum Analyzer: It is a must have in any mastering studio. This meter displays, on two axes, the frequency content of the track you are burning. Spectrogram: It displays the frequency content of the audio material over time, so that the energy content in each frequency range is displayed on a variable color scale. This tool is very useful to bring to the surface hidden resonances and, in general, areas that could benefit from additional compression/equalization. Phase Meter: It provides a measure, on a scale ranging from -1 to 1, of the degree of ‘mono-compatibility’ of a mix (where 1 is mono and -1 is completely out of phase). Generally values between 0.5 and 1 represent a good compromise between a good stereo image and acceptable mono-compatibility. Dynamic Range Meter: It measures the maximum peak and RMS levels, but also the difference between them, the so-called ‘Crest Factor’, not to be confused with the headroom, which refers to the ‘exploitable space’ between the highest peak value at the absolute limit of 0 dBFS. A track may have an average crest factor of 12 dB and a null residual headroom (maximum peak at 0 dBFS) and still sound lovely, whereas another track may have a headroom of 10 dB and a crest factor of 3 dB and have non-existent and muffled dynamics.


LUFS Meter: It measures the loudness as perceived by the human ear, according to the ITU-R BS.17703, EBU R128 standards. Many radio and TV stations are adhering to this standard so that the audience can enjoy the sound without annoying ‘jumps’ in volume. The target level is set to -23 LUFS. TIP: The procedure of loudness normalization has become automatic for all broadcasters where the standard is mandatory. As a result, it is totally useless to produce masters with high volumes and squashed transients, because their volume will be lowered (normalized) to -23 LU. In a few words, a rock mix will sound as ‘loud’ as Brahms’ Ninna Nanna in order to have a uniform perception of the sound, without the audience having to use the remote control for frequent volume adjustments.

Focus on the dynamics quality and do not enlist for the Loudness War! If you are curious to know what the dynamic range of your favourite album is, you can have a look at: http://dr.loudness-war.info/


Chapter 2

Equalization “Mastering is 90% EQ” - Craig Anderton - Mastering Engineer - Writer - Educator

E

qualization is perhaps one of the most difficult aspects to ‘master’ in the world of mastering. Even the slightest variation has a crucial impact on what the track will eventually sound like. In addition to one’s personal taste, this makes you stumble around in the endless search for the perfect sound. Moreover, even the most accurate listening system has its own way to represent the sound… which the ear has to get used to. TIP: Both on headphones and monitors, be careful not to go overboard with volume. Besides being dangerous, your ear fatigue will cause you to make wrong decisions that will mean you having to go back to them at a later stage.

There are several types of equalizers. In mastering you usually use linear-phase digital EQs and digital or high-priced analog parametric EQs. Without focusing too much on technicalities whilst still covering the main practical aspects, you need at least to become familiar with the technical terms and functions associated with equalization. TIP: Cut/boost levels at different frequencies are usually very subtle in mastering. Provided that the mix spectrum is well-balanced, even differences of half a decibel could be significant and ‘shift’ the listener’s focus to some elements over others, as well as alter the track’s character. If a mix is unbalanced, stronger corrections can restore the order of things (always limiting EQ adjustments to a maximum of +/- 1.5 - 2 dB makes no sense at all). It is clearly better to go back to the mix and correct the causes of significant imbalances. Mastering can make the difference, but you cannot always expect miracles, can you?


Below is a small glossary of the most commonly used terms in the EQ world. Filter: A circuit which alters a limited range of frequencies. Bandwidth: The width of the frequency range altered by a filter. Peaking Filter: A filter which boosts (or cuts) a specific band of frequencies. Centre Frequency: In a filter it is the frequency at which a Peaking Filter applies maximum gain (e.g. a gain of 2 dB in a filter with a centre frequency of 220 Hz will apply a 2 dB boost at this frequency and a smaller boost at the surrounding frequencies based on the Q value). Q Value: A measure of the width of the ‘bell’ of a Peak Filter (no wonder it is sometimes called Bell Filter). High Q figures indicate a narrow bell, whereas low Q figures indicate a wider bell. High-Pass Filter / HPF: A filter which progressively attenuates frequencies below a certain frequency. Low-Pass Filter / LPF: A filter which progressively attenuates frequencies above a certain frequency. Band-Pass Filter: A filter which allows a certain range of frequencies to pass. N.B. A first-order filter produces a roll-off of 6 dB/oct.; a second-order filter has a 12 dB/oct. slope; a third-order filter will have a 18 dB/oct. and a fourth-order filter 24 dB/oct. In some digital EQs, you can also find slopes of 48 dB per octave but such steep slopes may often cause unnatural or undesired effects (resonances). (High / Low) Shelving Filter: A filter which boosts or cuts the signal beyond a certain reference frequency. Cut-off Frequency: The frequency at which a HPF or LPF has attenuated the signal by 3dB.


Frequency Ranges What is essential in the equalization process is to learn to precisely identify the frequency ranges where the different elements of a mix ‘reside’. The older Hi-Fi systems, which you can still find in some houses, often feature control knobs saying everything and nothing about their use: they are called Lows / Mids / Highs. Kind of vague, isn’t it? If you want to make your way in this discipline, you definitely need to start thinking at least in terms of Lows, Low-Mids, Mids, High-Mids, Highs. You will start from here to increase your ‘array’ of frequency intervals. TIP: One of the most common mistakes beginners make is thinking that an increase in high frequencies corresponds to an automatically higher ‘definition’ of a mix, only to then find out that their dog is suffering from a terrible headache.

Let’s start to split our ‘action field’ into small groups: 20 – 40 Hz : This is the lowest frequency region where you can still hear the sub-harmonics of some instruments such as the bass, kick drums, synthesizers, subkicks, etc. HPFs are very often used to limit (or filter out) this range of frequencies. Our suggestion is tdecide what to do each time. In vinyl mastering you try to limit the amount of energy in this frequency interval (as well as high frequencies), in particular in the SIDE channel (see page TBD), in order to avoid potential vibration of the cartridge while playing the records. Elliptical EQs, which deliver these frequencies in mono, are therefore used. 40 – 80 Hz : The ‘lows’, as they are often called. This is perhaps one of the most favourite frequency ranges, as it adds more power and body to the sound. How many times did we bump into consumer headphones whose strength (or weakness, we should say) lies in the ‘bass-boost’? The problem is that this frequency range is very difficult to govern so as to sound good on most listening systems. Cheap speakers rarely extend up to 60 Hz, not to mention the full scale. Since very common genres of music, e.g. Rap, EDM and Hip-Hop, often have an extra amount of energy in this range, the aim is to achieve a balance that gives body without muffling the sound. 80 – 250 Hz : The warmth and body of many instruments reside in this frequency interval. The guitar and bass strongest frequencies is around 100 Hz. The lowest frequency on a guitar is around 80 Hz. The vocal warmth, as well as the snare drum body, often resides between about 200-220 Hz. This frequency range is often reduced to make vocals more intelligible in some musical genres. Sometimes the solution is to increase the distinction between the bass and guitar bits by reducing the signal to around 100 Hz, yet with the risk of excessively ‘depriving’ the mix of its basis. In any situation your ears will be the final judge (a practical example of a track with a good measure of energy in this range is ‘Forever in your Eyes’ - Mint Condition). 250 – 500 Hz : Low-Mid frequencies include spaces, mid tones of basses, acoustic and electric guitars, snare drums, and add more tonal consistency to instruments such as the double bass and piano. Even the drum toms have a considerable content of these frequencies and their excessive energy


may make the sound muddy. No wonder this range is often slightly attenuated in the tracks of the drum mics. This helps focus better on cymbals and drum attacks (a good example is Bruce Hornsby’s famous track ‘The way it is’, which is rich in these frequencies). 500 – 2000 Hz : Two complete octaves whose ‘focus’ is crucial to a successful master. We are talking about Mid frequencies. In many commercial masters, both on peaks and the average level, you can hear how these frequencies are (or sound) hollow. Whereas it is true that too many Mids may make the sound nasal and messy, too few of them will make the mix sound really weak and ‘precarious’. This is a fulcrum range where a good balance will provide a good starting point for the master to sound good almost anywhere. One of the most common problems typical of these frequencies are resonances (see page TBD), which can be partially treated and rectified at the mastering stage, but they should ideally be identified and fixed during the mixing process (listen to the track ‘You’re the voice’ by David Foster to have a general idea of how, in a mix, the predominance of frequencies in this range affects the final result). 2000 – 4000 Hz : The High-Mids are essential to the percussive attack of bass drums, snare drums and toms. An increase to around 2700 – 3000 Hz (a very sensitive frequency range for the human ear) may help increase the ‘projection’ of electric guitars or a piano’s high notes, as well as the vocal strength. Excessive High-Mids will make the mix sound difficult to hear and quite exhausting (e.g.: ‘Swerve City’ - Deftones). 4000 – 6000 Hz : Definition and strength of melodic instruments and vocals. Do not exceed or your ears will ‘bleed’ due to a cutting and sharp sound. 6000 – 12000 Hz : These frequencies add detail and sparkle but… careful with sibilants! It is not uncommon to hear consonants and sounds like ‘ch’ and ‘s’ (around 6-7 kHz) being really out of control. Using multiband compressors and dedicated de-essers in mastering can be a solution (not the best one, though) only if you cannot go back to the mix and deal with these problems. Many Mastering Engineers refuse to use de-essers since they also significantly affect other elements of a mix. Not only that, but a significant reduction in sibilance would make the singer seem to have… ‘a lisp’ (a tangible example of the problem arising from sibilance is the track ‘Heatwave’, composed by Dave Stewart & Barbara Gaskin). 12000 – 20000 Hz : This range imparts a sense of openness and ‘air’ to the sound, in particular on vocals and cymbals. In conclusion, it is clear how you cannot act on a single element of a mix without affecting the others. Mastering is the art of compromise, where each action has multiple outcomes. Some are choices of production, others are driven by personal taste, some others are caused by misjudgements, but perfection does not exist! … “one of the most important aspects of “Mastering” is the correction of the spectral content of an audio track. The relationship between the frequency range of 20 Hz to 20 kHz should be even and well balanced. If this initial integral function is ignored you will find that there is a need to test your


mastered project on a variety of speaker systems. The genre of the music has no bearing on the final output even if the genre calls for heavier bass, lower mid range, or sloping highs. Additionally, it does not make a difference whether your mastering chain is digital or analog. It is very important that sound is distributed equally throughout the audio spectrum. Balancing the harmonic content of your track should be a function you become extremely familiar with”. Earle Holder - Mastering Engineer ( HDQTRZ Studios - Public Enemy - Paul B Allen III).

Basic EQ Quick Tips: Filtering out / attenuating frequencies below 25 – 35 Hz may help clean the mix up of useless noise, which only sucks out the headroom and is of little help to the musical information, especially in fast tracks. In general, it is always better to attenuate frequencies above those of interest rather than amplify the frequencies you want to hear. If your mix sounds too clustered, try to attenuate the 100 – 300 Hz range by 1-2 dB. If your mix sounds too nasal, try to attenuate to around 700 – 1250 Hz. If your mix sounds too sharp, it is because of the frequencies lying between 5000 – 8000 Hz. If you want more detail, try with a slight increase (0.5 – 0.8 dB) to around 8000 – 10000 Hz. If your mix sounds a bit dull, try with a slight increase to 12000 – 15000 Hz. Use your ears. Train your ears. Protect your ears. Experiment! TIP: A good tool to keep on hand is this useful interactive chart graphing many instruments’ ranges across the full audible range. http://www.independentrecording.net/irn/resources/freqchart/main_display.htm


Resonances Often overlooked and wrongly underrated, resonances are like ‘woodworms’ devouring ‘useful sound’ and taking up precious space. They may lie in any of the above-mentioned ranges, but become particularly challenging in the mid-range. Resonances have diverse and sometimes unusual profiles, and it is not always easy to find out what causes them: persistent frequencies ‘ringing’ throughout the track without ever stopping (some are even off-key, so to speak), ‘puffing’ noises (like when you blow into a bottle, a glass or your slightly concave palm placed on your mouth), static noises, etc. An accumulation of resonances caused by different elements of a mix may seriously ‘congest’ one or more areas of the track spectrum, and not even spasmodic attempts of accurate equalization may help without distorting the sound and altering the balances. What could be fixed by simply taking routine measures (such as, for instance, filtering out the useless low frequencies) during the mixing stage becomes seriously challenging in the mastering process. Not only that, but getting rid of these hidden enemies can make the difference, at the end of the work, between a good master and a simply great one. Think about it!

M/S (Mid-Side) Technique What we are going to explain briefly in this short paragraph is an approach used in the mastering process, which has made great inroads in recent years: the Mid-Side technique. For those who are not yet familiar with this approach, we need to introduce the concept starting with some simple formulas (don’t panic, in practical terms it is easier than it sounds). Starting from a L/R track, you can decode the Mid component (which contains the information shared by the Left and Right channels, namely the mono component) and the Side component (the stereo component, which contains all the differences between the two channels). Thus we have two simple formulas: Mid = Left + Right Side = Left + ( - Right ) = Left - Right (phase-inverted Right channel) In order to recreate the original track starting from these two components, we need to use the second (and the last, I promise!) formulas: Left = (M+S)/2 Right = (M-S)/2 (phase-inverted Side channel)


The fraction with the number 2 shows that the volume is halved (equivalent to -6 dB). If you do not want to use plug-ins automatically decoding and recoding the Mid and Side components of a track and you want to create a custom M/S routing with sends and groups in a DAW (e.g. for stem mastering), you need to… use maths. Having the chance to work on the components separately has several pros, especially that of using all the instruments available on the two components. By listening to the Side channel in ‘solo’ mode, we can easily identify phasing issues and redundant frequencies, as well as balance the environments of a mix and make them more uniform, and correct resonances in reverbs or panned instruments at the sides of the mix. Filtering out of the Side the frequencies below 40-50 Hz is already a good move for focusing the power of the lows in the centre (where we find the kick drum and bass, the very heart and basis of the mix); this little move is also generally well appreciated by Cutting Engineers for vinyl! Similarly, we can work in Mono to try to improve the separation of instruments sitting in the ‘centre’ of the mix and correct any imbalance without affecting the stereo components. Altering the Mid rather than the Side volume, and vice versa, is another way to widen or narrow the stereo image of the track, however it is always advisable not to exaggerate. Even a single dB variation may have striking effects. TIP:If you want to widen the stereo image of a track, try to apply subtle compression on the Side channel only. In many cases, just a few dB of compression are enough.


Stem Mastering Another ‘modus operandi’ at the mastering stage, which could open up a lot of new possibilities, is called Stem Mastering. Instead of working on a stereo track, you work on the stereo groups forming this track (stems) and on their final sum (e.g.: drums, guitars, vocals, bass, percussions, etc.). In addition to this, you can also add the M/S technique previously described to have almost total control over the final outcome. If you have a problem with drums, such as for instance too sharp cymbals, you can intervene on the specific stem with a stereo EQ or by simply equalizing the spectrum extremes in M/S. If guitars have too much energy around 200-250 Hz, you can reduce it without affecting the bass, snare drum body and vocal warmth. If the bass has fluctuating dynamics, you can compress it without affecting the kick drum, bass synth or other instruments sitting in the low-end of the track. The options are endless.


MAGENTA

Mastering Equalizer


Magenta EQ Magenta is a simply unique equalizer and the first Acqua plug-in created with the purpose of providing those who are involved in mastering, both professionally and personally, with a top quality tool. Magenta has accurately and rigorously recreated one of the most highly revered and most frequently used tube Mastering EQs in the world. The sound of this unit is easily identifiable and, in some cases, you just need to turn this EQ on and let the sound pass through it to make it clearer, more compact and lively. In case you do not have the sort of money required to buy this hardware unit (about 6000 euros, an amount that is likely to cause many sleepless nights), never mind. Today, it is much more affordable to get the same sound.

Controls The interface has been ergonomically designed to be intuitive to operate. Power: Once on the master bus, you will hardly turn it off. IN Switches: Used to enable / disable the filter or band they belong to. They are also very useful to temporarily bypass the EQ effect so as to hear the sound before and after the equalization process. In / Out gains with I/O meters: Used to balance the input and output signal, they can also be used ‘creatively’ to saturate the tubes, thereby increasing the harmonic distortion. High Pass & Low Pass Filters: Included in the plug-in Mastering Pack. These filters have a remarkable musicality and, when used wisely in combination with the other bands, they provide an essential control on curves. High Pass: 22Hz; 39Hz; 68Hz; 120Hz; 220Hz Low Pass: 18kHz; 12kHz; 9kHz; 7.5kHz; 6kHz Band 1. 28 Hz – 1.5 kHz Selectable Low Shelf / Bell Filter, CUT/BOOST Gain Range -/+ 0-20dB Bandwidth: Sets the Q values (maximum on the right and minimum on the left) Selectable stepped frequencies: 28Hz; 38Hz; 68Hz; 100Hz; 150Hz; 220Hz; 330Hz; 470Hz; 680Hz; 1kHz; 1.5 kHz


Band 2. 82 Hz – 3.9 kHz Bell Filter, CUT/BOOST Gain Range -/+ 0-20dB Bandwidth: Sets the Q values (maximum on the right and minimum on the left) Selectable stepped frequencies: 82Hz; 120Hz; 180Hz; 270Hz; 390Hz; 560Hz; 820Hz; 1.2 kHz; 1.8kHz; 2.7kHz; 3.9kHz Band 3. 200 Hz – 10 kHz Bell Filter, CUT/BOOST Gain Range -/+ 0-20dB Bandwidth: Sets the Q values (maximum on the right and minimum on the left) Selectable stepped frequencies: 200Hz; 300Hz; 420Hz; 600Hz; 900Hz; 1.5kHz; 2.2kHz; 3.3kHz; 4.7kHz; 6.8kHz; 10kHz Band 4. 650 Hz – 27 kHz Selectable High Shelf / Bell Filter, CUT/BOOST Gain Range -/+ 0-20dB Bandwidth: Sets the Q values (maximum on the right and minimum on the left) Selectable stepped frequencies: 650Hz; 900Hz; 1.2kHz; 2.2kHz; 3.2kHz; 4.7kHz; 6.5kHz; 10kHz; 12kHz; 18kHz; 27kHz TIP: When you compare the equalized and non-equalized version of a track, you need to make sure that the levels perceived are the same! At first, your ear will naturally tend to lean towards the higher volume version of the track, misleading you.


You need to ‘get to know’ how it works on the signal and how you can use its ‘natural curves’. Clearly your ear will guide you through this process. Let’s see some examples: Here is Magenta’s response when, having been turned on, no band or filter is enabled. As you can see, the sound level tends to be slightly enhanced around 50 Hz. Keeping this EQ intentionally on without further adjustment can indeed be considered as a ‘choice of sound’ to be made while mastering your track, in order to exploit Magenta’s natural behaviour and use its typical tube ‘character’.

By increasing the input volume and balancing it with the output volume, you can exploit the natural tube saturation and see the harmonics rise and ‘peek out’ (can you see the ripples between 4 and 8 kHz?). You will have a fuller and brighter, in other words more ‘saturated’, sound. Even when used alone, high-pass filters are extremely useful and musical


With no other filter on, it helps restore a certain degree of response linearity.

These two settings may be useful to attenuate frequencies at the spectrum lower limit for a more defined and focused sound. For faster tracks, the high-pass filter at 68 Hz could be a good starting point (note that no curve is too extreme).


Excellent filters on vocals and backing vocals at the sides of the mix and drum mics (especially when specifically suited for recording cymbals).


The first one is very useful in mastering to obtain the soft sound of a gentle ‘roll-off’ on the very high frequencies and the second one could find a strategic use during a session of stem mastering on the stereo groups of electric guitars, helping focus on the high-mids (used in conjunction with the HPF at 120 Hz and a gain of a few dB at 82 Hz with a medium Q value to restore some ‘warmth’ and linearity, while the filter cleans the spectrum of useless energy in the low range).

Magenta also provides an extremely versatile control of the ‘tonal shaping’ of a track. We can recreate:


Classic ‘Smiley’ or Hi-Fi curves:


Or we can simply play with rather ‘bizarre’ curves, which is not so typical for a Mastering Engineer! The possibilities are really quite endless.


TIP: If you want to keep the character and body of the tape without losing detail on the highs, you can use Magenta to create a small pre-emphasis curve and then use one of the many Tape programs for Nebula to ‘exaggerate the whole thing’ with some saturation and restore the balance on the highs .


The mastering processing chain “Have a light touch and get out of the way” Bob Ludwig – Gateway Mastering There are many different ways of mastering a track. Some approaches can be considered as standard practice, whereas others require a necessary debate and careful analysis. Variables are so numerous that cannot be included in a magic formula always delivering the best achievable outcome. First of all, in order to have an excellent master, you have to start from at least a good level mix, leaving little room to chance and having the right ingredients to achieve a – so-to-speak – ‘professional’ sound. Unfortunately mixes are often sent in for mastering with problems that cannot be fixed without going back to the mix. Let’s see the most common ones: Too high volume leaving no room for manoeuvre, detrimental compression and clipping/limiting, static arrangement with poor element separation, out-of-control lows, loud and disconnected vocals, excessive sibilance, razor-sharp cymbals, unwanted residual resonances, clumsy editing, remaining noises and clicks, phasing issues, high-volume Charleston, too short fade-ins/outs (cutting off reverb tails), large imbalances between the right and left channels, monaural mixes (the claim being that an impeccable stereo image is achieved), etc. Be careful, because these problems can only partially be fixed in mastering. TIP: Try to keep the average volume of the mix around -18 dBFS with maximum peaks at -5/-3 dBFS and always try to deliver the best possible quality work (generally 48 kHz / 24 bit is enough, but if you recorded at 96 kHz keep this Sample Rate).

Let’s see some examples of mastering chains to draw on for your works. Ex. 1: Mid-speed modern rock/pop (drums, bass, two guitars, vocals, various effects) received (exported) at 96 kHz / 24 bit. Minimal mastering: 1. HPF with cut-off frequency of 25/30 Hz and a slope of 12 dB/octave 2. Broadband compressor (with sidechain filter at 150 / 200 Hz) with medium attack and release times, ratio 1.2:1 and medium threshold. A few dB of gain reduction in the high volume sections will be good. Compensate with make-up gain 3. EQ 4. Limiter (ceiling at -0.3 dBFS with a target RMS level of about -12/11 dBFS) 5. SRC & Dithering (for the final bounce to 44.1kHz / 16 bit)

Standard mastering:


1. HPF with cut-off frequency of 25/30 Hz and a slope of 12 dB/octave 2. Broadband compressor (with sidechain filter at 150 / 200 Hz) with medium attack and release times, ratio 1.2:1 and medium threshold, Target Gain Reduction: max 2 dB. Compensate with make-up gain 3. EQ 4. Stereo Processing for a wider stereo image 5. Limiter (ceiling at -0.3 dBFS with a target RMS level of about -12/11 dBFS) 6. SRC & Dithering (for the final bounce to 44.1 kHz / 16 bit) Creative (loudness-oriented) mastering: 1. HPF with cut-off frequency of 25/30 Hz and a slope of 12 dB/octave 2. Multiband compressor (to level out the dynamics of the bass and bass drum, vocals and cymbals) 3. EQ with particular emphasis on the low end (e.g.: +2 dB at 80-100 Hz with a low Q value) to affect the behaviour of the next compressor 4. Broadband compressor (without sidechain filter) with an attack time of about 50 ms and medium-fast release time to cause a ‘pumping’ effect; ratio 1.2:1 and medium threshold, Target Gain Reduction: max 1.5 dB. Compensate with make-up gain, if necessary 5. EQ to add air and detail after compression (e.g.: -0.5 dB at 300-400 Hz, 0.7-0.9 dB at 2.7/3.3 kHz, High Shelf at 8-10 kHz, +1.2 dB) 6. Stereo Processing 7. (Mastering colour) Nebula Programs for Saturation / Tape / Console or Preamp 8. Limiter (ceiling at -0.3 dBFS with a target RMS level of about -10/9 dBFS) 9. SRC & Dithering (for the final bounce to 44.1 kHz / 16 bit) Conservative mastering (without compression): 1. HPF at 30 Hz with a slope of 6 dB/octave 2. EQ (e.g.: -1 dB at 45-50 Hz, high-mid Q; +0.8 dB at 70-90 Hz, low Q; -0.4 dB at 450 Hz, low Q; +0.3 dB at 4000 Hz, low Q; High-Shelf at 8000 Hz +0.2 dB; LPF at 18 kHz) 3. Subtle Stereo Processing 4. Limiter (ceiling at -0.1 dBFS with a target RMS level of about -14/13 dBFS) 5. SRC & Dithering (for the final bounce to 44.1 kHz / 16 bit) Simple corrective mastering (scenario: too loud bass drum and weak bass, annoying peaks on guitars, mix normalized to 0 dBFS, impossible remix): 1. General gain – 5/6 dB. Let’s bring the mix back to an optimal level of functioning to get more headroom and scope for action 2. HPF at 40 Hz, slope 12 dB/octave 3. Multiband compressor: Band 1 (50 – 150 Hz): our aim is to pitch down the bass drum and restore the bass. Side effects: loss of ‘punch’ in the bass drum. Band 2 (500 – 2500 Hz): fast attack times, medium-fast release time, medium-high threshold (we want to intervene on the guitar peaks and leave the other elements as unchanged as possible); ratio 1.3:1, Target GR: max 1.5 dB


4. EQ 5. Stereo Processing 6. Broadband compressor for a general levelling out before using the limiter (medium-slow attack/release times; ratio 1.1:1 and medium threshold so as to obtain a max of 1-1.5 dB of GR in the high volume sections of the track) 7. Limiter (ceiling at -0.3 dBFS with a target RMS level of about -12 dBFS) 8. SRC & Dithering (for the final bounce to 44.1 kHz / 16 bit) ‘Optimization’ mastering: 1. HPF at 25 Hz, slope 6 or 12 dB/octave 2. LPF at 18 kHz 3. Identification of any residual resonance on the Mid channel by Notch filtering 4. Identification of any residual resonance on the Side channel by Notch filtering 5. Low Shelf at 75 Hz on the Side channel, medium Q, - 4 dB 6. Broadband compressor (RMS) with a max Target GR of 1.5 dB. Compensate with make-up gain 7. Second compressor in the signal path (PEAK Compressor). Fast attack and release times with a max GR of 2-3 dB to obtain a transparent result 8. EQ 9. Stereo Processing 10. Limiter (ceiling at -0.2 dBFS with a target RMS level of about -14 dBFS) 11. SRC & Dithering (for the final bounce to 44.1 kHz / 16 bit).


AMBER AC-55 Equalizer


Amber EQ Amber is an extremely versatile EQ used both in the mixing and mastering stages. The option to enable or disable the preamp section allows to use the colour naturally imparted by this machine to the sound at will. Amber interface is very intuitive. At its sides you can find the input and output gain controls to control the gain staging. The first band covers the frequencies from 18 to 450 Hz and can be operated in shelving or peak modes. Likewise, the fourth and last band covers the frequencies from 1.5 kHz to 25 kHz; Gain range: +/- 24 dB and +/- 20 dB respectively. It is to be noted how this EQ has a wide range of uses, so you need to become familiar with it in order to make full use of its features.

Broad Tonal Character

In this figure you can see how Amber is responding when the first and last band are on. The first band was set to Low Shelving at 50 Hz (-5 dB) and the last one to High Shelving at 20 kHz (about -6 dB). Looking at the curve in the graph, you can easily see how the gain reduction affects a larger portion of frequencies because the curve has a broad trend. The first band affects frequencies up to about 400 Hz, with a significant action even between 100 and 200 Hz. Likewise, the last band affects frequencies below 20 kHz down to 4 kHz. By enabling Amber’s preamp section, you can immediately perceive a velvety colour that can be used to add a touch of character. In the figure below you can see how Amber responds with the preamp on.

Pre On


The figure shows how, with the Q and gain values being equal, the response varies across the spectrum. Another ‘inaccuracy’ of its ‘analog’ heart?

Varying Response

Back to the Future! Here are the actual settings used on Amber for mastering an 80s-style synth wave track in the studio. As you can see, the corrections are minor, yet the difference, also due to the colour of the preamp section, is enough to make this well-balanced track absolutely perfect.

80’s Synth Track EQ Example


GREEN EQ Equalizer


Green EQ Green is a great mastering EQ providing extremely versatile tonal control. The curves obtained can be soft or aimed at more specific corrections. Like Amber, Green also provides the option to enable or disable the preamp section. When on, this adds some colour and warmth to the sound, although it is not as clear and visible as with Amber (which may be more suitable for creative mastering). Green offers five bands with variable Q values (from 0.4 to 4, including the Shelf setting in the low and high bands). The gain range rises from -15 dB to +15 dB per band. In the figure below, you can see how the interaction between the mid-bands can help create more or less complex but precise curves.

Middle Bands


Green and Magenta have one feature in common: they can work in the so-called ‘air band’ (the very high frequency range from 15 kHz upwards), which is emphasized by many tMastering Engineers to make the sound ‘crisper’ and provide a sense of ‘air’. Magenta works on 27 kHz, Green on 26 kHz. There is hardly any difference in terms of frequencies, although the way these two EQs work is different. Many Mastering Engineers do not agree on the psychoacoustic effect of emphasizing harmonic frequencies out of the audible range (the audible effects are due to the equalization effect on frequencies below 20 kHz), which are often filtered out as a result of the Sample Rate Conversion. There is little point trying to give ‘definition’ to an old tape recording by emphasizing this range: the background hiss will only be louder. In any case, it is better to use a harmonic generator for this purpose. Remember that the largest number of consumer reproduction systems are neither accurate nor consistent with the spectrum limits. Sometimes it is better to softly attenuate the very high frequencies rather than emphasize them. In the figure below, Green’s ‘air band’ is shown.

Air Band


Chapter 3

Compression Dynamics processors

C

ompressors are among the most used (and misused) tools in modern mastering. They are used for different purposes, from adding colour to altering the dynamics of a track and acting as a ‘glue’.

Talking about compression in mastering can be extremely easy as well as enormously complex. Leonardo da Vinci is quoted as saying that “simplicity is the ultimate sophistication” and we will try to stick to this idea. If I may use an analogy, just as important as the right wine is to a dinner main course, such is the right compressor to the track it is used on. It is not uncommon to have more than one type of compressor during the mastering process, with some being extremely sought-after due to their distinctive sound. In the analog world, there are some revered machines which, although out of production for several years, still fetch staggering prices on the second-hand market. Just think of the legendary Fairchild 670 compressor, the undisputed king of tube limiting (with 20 vacuum tubes!), which can cost up to 50,000 dollars (slightly more than 40,000 euros). No misprints, unfortunately. In recent years, the use of compression has laid down the law on the final result of millions of tracks to such an extent that it has become involved as the ‘accused’ in the case against the Loudness War. The over-use of compression literally kills the natural dynamics of individual instruments as well as whole tracks. Killing the sound natural features has several effects and introduces artefacts that make you lose the musicality of the track you are working on. The stereo image narrows, musical expression


is minimized, the balance between the ambience and its elements is so altered that three-dimensionality is completely lost. Anyone who really knows how to use a compressor is also able to turn it off at the right moment. You also have to consider that, with the new broadcast and TV audio standards that have been adopted on a worldwide scale (see, for instance, EBU R128), any audio content is automatically normalized to be perceived at the same loudness as other contents coming from different sources. This method makes it completely useless to try to ‘sound louder than competitors’. In the two figures below, we show a clear example. Two versions loudness-normalized to -23 LUFS (the reference standard) of one track by Spandau Ballet and another one by Deftones. We let you draw your own conclusions.

Now, all this having been said, you will study the basic concepts of dynamics processing that will allow you to make the best possible use of compressors, for both artistic and corrective purposes, without losing sight of the ultimate goal: to emphasize the musicality of a track.


Types of compressors Without focusing too much on electronics, whilst still covering the main practical aspects, let’s see the various types of compressors and their uses.

Pure Digital Compressors/Limiters They use a set of mathematical operations to adjust the input and output levels and each function generally provides a degree of accuracy and transparency hardly achievable in the analog domain (especially top-quality digital compressors designed for mastering).

VCAs (Voltage-Controlled Amplifiers) Providing high accuracy on the attack and release controls as well as on the gain, this circuit design makes these compressors extremely versatile.

Vari-Mu Compressors Their internal architecture is based on tube technology. This type of compressor does not come with the ratio control, but the level reduction is incremental depending on the input volume. They work extremely well on percussions, drums and, in general, whenever you want to keep transients intact. Generally Vari-Mu compressors have higher response speed on the attack and release than optical compressors (not as high as FETs or VCAs, though).

FETs Capable of very fast attack and release times. Typically they are more commonly used in the mixing stage, where an extremely creative use may be justified.


Opto Compressors It is interesting to see how they work. A light bulb or LED are controlled by the compressor side-chain and a photoresistor in the gain stage responds to their light by varying the gain reduction. Optical compressors react quite slowly to transients and are often used in mastering to ‘massage’ the track. This circuit design is also sought-after for the character it imparts to the sound.

RMS vs Peak A compressor may be able to measure both peak (the highest levels in a very short time frame, usually transients) and RMS signals. Some compressors offer switchable RMS/Peak operation, others are designed to work best only in one of the two modes. A tube optical compressor will work well in RMS mode, whereas a FET peak limiter will be suitable for limiting ‘dangerous’ transients.

Compressor controls Threshold The threshold control helps define the ‘portion of sound’ we want to process and is the level above which compression starts. If we want to limit signal peaks, we will most likely set a high threshold with a hard knee. If, however, we want a larger ‘mass’ of sound in the track to be levelled out to our needs, we should use a lower threshold with a soft knee to soften the compression effect. TIP: In some cases, two cascade compressors are used in mastering to work on different components (Peak and RMS) or, in order to reduce the dynamic range, they can be used to split the ‘workload’ so as to make the compression process as uniform and transparent as possible. Without pumping effects. In general, an overall compression of 2-3 dB is acceptable. Higher values could alter the original dynamics of the track a bit too much and we need to decide whether this is what we want. Often in jazz an overall compression of 0.5/1 dB is already considered at the limit. Compressors should not be used at all in classical music, in particular analog compressors which tend to colour the sound.

Knee The knee is the point where the ratio changes from ‘unity gain’ to a set ratio. With a hard knee, compression is introduced as soon as the signal exceeds the threshold, whereas with a soft knee, the gain reduction process starts as the signal approaches the threshold limit, before exceeding it. This makes the transition softer and the compression effect less perceptible.


Ratio There is no threshold without a ratio, which numerically determines the amount of compression applied once the audio has crossed the threshold. For example, a ratio of 2:1 means that for every 2 dB of volume over the threshold, the compressor lets out 1 dB of volume. Compressors start to be called limiters at ratios of 20:1 all the way up to Infinity:1 (rather extreme). In mastering ratios are kept at very conservative levels, especially if a chain of multiple compressors is used (it is usual to set a ratio of between 1.1:1 to 2-2.5:1). The first important concept you need to understand is that if it is true that high ratio values result in high compression values, their effect is not necessarily more audible than low ratio values with a lower threshold, especially if the attack and release times, which we are going to see shortly, are set approximately without considering the speed of the track. Another thing that is going to sound counter-intuitive is that the higher the ratio, the lower the amount of compression. How? If, for instance, we have a 16 dB overshoot above the set threshold and increase the ratio from 1:1 to 2:1, we will have a gain reduction of 8 dB. If, having the same overshoot, we increase the threshold from 8:1 to 16:1, we will have an additional gain reduction of only 1 dB (compared to 8 dB in the previous example). Now we see how the difference in the signal reduction is higher in the lower ratio values than in the higher ones. By this we are not saying you need a scientific calculator to use a compressor, but with experience and the right musical sensitivity you will not become a slave to your compressor, as is often the case, making it your slave instead. Many Mastering Engineers prefer not to rely on numbers, but on their perception.

Attack Time The attack time defines the time it takes the compressor to reach the maximum value of gain reduction, obtained by combining the functions of threshold and ratio, once the set threshold is exceeded. Returning to the previous example of a 16 dB overshoot with a ratio of 2:1, if we set the attack to, let’s say, 200 ms, that would be the time it takes the compressor to produce a gain reduction of 8 dB (if the overshoot signal was constant and static for a period of time sufficient to allow it to reach this value, which never happens in a song where levels and volumes always change). In the figures below you can see in detail what has been said so far. We used a standard digital compressor, setting a threshold to -23 dBFS with a ratio of 1:1 and then 2:1. We generated a tone burst instantly increasing the volume by 16 dB compared to the threshold.


1.1 (200ms attack)

2.1 (200ms attack)

8.1 (200ms attack)

16.1 (200ms attack)


As you can see, the most critical difference is between 1:1 (i.e. no compression) and 2:1. Compression is not doubled when going from 8:1 to 16:1. The difference is a mere 1 dB. Let’s run the experiment again with a faster attack time, 2.5 ms, and the ratio set to 2:1

2.1 Fast Attack 2.5ms

We notice that the initial transient has died away and the image is smoothed out. Fast attack times alter the dynamics of natural transients. In this case, 8 dB of gain reduction were produced within only 2.5 ms. The natural attack of instruments such as drums and percussions would be suppressed with these settings. Attack times should be quite long if we want to allow transients to pass ‘safely’ and then compress the signal as needed. However, very long attack times, like for instance on a snare drum, would not give the compressor enough time to ‘react’. TIP: Compressing low frequencies with very fast attack time settings may cause annoying distortion and clicks. This effect can be deliberately used in mixing (to add more attack to the bass drum, for instance), but not in mastering.


Release Time Generally, the release time is the time it takes the compressor to return to unity gain (to stop compressing) once the signal falls below the set threshold. If you look at the following pictures, you can see how the release of a generic compressor behaves once the signal drops back below the threshold. We used three different release times: 20 ms, 330 ms and 1.5 seconds.

Release 1.5 sec

Release 20ms

Release 330ms


The first thing you notice is how the compressor keeps on working even after the signal has fallen below the threshold, before returning to unity gain. This is why, if we do not want this to happen, we have to measure out the release time and, accordingly, the threshold, and process only the signal exceeding the threshold. This feature is clearly emphasized now, because we are using a tone burst; in any case, it should always be considered when using a compressor on any type of material, apart from when we want to deliberately ‘remodel’ the natural envelope of an instrument (as we are going to see later). As for the attack time, too fast a release setting can also result in unwanted distortion and clicks on low frequencies, as well as in pumping, especially if the compressor was set to such values as to produce high levels of gain reduction. A long release time also tends to modify the natural decay of many instruments’ sound tail, thereby altering the timbre (which sometimes can be just what we want). TIP: Both in mixing and mastering, it is important to give a compressor’s release time such values as not to result in a change to the rhythmic nature of the track. In a few words, we have to ‘train’ the compressor to keep time! ‘Arrhythmic’ release times tend to be clearly perceptible and are not acceptable in the mastering stage, where transparency should ‘theoretically’ be the rule.

When setting the attack and release times, it is very useful to know roughly how long some frequencies take to complete a frequency cycle. In so doing, even when using a side-chain filter, you will more precisely identify your scope for action depending on what you want to achieve with your compressor. Here is a small chart intended as an additional reference tool. 20 Hz – 1 kHz Frequency Cycle Duration Chart 20 Hz : 50 ms 30 Hz : 33 ms 40 Hz : 25 ms 50 Hz : 20 ms 60 Hz : 17 ms 70 Hz : 14 ms 80 Hz : 12.5 ms 90 Hz : 11 ms 100 Hz : 10 ms

200 Hz : 5 ms 300 Hz : 3.3 ms 400 Hz : 2.5 ms 500 Hz : 2 ms 600 Hz : 1.6 ms 700 Hz : 1.4 ms 800 Hz : 1.25 ms 900 Hz : 1.1 ms 1000 Hz : 1 ms


Side-Chain Control The side-chain allows us to drive the behaviour of the compressor based on an input signal that is an equalized or filtered version of the main input or any other input signal (more commonly used in mixing and for a creative use of the compressor, an aspect that is left out of our analysis). Compressors, by their nature, respond more to low frequencies than the rest of the spectrum. This happens for two reasons: low frequencies have much more energy and a longer wavelength than high frequencies. This makes the signal stay longer above the threshold, once it is exceeded, and the risk is that the compressor is triggered even when we do not want it. The side-chain allows the input signal to be filtered so as to prevent the compressor from behaving like this. Note that an equalized or processed side-chain is always independent of the signal to be compressed. So, even with a highpass filter enabled, every audio signal will be compressed according to the side-chain instructions. It is a common misconception that only the ‘residual’ signal is compressed, leaving the filtered portion unaltered. We are not talking about a multi-band compressor, where each filtered frequency band is brought to the ‘attention’ of a compressor operating independently! De-essers are based on the same principle, but the side-chain is here emphasized in the sibilant range (6000-9000 Hz) so that the compressor, with fast attack and release settings, kicks in on the peaks caused by this annoying problem generally localized in this spectrum range.

Look-Ahead Function Although many compressors are fast enough to respond to sudden volume changes, some transients in the signal are so fast that trying to compress them can cause very unmusical effects. Imagine there is a fly zipping past your dish and you clap your hands in an attempt to catch it, but it’s too late. Now, what if you were able to ‘predict the future’ and know up front where the fly is going to land? By introducing a few millisecond delay (usually 2 to 5 ms) between the signal sent to the compressor sidechain and the actual signal to be processed, every transient will be detected and processed without risk of artefacts. The look-ahead option indeed allows to use less extreme attack times, with fewer unwanted effects. In the digital domain, auto delay compensation can make up for this delay.


Parallel Compression This is a common trick used in mixing on vocals, drums, and rhythm section, but it appears to have found more fertile ground in mastering. The theory behind this technique is very simple: instead of compressing the highest peaks of the signal to add body and sustain to the sound, yet often losing the attack, a copy of the signal is compressed and then blended back in with the main signal. This allows to leave the original dynamics unchanged, while adding volume to subtle details that are often left overlooked. When should we use a compressor in mastering? Here are some scenarios: • • • • • •

To make dynamics more homogeneous To attenuate the peaks and increase the general RMS of the track To ‘create’ a more compact and powerful sound To exploit the euphonic characteristics (e.g.: tube compressors) (On some stems) To ‘remodel’ the envelopes To create a sense of movement in static genres (digital electronic or synth music).

TIP: Never use a compressor only because you have it! And do not compress the left and right channels separately in order to avoid an image-shifting in the stereo image of the sound.

Multi-Band Compression In this type of compressors, the input signal is divided by dedicated filters into separate frequency bands which are then compressed individually. Each compressor operates completely independently. The bands are then mixed and sent to the output gain control. This type of processor tends to ‘re-equalize’ the sound much more than broadband compressors and should be used carefully so as not to drastically change the balance of the mix.

Limiting Limiters are used in mastering for two specific functions: to prevent a signal from going above a set level (the ‘ceiling’) and to lift the final level of a mix. Although well-designed limiters are very transparent-sounding, it is always important not to be misled by the perception of a higher volume and to listen with a critical ear in order to see if some important detail is missing.


TITANIUM

Mastering Multi-Band Compressor


The classic sound of analog multi-band compression Mastering Engineers may not always agree on the use of multi-band compression in mastering, because these compressors not only affect the dynamics, but also have a strong impact on the mix timbre. However, multi-band compression can be applied to specific genres, such as pop, rock, EDM and hip-hop for artistic effect, especially in the low frequency range, to knit the bass and kick drum all together or give a static mix more character and life. Titanium is an excellent creative and corrective tool. You can freely choose whether you want to use the natural response of its compression or to keep the dynamics under control as transparently as possible. Let’s become familiar with the interface.

As you can see, you can choose among three different configurations. The first one, TITANIUM 1B (One Band), is not a multi-band device, but a stereo version of Titanium. The second one, TITANIUM 2B (Two Bands), splits the spectrum into two regions. There is a control knob called X-Over High, which is a crossover filter for high frequencies. Available values are: 1.2 kHz / 2 kHz / 3.2 kHz / 4 kHz / 4.8 kHz and 6 kHz. At 3.2 kHz, for instance, the High Band will operate on the frequency region above 3200 Hz, the Mid Band on the frequencies below. The ON switches on the bands are very useful because, when off, they disable the reference band in order to better hear what the complementary band is working on. Main controls (on each band in the two- and three-band compressor): Threshold (0/-48): This knob sets the level above which the compressor starts acting.


Ratio: Available values range from a minimum of 1.5:1 to a maximum of 10:1. Attack: This knob sets the attack times. No values, you go by ear! From Fast to Slow. Release: This switch sets the release times. Again, no values, from Fast to Slow. Gain (0-25 dB): This control compensates for the Gain Reduction carried out by the compressor. Each band has its set of meters, which are very useful. Meters I / O / R: They allow to monitor the Input, Output and Reduction levels. The third configuration is TITANIUM 3B (Three Bands). Now you can apply three different compressions to three different frequency spectrum areas. The main controls are the same, with two additional knobs: a second crossover filter and a frequency multiplier (f x 4) which, once enabled, allows the second crossover to have the bracketed frequencies as reference values. The second crossover allows to separate the mid-frequency band from the low-frequency band; selectable frequencies (in Hz) are: 60 (240) / 100 (400) / 150 (600) / 200 (800) / 250 (1kHz) / 300 (1.2 kHz). Here are a few ideas on how to use Titanium in a hypothetical mastering session starting from specific stems. The following settings have been determined based on the audio files we have been working on to give you these suggestions. The options are endless, but it is easy and quick to see how many different paths can be taken! 3B on Triplet-Feel Drum Stem.

Drum Stem


This drum stem was essentially static and we had some annoying resonance from the kick drum in the low frequency range (the kick drum tail was too long and the volume too loud). Although pleasant, the timbre needed more control and power. The Low Band (from about 200 Hz downwards) was set to such values as to emphasize the kick drum attack (residing around 70-75 Hz) and ‘squash’ the tail for a stronger impact. The Mid Band (ranging from about 200 Hz to 4 kHz) steps in slightly and, with moderate attack and release times, aims to ‘massage’ the kick drum percussive mid-range, making it softer. The band from 4 kHz upwards was configured to emphasize the snare drum ‘snap’ and the kick drum beating, with a deliberately long and out-of-time release compared to the drummer’s triplet groove; the use of the compressor is more evident and especially audible on cymbal tails (creative use). TITANIUM 1B, followed by Amber Mastering EQ (see page TBD), on a rhythmic jazz vocal group.

Jazz Rhythmic Vocals

We used the Titanium Stereo Compressor to emphasize the labial and dental consonants in some syllables, in particular ‘Paa’, ‘Baa’ and ‘Daat’, which needed more character. Corrections with Amber: • 1 dB at 170 Hz to give a deeper sense of warmth. Low-mid Q • - 1 dB at 2300 Hz to soften the sound. Low-mid Q • + 1 dB at 5000 Hz where vocals have gained more detail. Low-mid Q • - 10 dB at 50Hz (although Amber has a much wider response – see specific section) to attenuate the low-frequency region of the spectrum, where other elements of the mix reside. • The preamp section was left on to add colour and three-dimensionality. TITANIUM 2B on Electric Guitar Stereo Groups.


Stereo Guitar Riff

Titanium works greatly on electric guitar groups, both to control the attack and add more ‘grit’, without necessarily resorting to a static EQ or effectively compress the signal. These settings have turned a rather muddy group of three Fender Stratocasters into a roaring lion (we operated Titanium from 2 kHz downwards, trying to emphasize the content of the complementary band, kind of a reverse approach). TITANIUM 2B Conga Stem.

Conga Stem

Congas tend to be rich not only in harmonic content, but also in the frequency range 125-400 Hz. To work on this group, we chose to use Titanium mostly below 1200 Hz: a relatively high ratio, fast attack and slow release times. The conga body was extremely robust and the dynamics a bit variable (the beats were not clearly audible in the mix and blended in too much with the drum sound). The compressor was set from 1200 Hz upwards so as to emphasize the strokes made with the palm of the hand. The result is a fuller, clearer and more defined sound. Note: Indian percussion instruments produce a similar but more ‘sustained’ sound. This is the case of tabla drums, where transients should be clear, clean and well-defined because they correspond to specific beats in the traditional forms of Indian music, such as the insanely fast ‘tabla relas’, where a wide variety of stroke patterns are used. If you happened to work on a track with this instrument, which is more and more used in jazz fusion, Titanium would serve as an excellent tool to model the sound even better than an Eq!


TITANIUM 3B on the Stereo Mix.

This is the most common way to use a multi-band compressor. Some Mastering Engineers use it before applying a broadband compressor or an EQ preceding another compressor, others do not use it at all so as not to affect the original dynamics of a track of even half a dB. Many jazz artists do not like the effect of multi-band compression on their music. It is up to those who employ this tool to find a suitable use for it. The settings for this example are much more conservative than the previous cases. Ratios were all kept at the minimum value and we tried to never exceed, in the three bands, a gain reduction of 2 dB. Our aim was only to ‘glue’ the mix without depriving transients of attack or character (they were only slightly controlled in the band from 2 kHz upwards and emphasized in the mid band from 400 Hz to 2 kHz to make them slightly more visible). The low range was made more compact by using slower attack and release times with a deeper threshold to ‘embrace’ a wider portion of signal. Finally, a light stage of make-up gain was added.


Aquamarine Mastering Compressor


One of the most coveted mastering compressors, Aquamarine is a triumph of sound and energy. The hardware model on which the plug-in is based is currently one of the most well-respected mastering compressors of all time. Also great as a mixing device, this compressor can model the sound both softly as well as forcefully, based on its users’ tastes. Compared to the original model, a ‘Mix’ knob has been added to adjust the mix, in parallel compression, between the dry and processed signal. Moreover, the attack times achievable in the ‘Discrete’ stage are quite close to 0, thereby allowing very fast response times. The plug-in response also varies based on the use of the gain staging (see chapter 8), so you should keep this rule in mind if you want to get the most out of this tool.

The preamp section allows to add up to three different ‘colours’ to the sound, from the most noticeable ‘Bronze’, through a mid-way ‘Silver’, to the less coloured ‘Gold’, which has a linear response even at low frequencies. The first compression step is the ‘Opto’ stage, that emulates the response of the original optical attenuator. The controls are Opto Threshold (Aquamarine starts with a -1 dB threshold) and Opto Gain, which controls the amount of make-up gain. Although it looks easy, multiple combinations are possible, if you also consider that you can influence the compressor response by adjusting the Input Trim. If you do not need it, this section can be bypassed by using the dedicated switch. The second step is the ‘Discrete’ stage, that emulates the compressor VCA. Its behaviour depends on whether it is complementary to the first stage or it is used alone (in this regard, you can refer to the section ‘Compression’ on page 42 about the different types of compressors). Ratio values are 1.2:1, 2:1, 3:1, 4:1, 6:1 and FLOOD (20:1). In order to better control the compressor response, the Sidechain switch cuts off frequencies below 90 Hz to limit the risk that compression is triggered by low frequencies. In the example below, you can see how Aquamarine was used with other processors in a short but ef-


fective mastering chain of an a cappella track. It is a kind of New Age music with a dreamy atmosphere, where spatial realism is essential. The only two elements are female vocals and reverb. However, do not be fooled by the apparent simplicity of this scenario, because having many spaces can lead to evident corrections, at the risk of unsettling its subtle balances. Although the mix was well-balanced, the dynamics were ‘unstable’ and the voice felt detached from the surrounding environment (imagine you have the singer in front of you and you feel like the ‘room’ is behind her). The first processor we used was Aquamarine. With the Sidechain enabled, we ‘examined’ how we could handle the unstable dynamics (the general volume of the mix was fine and we did not adjust the Input Trim too severely. As you can see from the figure below, the settings are quite conservative). We used the Opto section to ‘bridge the gap’ between the vocals and the reverb tail, without exceeding 2 dB of gain reduction.

We made the Discrete section achieve a maximum gain reduction of 1.5-2 dB on the evident peaks with a medium attack time (there were no particularly fast transients) and a quite fast release time, trying to make the process as transparent and progressive as possible. It is worth noting that the compressor’s general volume was slightly turned down so that the volume levels could be equal before and after compression. Do not try to add volume with compression. Other tools are more suitable for this purpose! The settings used for Green helped make the vocals more prominent and less sharp. The ‘air band’ was also emphasized (about 1 dB or so) to bring a further ‘sense of air’ and determine the density of the reverb. Since there were no cymbals, we could add detail without making the track sound too clear and bright (or simply unnatural). Before the final limiting stage, we used a Tape program for Nebula at 15ips, with a subtle drive (one of my favourites for this kind of arrangements, created by CDSoundMaster) producing a deeper and more three-dimensional sound and adding an additional good-working ‘character’.


This example is but a drop in the ocean of possibilities! Here we used compression before EQ, but I am sure there are plenty of Mastering Engineers who would have done just the opposite, it is just a matter of personal preference and approach. There is no absolute rule. Never underestimate the power of automation in mastering, especially when dealing with the macro-dynamics of a track (i.e. volume variations on long-middle passages or different track sections: Intro, Verse, Refrain, etc.). Compression does not always lead to optimal results! TIP: Leaving the thresholds at 0, use the Input Trim to bring the input volume up to such levels that the compressor starts working (it is better to work on the highest volume track sections), then reduce the input level until compression is no longer applied. This will be a good starting point for modelling the dynamics. Now you can start working on the threshold, attack and release parameters (if you want to use the Discrete section) to model the sound as you wish. Just keep in mind that huge gain reduction variations will probably be always audible, especially when release and attack times are not precisely set based on the dynamic content of the track. Only a digital mastering peak limiter, designed to be totally transparent, can apply 4-6 dB of gain reduction without this being audible. In any case, this would be useful to drive the track towards the target volume rather than reducing the perceived volume between the track sections.

For instance, with a volume difference of 6 dB or more between two track sections, you would expect the effects of compression to be audible in this transition, especially if you have not used the highest volume sections as a reference for setting the compressor. A song can be considered as a huge envelope; narrowing down the dynamic range of the refrain (based only on its dynamics) of 3 dB and making it up with the make-up gain means reducing by the same amount the perceived volume difference (let’s say, on a ‘large scale’) between this refrain and a lower volume verse. You should always try to return to unity gain after going into compression by making good use of the release time. Having a fixed compression value of 3-4 dB would make little sense because it is just like turning the volume knob down by the same amount. It is better to use two compressors chained in a series to achieve higher levels of gain reduction and then resort to the automation of the general gain of the track (be careful not to use the master fader for this purpose. There is a good free plug-in, the Sonalksis FreeG, that you can use, for instance, as the last tool in the chain before the final limiting stage).


Chapter 4

Stereo Processing

T

his technique is commonly used in mastering to give a track more three-dimensionality and ‘air’, widen the image, tighten it up, or correct phase issues (especially in the low frequency region of the spectrum).

There are different ways to work on the stereo component of a track (i.e. on all the differences between the right and left channels of the mix… Remember the Side component?), but it is useless to use stereo processors on essentially monaural mixes in the hope of adding a little extra ‘size’ to them. There is simply not enough ‘raw material’ to feed these processors, irrespective of the way they work (some use the principles of psychoacoustics, others work on phase relationships or the volumes of the Mid and Side channels and so on…). The additive effect of stereo expanders is generally desirable and euphonic, but only when the original stereo image is strong and compact and can be further emphasized without loss of detail and impact. Some of these processors also provide an additional stage such as, for instance, the harmonic exciter. The corrective effect of stereo processors (especially multi-band processors) is particularly useful in mastering. A lot of mixes have phase problems in the low frequencies around 35-65 Hz that become extremely clear on headphones and cause the mix to be highly problematic on monophonic systems. Making this frequency interval ‘mono’ (or reducing its width) can be a good solution. Some mixes are too wide and lose impact. Narrowing the image can help restore a sense of cohesion, but sometimes changing the arrangement and the position of some elements in the mix gives far better results. The best recipe is to experiment and always check the mix by monitoring in mono.


Chapter 5

Sample Rate-Bit Depth Dithering

S

ample rate and bit depth are two fundamental, but often overlooked, concepts governing the audio world. A third one, dither, actually follows them, but we will deal with it later.

The sample rate can be defined as the ‘rate’ at which a digital system takes a series of snapshots (samples) of a waveform recorded within a time frame. Ideally, the more samples you take, the closer the sound will be to the original. If you consider the film or cartoon frames, designed to create the illusion of motion in the audience, it is easy to see how frames should be captured fast enough to create the illusion of smooth motion. Otherwise the result will be approximate. The higher the sample rate, the more accurately very high frequencies (having very short wavelengths) can be captured. Without focusing too much on technicalities, it is important to know that, according to the Nyquist


theorem, a signal is properly sampled without loss of information if the sampling frequency is equal to or greater than twice the maximum frequency of the signal spectrum. Undersampling an analog signal produces frequencies that were not within the original signal, which can be perceived as a sort of metallic sound. This phenomenon is called ‘aliasing’ and must be avoided or it will have a significant impact on the quality of the sound. The bit depth determines the range of amplitude values that can be stored for each sample and its potential dynamic range, from 0 dBFS downwards. In round figures each bit provides a dynamic range contribution of 6 dB. This means that a 16-bit recording corresponds to a dynamic range of about 96 dB with 65,536 possible levels, and a 24-bit one offers 144 dB with nearly 17 million possible levels (16,777,216). Imagine you are walking barefoot on a fine wet sandy beach. You leave your footprint in the sand and each fine grain, which will probably get trapped in the folds of your skin leaving a mark, can be considered as a high-resolution sample. What if you were walking on a pebble beach? Aside from the pain, your footprint would never be like the footprint left in the soft clay sand and would not be as deep either. What you would see is just a vague approximation of your foot shape! The bit depth of an audio signal is originally determined by quality at the recording stage. Today recording at 24 bit is the standard, while the 16 bit is considered as a target value, when no further processing of the signal is required. So, when does this value change? Even at the slightest processing, because algorithms apply functions and multiplication to the signal, whose bit depth is immediately extended. Any time a calculation takes place, even with a variation of half a dB of gain, this change occurs. At some point, however, after several chained processing stages, the least significant bits have to be thrown away until only the bits containing useful information remain (as in the case of the final bounce to 16 bit). This procedure, called ‘truncation’, simply lops off (truncates) these bits of the signal. A major problem caused by this procedure is the generation of non-linear distortion and harmonics introduced by a ‘truncation error’ correlated with the main signal. Needless to say, this quantisation error ‘sounds’ horrible to the ear.

1 kHz - 24 bit

1 kHz - 16 bit


Let’s suppose that you have a sine wave at 24 bit and reduce the bit depth by simply truncating it: you will have quite a problematic scenario. The main signal is still there, but the amount of distortion introduced has literally obstructed its faithfulness to the original signal. Here is where dither, one of the most widely misunderstood concepts in mastering, enters the scene. Dither is noise. How can noise help remove the quantisation distortion and enhance the dynamic range? We need to stress that noise cannot always be used as dithering noise, because the latter is generated by considering statistical probabilities which aim to decorrelate the quantisation error from the main signal through a deterioration of the signal/noise ratio. In doing so, unwanted distortions are removed. Just try this simple exercise to get a clear idea of how important this step is. Starting from a 24-bit file, attenuate the signal by 80 dB. The sound will become almost inaudible. Save two 16-bit versions with and without dither. Reset and normalize them. Now listen: the undithered version will be distorted and damaged, with the signal sometimes disappearing below the 16-bit quantisation level, where only absolute digital silence reigns. The dithered version will be rich in noise (of course, we have deteriorated the signal/noise ratio by 80 dB!), but the main signal will be very easy to discern and will not contain the same distortion components as the first file. There are several dithering algorithms that work properly and each one of them colours the sound differently. With time and experience, you should be able to accurately select the type of dithering to use, depending on what you are trying to achieve. The simplest algorithm is the TPDF (Triangular Probability Distribution Function). Various algorithms, for instance, use ‘noise-shaping’ techniques to shift noise into a less audible frequency range, where the human ear is less sensitive, in order to reduce subjective noise perception. In the figure below, you can see some of the many possibilities a Mastering Engineer is often faced with. Again, your ears will be the final judge.



Generally, you need to be careful when using dithering algorithms with a strong boost to high frequencies. Although dithering levels are usually very low and hard to hear (and although the addition of dither does not correspond to an increase of 6 dB of noise each time, since algorithms are not interrelated), dithering could eventually become audible, particularly in the cymbal or reverb tails or when instrumentation is less dense. Can dithering help ‘extend’ the dynamic range? Here again an example serves to illustrate the point. We started from a remaster at 96 kHz / 24 bit of ‘Thriller’ by Michael Jackson and reduced the general gain of the song by 100 dB, a value below a standard 16-bit file capacity (our intention is to see whether, following conversion to 16 bit, which does not have this dynamic range, we can ‘gain’ a few extra dBs of signal by using dithering, enabling a ‘deeper listening’ than a 16-bit file would allow). We exported two 16-bit versions without and with dither (TPDF). We normalized them with a maximum peak of 0 dBFS. What we see is that the ‘normalized’ undithered version does not change. Absolute digital silence. The file is not able to quantize the information.

MJ - Thriller No DITH - 100dB Normalized


The dithered version enables us to include even more digital information. What we can clearly hear, although at a lower volume than the noise introduced, is Michael’s voice, the kick drum, the guitar and other elements of the mix. In a few words, we were able to ‘listen deeper than a 16-bit file would allow’.

MJ - Thriller - 100dB TPDF Normalized

So, when should dither be applied? It is very commonly said that dither should be used only once, before the final bounce to 16 bit. This is true (and useful) only if you always work in digital, saving at 32-bit floating point all along the way. However, it is often used more than once along the production process. Let’s see the most common cases: In the mixing stage (ITB at 24 bit), dither (24 bit) is usually applied to the aux sends to hardware processors if a plug-in has been previously used (as we have seen, plug-ins increase the bit depth of a digital signal any time a calculation takes place). In case you use a reverb/digital processor, it would be better to make sure that dither is also applied to its output so as to prevent the signal from being affected by quantisation error which, although low, would still be there.

During the bounce, you need dither at 16 bit (after the Sample Rate Conversion) for CD production, streaming, downloading or the iTunes Standard Master (24 bit for Mastered for iTunes™ Encoding).


“Mastering is not the icing on the cake, or even the cherry on top of it. Mastering should really be the final polish on that cherry, so that it helps everything in the cake mix to stand out in perfect detail. I’ve enjoyed albums for years and then heard a remastered version that is a revelation to listen to, simply because you can suddenly hear a host of previously submerged details. Don’t assume that mastering always involves boosting the low and top ends and applying multi-band compression – just listen, then decide if there’s something that needs enhancing. If you find yourself piling on more than a few dBs of EQ or compression at the mastering stage there’s something wrong with the mix, so unless you have no choice, go back to the mix first. Also, unless your ears have a very good memory, refer every half an hour or so to a selection of carefully chosen commercial tracks that you admire in a similar genre, to avoid going too far at the mastering stage – it’s easy to end up with a harsh sound that’s initially pleasing, but which soon becomes tiring to listen to.” Martin Walker – Author/Composer (Sound On Sound Magazine / Yew Tree Magic)


Chapter 6

DDP and Track Sequencing

O

ne of the most enjoyable steps in mastering, after tracks have been thoroughly subjected to the labor limae, is when the end product is produced.

Mastering not only means ‘messing around’ with a lot of awesome tools and their little coloured lights, but also never losing sight of the overall message of the CD. Mastering a single track is much easier because you have more freedom. What is more complicated (but also much more exciting) is mastering a track bearing in mind the album’s genre and background, the features all tracks have in common, such as the main instruments, voice, stereo image, same perceived volume between the tracks, climax, etc. Unfortunately, the distribution model adopted in the last years (based on downloading single tracks ‘taken out of context’), with rare exceptions, has come at the expense of creating a ‘Glass Master’ for the replication process. This feature, which has actually become a rule in the field of independent record industry, has distorted the idea of the album as a collection of ‘stories, colours and pictures’ with specific start and end points. Up to a decade ago, a single track acted at least as a ‘lure’ to sell the whole album, but now this has changed. The single track itself has become the end product! There are several programs and DAWs that allow to create a Red Book Compliant mastering from single tracks. So how to proceed? First, you should clearly recall the track sequence. Then, you will be able to place the right pauses between the tracks: remember that silence in music is as important as notes.


Now listen with your eyes closed and ask yourself some questions: •What is the feeling you want to convey? Energy (e.g. with short pauses between each track not giving respite and following the natural rhythm of the tracks)? Calm and peace (long and meditative fade-outs, long pauses, etc.)? •Is it a live CD? So you need to find the right points where to ‘separate’ the tracks from one another as naturally as possible. Generally the audience clapping between two pieces is the point where a track ends and the next starts, but it could also be a start with a drum solo (on two separate tracks on the album’s CD version but put together in a file for digital downloading, for instance), the beginning of a singer’s speech in the case of integral live music and so on. Once this step is completed, you need to export the DDP image file. What is it? The DDP (Disc Description Protocol) contains and describes a set of data necessary to create an error-free Glass Master for reliable CD duplication and printing. It is always recommended to send a DDP by e-mail rather than delivering a mastered CD: this could contain damaged sectors that may compromise not only the duplication, but also the playing of the CD in some players. By checking the ‘MD5 Checksum’ (one of the files generated as a result of the export) included in the DDP, you can easily detect any loss of data during the transfer and try it again. It is based on comparing the string produced by the file sent with the string produced by the file received. A DDP file includes: •Files called DDPID, DDPMS, SUBCODES describing the CD •The AUDIO.DAT file including the audio files •The MD5-Checksum.md5 file that we mentioned before •All metadata composing the CD Text Information (correctly decoded by most players), e.g.: Composer’s Name, Artist’s Name, Track Titles, ISRC, EAN/UPC codes, etc. The ISRC (International Standard Recording Code), defined by ISO 3901, is a unique identification system for sound recordings and consists of a 12-character alphanumeric string (in the form: XX – Y12 – 12 – 12345). The first two characters identify the country code: IT for Italy, GB for the UK, US for the USA and so on. The next three digits identify the registrant code, the next two numbers are the year of assignment and the final part is the code assigned to the track. The ISRC is issued by the National ISRC Agencies. Here is the official list: http://isrc.ifpi.org/en/contact/national-agency-contacts Should you worry about using ISRC codes?


If you release a CD on your own independent record label and you expect to earn something from it, then the answer is yes. Using an ISRC has several pros. ISRC codes are only required to release, distribute, sell or broadcast your music on a radio station (on the web or not). When an ISRC is embedded into your track and identifies it as belonging to you, it will help track your sales and the audience ratings in streaming or on the radio. However, ISRC codes are only required for the final mastered (stereo, not multitrack) versions of your tracks, which will be sold and distributed. No code is required for demos and other ‘domestic’ recordings. You also need to know that any further version of the final track intended for distribution, which has been rearranged, edited, recorded a cappella, remixed and so on, must have its own code.


Chapter 7

DC Offset

T

he concept of DC offset needs to be mentioned in any discussion of mastering. A waveform with a DC offset has different amounts of signals in the positive and negative fields: in a few words, its centre will never be equal to 0 but will be slightly more or less than this (positive or negative), as a result of a component of the recording chain introducing this problem before the signal is converted from analog to digital form. Why could this be a problem? First, a signal with a DC offset will never reach its maximum possible volume before a digital clip occurs, or even if only normalized. This happens because DC offset uses up useful headroom for the signal, whose maximum peak is closer to 0 dBFS than in normal conditions. DC offset may also introduce some slight low volume distortion, which can become audible during the file processing and cause annoying clicks at the beginning or end of each file (or during editing). You can use a high-pass filter to get rid of this problem and remove low frequencies below 20/30 Hz. Moreover, there are a lot of detection and correction software that can be used to pre-emptively solve this problem. Some plug-ins, including sub-harmonic exciters, may also introduce DC offset. It is very important to remove it.


Chapter 8

Gain Staging. What is it and why is it important?

G

ain staging is the process of managing the level in each component of the recording and processing chain (from preamp to the master bus sum), in order to optimize the signal-to-noise ratio without incurring distortion. The aim is to have plenty of headroom that, even in case of unexpected peaks, provides a good safety margin. This problem is much more pronounced in the analog domain than in the digital domain, where there is no presence of background electronic noise and the only impassable limit is 0 dBFS. However, the best approach is not to make this distinction and just use the analog gain staging in the digital domain. Indeed, we are working more and more with hybrid systems that somewhere in the chain combine the use of an analog component (e.g. an EQ or a compressor). Besides, many plug-ins (particularly those emulating the behaviour and curves of hardware units) are generally inaccurate when they receive tons of gains and can introduce unpleasant artefacts. Many home recording users are ‘scared’ of conservative input volumes because they fear they will not be able to achieve appropriate final levels once the mix is completed. This shows not only a lack of theory underpinning recording practice, but also some inexperience in the production process and its steps. How many times do you hear people say, ‘If I set a low volume while recording, how can I get the same sound as if I were listening to a CD’? But then, if you are reading this guide, you probably already know the right answer to this.


If a signal at 0 VU on one of the meters appears on an analog console, you should bear in mind that there is roughly 20 dB of headroom left before incurring distortion. Peak levels may exceed 0 VU without compromising on sound quality because analog circuitry naturally smooths out these peaks (an effect and colour looked for by many). Remember that VU meters do not measure a signal’s peak levels, but its RMS; if you want to keep transients intact, you have to keep peaks at a maximum value of 0 VU. If you have a tape recorder with a drum track at 0 VU (RMS), don’t be surprised that peak LEDs are red and that some degree of compression is already being applied, since this instrument can have a dynamic range of more than 18-20 dB! To give you a concrete example, 0 VU corresponds, in the digital domain at 24 bit, to around -18 dBFS. With current converters and DAW resolution, it is perfectly normal to record, for instance, a snare drum with very low RMS values and peaks at -14/12 dBFS. Since a value of -10/12 dBFS on a digital scale corresponds to around the nominal peak value on an analog console, there is no need to record trying to reach the ceiling of 0 dBFS. [Fig. Reference Levels] If you are working ITB with third-party samples (often sold normalized), it is good practice to use a gain plug-in as a first insert in the DAW (simulating the input gain of a channel strip of a mixer) and bring the level to more conservative values, which are more ‘compatible’, so to speak, with other plug-ins or external hardware, when used. If it sounds too low, you just need to turn up the loudspeaker volume, don’t you? To sum up: If you use the highest peak sound as a reference level and set it to such values as to make it reach a maximum peak level of -18/-12 dBFS on a digital scale you will not have any problems. In mastering, gain staging refers to the processing chain as the signal is ‘processed’. However, there is another scenario where, if the level is too high, some problems may arise, even if you worked in digital and used a limiter to chop off everything above 0 dBFS: inter-sample peaks. It may be the case that, while reconstructing a continuous waveform from the digital information, a converter produces a signal with a higher peak than the value of the samples ‘squashed’ against the limiter ceiling. The effect this has on the sound you will hear depends on multiple factors, including the quality of the DAC and the way it is designed but, although sometimes inaudible, many other times some distortion may be apparent. In any case, you’d better be ‘down-to-earth’!



Chapter 9

The interviews Michael Angel CDSoundMaster What is the true role of a mastering engineer? Michael Angel: This is a great question! The role of the Mastering Engineer has changed so many times over the years, but the fundamentals remains the same; the Mastering Engineer is to be the last human decision maker in the recording process who preserves the best elements of the performance and its recording, while preventing the technical and sonic issues in translating these to a master and the subsequent mass-produced copies. In earlier days, the Mastering Engineer had a crucial role in making sure that the structure of frequencies, balance, volume, and dynamics were under control so that the needle would not dance out of its groove on a vinyl album. Control of the bass frequencies was vital, and typically mastering tools were designed to limit extreme changes to low frequencies and to place these frequencies towards the center of early stereo vinyl recordings. In the boom of the record industry from the 70’s to the early 80’s, there was a shift in thinking in the entire structure of the recording industry. What used to be a tightly run ship was being seen more and more for the vital artistic role of each engineer in the process. The Mastering Engineer was often previously used for redundant, even menial tasks, transfers, edits, and preparations for masters, with the hope of inching their way up the ladder to primary Mix Engineer, or Balance Engineer. The amazingly overlooked final stage was often left to a machine-room mentality. Some of the most talented and gifted engineers were often sitting side by side in a room of folks with headphones wondering if


they were served a dupe of a dupe of a dupe of a multi-million dollar album. Incorrectly marked or incorrectly cued masters were pressed with unintentional wide stereo pans that were absolute process errors and later attributed to the artistry of the bands and engineers. Some of this changed in the handing over of the industry from the 50’s-60’s organization of creatives-meet-scientists, into a messy new aesthetic of “everyone is a producer” idealism. Because the mix had to survive the process from tape tracks to tape masters to dupe masters, on and on, it was crucial that the mix account for each stage in the process (often adding vast amounts of high end gain to account for later high end loss at each generation dupe). The process has always been about balancing one trick after the next. The engineers of the 50’s were geniuses at mastering speed, technique, and stream-lined processes and the Mastering Engineer had to make it stick. The 50’s into the 60’s saw the first truly stunning potential for capturing sound but the process still had a lot of loss from one stage to the next. Engineers were trained in-house with very precise rules. In some cases, there were documents about how far microphones had to be placed from instruments and consoles were marked with maximum fader positions not to be crossed. The “balance” engineer was the top of the food chain for a long time. The more complex the mix could become, and the finer the perfection of late 70’s, early 80’s tape machines had become, the more value was placed on the protection of final masters once again. As albums, cassettes, and yes 8-tracks made their way to a new frontier, the Compact Disc, the Mastering Engineer became the new top of the audio food chain in the careful transfer of countless treasures of 1/4” and 1/2” masters into digital. This required restoration in some cases, and revealed hidden weaknesses to recordings, splices that could be heard for the first time, and even revealed certain artifacts in the high-end hardware that had been so well suited to the pleasant masking of such sounds before digital clarity. Remasters became a critical new path to huge new sales for record labels, featuring the hard work of Mastering Engineers, further giving a prominence to the important position that had always been there, but for different critical purposes. Although there have always been rare renegades in the industry, the financials almost completely eliminated the outside world from the truly professional sounding mixes and masters, and the Mastering Engineer was even able to present their own independent facilities for the sole purpose of serving excellence in their one particular field. The first generation of truly independent Mastering Engineers had worked hands on with the most famous names in history at the most prominent labels, and they earned their place at an hourly rate of $300-1500/ hour. As insane as this may sound to many, even a $15,000 mastering job on an album guaranteed to sell 500,000 copies would come to $.03 per unit. When comparing what the distributor does for roughly $2-4 per disc, I see quite a value in that 1/33 of a dollar per unit! As the digital world gave way to the first ‘affordable’ digital audio recording systems on personal computers, some as inexpensive as $4000-$8000 for 8 tracks of digital audio on a 486 with an $800 sequencing program that ran 8 tracks of digital audio (yes, thank your lucky stars you can now own a $100,000 recording studio for the price of a crappy used $150 laptop), there was a new era of ‘home’ studios beyond the previous demo setups of 4-track cassettes and Fostex 1/4” reel to reels. The Mastering Engineer was now, for the first time, having to start considering what to do about a new side to the market- the home-made independent studio recordings. This included a need to apply a heavy handed approach to sub-par, less professional mixes that required technical notes and redoing


certain elements of the mixes before supplying usable pre-masters. The CD required a glass master created off of a Sony edited master, replicated in minimum quantities of 1000 units. It was a few years later that affordable home studio CD Burners existed, with a 2X CD-R running $500 minimum, and individual blank CD’s costing $12 each. Technically, this opened doors to the first ‘Mastering Engineers’ that had no technical requirements, had not climbed the ladder at a major facility, and whose only opt-in was that they owned a CD-Rom and could burn a red-book master with their own computer. I was blessed to arrive just in time to have learned the true traditional craft of Mastering before this became the norm. I had one foot in the door of the classic way of doing things while watching the price drop radically to the rest of the music world, which was in itself a blessing to my bands that I recorded, my budding project studios and growing Record Label and BMI Publishing Company. The first computer software plug-ins to come along that featured do-it-yourself mastering settings were an industry joke for several years. Regardless of whether some of them were actually pretty decent in their audio quality, the premise that someone could come to understand the proper way to use them without having learned the traditional path to an intricate and complex technical career was ridiculous. Early online forums were born in the mid 90’s where discussions took place over the new trends in home-remedy mastering. It was understood in the industry that risking one’s own mastering, especially without any decent sounding home recording setup remedies as of yet, was considered as risky as do-it-yourself brain surgery. As sound cards improved and 16-bit audio was actually a step above 12-bit mini-disc recordings, some home setups were actually starting to sound competitive with commercial facilities. You may think this sounds hard to believe, but I actually owned the first pro-sumer 20 Bit audio card before there was such a thing as a computer that could smoothly operate 24 bit playback live from its pci buss. Somewhere between Napster and torrents, affordable Pentium computers, and excellent articles about how to use SM57’s and the world’s first affordable semi-pro mixers led to a new revolution in home recording. The small commercial facility was finally born, and skilled Mastering Engineers breathed a sigh of relief that more aspiring home producers were coming into their element as a new breed of skilled-and-learning Mix Engineers who trusted the expertise of Mastering Engineers to prepare the important translation into final glass masters. Ah, but panic was about to spring forth at the explosion of the plug-in software industry. It seems that there was a large profit in explaining to the average beginner recording hobbyist that they had to have their own mastering tools. For those who were eager to learn the trade, it was a wonderful time for exchange, learning, tutorials and the birth of the modern recording forums, but it was equally a time of novice panic and professional audio nightmares, as do-it-yourself audio surgery spawned from a commercial marketing plan to sell professional audio tools to an eager crowd that was not yet ready to understand how they worked. “What kind of stuff do I need to master my album with?” “What is MidSide and do I have to have it before my band releases our album?” “Is it ok to use the Eq in Cubase on the whole mix or do I have to have the T-Racks for it to go on CD?” “Where can I read a red book and what is an ISRC?” All of a sudden large software developers had spurred more than an interest in


recording music on increasingly powerful computers. People were terribly concerned that they better learn how to be a Mastering Engineer and quickly, or else! From the mid 90’s to circa 2005, Mastering had peaked as an exclusive niche that was left to the most skilled audio engineers who had earned their way to the top. The Mastering Engineer had to also consider flexible real-world pricing models for the new serious recording musicians who realized that there was a vital role to be played between the inspiration to record one’s own music and the serious investment of releasing hard copies that needed to be professionally prepared for replication. Mastering came to include online service, file transfers, and online communication in place of expensive in-person sessions. The real transformation to today’s model didn’t fully take effect until the success of iTunes guaranteed that the new era of recording was yet another drop in quality; the mp3. Now, the Mastering Engineer, having ‘mastered’ the changes in direction from lacquer to tape reels, then to CD’s and improving digital, to a now inferior compressed data that had been so improperly researched before taking hold that literally the bottom dropped out in the quality of preservation of priceless masters. Thousands of original platinum recordings were ‘ripped’ to 128 kbps mp3’s with little-to-no skill, and without the enlisting of the Mastering Engineer. Once again, the short-sighted record industry had missed the mark in assuming that “digital is digital” and people noticed all sorts of artifacts from rumbling to digital overs, shifting in stereo field to weird warbling artifacts. Re-enters the Mastering Engineer once again to figure out the minimum necessary quality to reproduce 16 bit, 44.1 kHz lossless PCM Wav files into mp3 format. Which encoding sounded best? What made for a plausible rendering? Ah, the Mastering Engineer’s job is never done, rarely understood, always poorly mimicked, and never the same for more than a decade. So, we have seen many choices to improve audio go by the wayside. The use of a dynamically superior AUDIO DVD 24 Bit, 96 kHz format was never established (though many hobbyists record at 96 kHz!) The first generation HD converters still sound better than the most expensive deeper bit converters of today, with parts that are no longer possible to source. The average user is listening with earbuds that cost $.50 to manufacture, or laptop speakers barely capable of ½ the listening spectrum compared to the average home stereo that easily matched 80% of the master’s actual quality playback. Today’s Mastering Engineer plays the role of trusting their ear and their nuance in what makes the right master for the job, but also in educating the potential client to raise the bar on hobbyist-goes-instant online distribution, while serving files digitally and possibly in multiple formats. It is a balance of excellence in sound in an industry that is now flooded with amateurs armed with sophisticated, often pirated software that easily does the job of the finest analog gear (not all plug-ins are equal in quality, but they certainly bring more benefits to sound quality than the areas where they are lacking). The professional Mastering Engineer plays a lonely role in maintaining their value at a fraction of the hourly they need to earn while competing with those who have a fancy website offering $10/song. I imagine that the next step for the Mastering Engineer can live another several years, perhaps even a decade, offering professional advice to those learning to mix and record, and standing out in a pool of audio software developers selling their own designs that truly make a difference. What are, in your opinion, the myths/clichés surrounding mastering?


Michael Angel: There are so many wonderful myths, legends, and clichés to mastering. I am so happy to be given the question! One of my absolute favorite myths has just recently come up again just the other day. Yes, the Mastering Engineer must be able to identify potential issues in recordings that others may not notice, requiring stellar monitoring and sharp, attentive concentration and hearing. But, the tools of the Mastering Engineer have long been given the same status as the engineer themselves. True, these are amazing sounding pieces of hardware built with artistry and skill, but the myth of their absolute precision and perfect redundant rendering is an absolute myth. Well-maintained, high-end analog gear has abnormalities that have been sonically accepted, but largely unknown by the very industry that relies too heavily upon them. The finest Eq’s in the world that provide the cleanest signals for the most surgical of adjustments have massive areas of discord between the marked settings and the reality of the actual audio path. There are levels that vary by a large amount from the marked +/- gain, frequencies that truly need to be visually calibrated by an oscilloscope to match the actual markings. This leads to another myth of the same topic; detented settings are best for mastering because they are repeatable. I sit at a different opinion from most other Mastering Engineers because I have learned what I have from my own work with testing and analysis of the signal. There aren’t many people that can give as accurate an account of what the actual waveforms of the most sought-after gear actually look like. I have seen it all, at every level, through every test and using high-end cables and low-end cables. I have been amazed at the precision of what many would consider mid-grade devices because they aren’t for the wealthy-only, and I have been blown away at the amount of adjustment required from some of the most expensive devices that exist. The truth is that the Mastering Engineer is the most valuable asset to the use of such gear - knowing how to tweak levels until they sound correct, and for those that truly love preciseness, there are analytic tools on more modern, less classic mastering-era devices that let them know for certain that a dB cut with the most narrow slope “Q” is actually doing exactly that. For all of the other wonderful gear, I am thankful to have a means by which I have collected a truly priceless catalog of recordings of every setting of every device to be perfectly replicated in digital, never to need the wondering of whether it will behave the same the next time - finally it is a museum captured in its perfection inside the DAW! Another myth in mastering is that you can always fix it in the mastering stage. I learned this early in my friendship with my mastering mentor. It has long been a rumor that the musicians can rely on the Mixing Engineer to make up for their mistakes (yes it is done all the time, but it never accounts for doing a better job in the actual performance). It is further incorrect that the Mix Engineer can ‘almost’ get it and that is fine because the Mastering Engineer will fix it. The final stage is the result of those which come before, and the best master in the world is going to be a representation of just how masterfully the performance was captured to begin with. I suppose a cliché to mastering is certainly the loudness wars, so I will save that for that particular question! But, another cliché to mastering that I have seen come up often times, is “did they actually do anything?” or “I can’t tell what they did.” When I take on a mastering job, I make notes for the client in case they want to analyze what has been done step by step. Especially if they have done an amazing job with the mix already, the Mastering Engineer’s best role is to know when to leave things alone.


But often it has been cliché that expectations are so high, that the master will sound so significantly different from the mix that the client wonders if something was done at all. My personal philosophy, though influenced from some different thinking on the topic, has come to a personal sense of artistry; I care about the consistency of the entire album, but I like to make each track stand out as something unique in itself and not just consistent with other tracks. I also like to leave a little signature “more” or “better” than where it began. If it is absolutely perfect from the mix already, I will have run it through many paces, including loudness options, before choosing to leave it completely alone. What genre(s) do you mainly focus on? How does that affect your mastering decisions? Michael Angel: Another great question! I come from a classical training background. My parents were both professional opera singers. My mother was a touring solo opera singer and the Owner&Founder of Capitol City Opera Company. I was trained in voice and piano at a very young age and took to falling in love with drums shortly after. After years of study and playing in concert band, performing all over the world with the Atlanta Boy Choir and also with the Alumnist Men’s Chorus, receiving a Grammy for our performance and recording of Britain’s War Requiem, singing and playing percussion under Maestro Fletcher Wolfe and Conductor Robert Shaw, I played in rock bands, ran my first project studios, recorded blues bands on location, and was a percussion major in college until I eventually got my BA in English Literature with a triple minor in Art, Music, and Psychology. My background shaped my love for music and gave me an appreciation for excellence in many different genres. Performing in the world’s finest concert halls gave me the fascination of the live experience but also the impact of sound and ambience on music content. Working with monitors and headphones fighting for the perfect analog mix (all tracks by hand no automation - imagine that outside Cubase!) strengthened my love for the impact that recording has on the song and the artist. I loved the sound and writing of groups like Queen, who were amazing musicians but also made the studio an active member of the band. I loved the craft and recording of guitarists, punchy studio drums, and the rare absolutely skilled mix, but none of it ever wins out over my absolute love for the perfect choral blending of Maestro Fletcher Wolfe. He is the world’s finest Mix and Mastering Engineer, and he has no idea what a console or what difference a ribbon speaker would make. What he has done historically and even is still doing to this day, with the sound of blended voices in choral music, is the most remarkable musical expression that any human will ever accomplish. I place this at the top including a love for Rachmaninoff, Beethoven, Mozart, wonderful piano works in general, and the sound of a brilliant mind controlling a wonderfully trained precision of hand movements and expression. My love for drums and percussion makes it easy for me to love working with virtually every “pop” genre, and as a multi-instrumentalist, I have just enough skill to understand what the members of a band hope to see preserved in the final recording. The most challenging part in this decision-making process is not the genre as much as the content. I spend 12-15 hours working every single day in a tough economy, so I rarely take on new mastering projects right now, but even when I was seeking to master full time, I made the decision that I wanted to be true to myself and to my faith. I am a follower of Christ, and this is the most important element in anything that I do. We live in a time where this is very unpopular, especially within the confines of


an art form that is based on science. I am certain and steadfast in my belief in one true God, and for me this means that when I master someone’s music, I want them to feel proud of the seal of my work added to their hard work. If there is a strong ideological or content clash with whom I choose to be in my walk with Christ, I would do a disservice to the client and I would feel less of a man in dishonoring such an awesome God. Beyond that, I can get passionate about anything that sounds good to me in any musical style. I just love sound! What is your opinion about the Loudness War? Does “loudness” have an impact on actual sales or not? Michael Angel: Yes. Don’t worry - it would be completely out of character for me to give a one word answer - lol. Sadly, loudness does impact sales, and it seems that the revelation of improper loudness and how it harms the content came about from a guitar hero game! As I understand it, there was a release of a song that was before the final commercial master. Fans were able to compare the decent sonics of the mix before it was absolutely crushed for mass consumption and they were shocked at the horrible things that happened to the sound. The truth of the matter is that it does not have to be a quality trade-off. There are tricks to the trade and enough excellent mastering tools that a skilled engineer can supply the Mastering Engineer with a mix that is well-prepared for big, impactful sound without ruining all that is good and holy about a wide dynamic range. A short story about dynamics. The movie industry has gone through this process, and because the production experts in movies are always more adventurous than us purely audio guys, they tend to reach conclusions with their target audience quicker and with more savvy than the audio industry. Let’s just be humble enough to admit it. In the purely analog days, the soundtrack of a movie had the same limitations as our audio mixes. They were going to lose something at each dupe generation, so levels needed to be a certain percentage above the noise floor at each stage, and transfers needed to be made with as few retouches and retakes as possible, as there was loss to high frequencies and more artifacts the more times tape was run through playback. Hitting the master too hard and with too little dynamic impact was exhausting to the listener, but too wide a dynamic range led to some dreadful sounding, hissy parts to some great movies. Too much of the wrong kind of loud was painful in the theater setting and silent dialogue was an element killer. Studies were done and the average and peak listening levels were established by the audience themselves. This led to a set of standards for the actual dynamic range of film soundtrack still used today, or should I say over-used today. The variance in the quality of location and dialogue recording has just as much to do as volume. Weak recordings at pre-set, pre-configured dynamic settings make it almost impossible for the home playback of movies to have any logic. We can barely hear the subtle


scenes and turn it up to a point where we hear the lesser quality of our consumer amplifiers. Then, the huge car wreck or monster appears and we have to jump to turn things back down as the cones in our speakers beg for mercy. My personal opinion - establishing this dynamic range had a nice logic but only the most skilled director, producer, and editor gets it right, and it is ok to have some dynamics without the overkill. The same is, I believe, true for music. Two things in the past decade have absolutely killed people’s natural instinct to mix a good song: the loudness wars and -18 dBfs. How many people actually have any real understanding of either of these? An actual real-world range of 12-14 decibels RMS between low levels and maximum peak works very well even for classical music, believe it or not, but nothing can stand in the gap for skilled listening in an accurate monitoring environment. Listen to comparable recordings that sound good, and listen to them very quietly and very loud. Listen on the worst case scenario but take it with a grain of salt and compare at the ideal volume on the best speakers. Loudness can be our friend with it is a balanced, excellent sounding record that has its dynamic moments as is appropriate to the genre. Overboard is horrible and can permanently ruin the mix beyond repair, but truly there is a narrow range between too dynamic to have maximum impact, and overly treated to where the life is squeezed out. If good dynamic decisions are made at the mix level, and skill is used to make things sound balanced and strong, fluid and musical, then it is not very difficult to make musical decisions that also make things sound a little bigger, fuller, deeper, broader, and more believable, and yes, commercially viable. More than anything, sales are driven by consumer confidence, marketing budget, social presence, a good video production, consistent live performance, etc. But, imagine if someone is the next aspiring pop star and you put their new single in the playlist up against 3 other currently charting singles, and the stress on natural dynamics won out over the perceived presence and loudness of the song in comparison. That artist is likely to be instantly judged by most untrained listeners as being ‘less’ professional and therefore the perception is that it is ‘less’ of a recording. So, until the commercial side of the industry caves (probably too far in the other direction like cinema) to a dynamic artistry, loudness is going to be one of the elements of success. If done wrong, it is irreparable, but if handled with skill, it can enhance everything that has already been done right. Let’s speak about compression: are there any tricks you want to share? Michael Angel: Sure! I love sharing thoughts on compression and EQ. When I get serious about making a full instruction series I hope to make it possible for everyone to try their hand at things they may not have considered. I am going to save some more complex compression techniques for more advanced tutorials, but in general, there are certain secrets that work beautifully in the mix, therefore making it much easier to attain success at the mastering stage. First, break your compression thinking into categories. Compressors were created for resolving the signal to noise ratio of tape machines, so the first place to consider the compressor or limiter is in the actual tracking process. Lump this in with the list of things people want to learn, and trust me - one of the reasons that your plug-ins are not getting you the sound of “that” 80’s or 70’s group is partly due to the compressor choice going TO tape - in this case going to digital capture in a DAW. If you


are just focusing on great sound and you want to make all dynamic decisions after recording, then keep in mind the performance can still suffer thin source material if too low signal gain is chosen in most average-to-nice microphone preamps. Too much signal can sound good on its own but can get messy when tracks layer up. There may be situations where you want to track to the DAW with good compressors that give a comfortable average signal going in as opposed to an extreme dynamic range. Next, compression on drums versus other instruments. I consider compression part of the creative process - it can be transparent to let the instrument come through or it can be a very strong musical statement. Make sure you know what the purpose of the compressor is. If you have an aggressive song that needs heavy-hitting drums, I personally like to set up compression on each individual track, on the buss, and on the master. I have a system for just about every scenario, so I don’t always need to wait until a full mix is up to make certain choices. I use a relatively slow attack and release on kick and snare, and often on all of the toms. Each drum and song and performance is unique, but somewhere between 15MS and 50MS gives a wonderful crack of the snare and kick. The trick to the sound, though, is the release time. I do not treat this in its own respect, but in its actual length between the attack and the tempo of the song. So, there is a sweet spot in the right measure of “crack” and “pump” or “breathing” for a song that will come across as aggressive and musical depending on the pace and tempo of the song. If this changes I will break up, for instance verse and chorus, and treat them differently. In general, something like 20MS attack 90-150MS release makes for some nice slow drum crack, but release may need to react in a shorter time for a faster song. I use less ‘effect’ compression on the buss and go for growing the signal naturally. I like to follow the compression on the buss with a limiter, which will react very fast. I normally will use some of my natural tape compression and saturation in the chain to make each stage sound bigger and more natural. What about EQ? Michael Angel: For EQ, this is purely a decision about favorite devices for favorite scenarios. In mastering, I will give my most general tips and tricks, because any user can apply these to their own technique and come out ahead. In general, for mastering specifically, I like to start my session just listening and making notes. When I first operate on the track, I will go through an extensive signal sweeping process with a terrible sounding narrow band Eq. I do this for the purpose of exposing the worst mix decisions whether they are obvious or not. Sometimes I may already hear the issue before needing this, but commonly I will make a very narrow “Q”, turn up the gain, and start at 20 Hz. As I raise up the spectrum, I listen where the huge bumps appear. I use a master limiter after the Eq to prevent overs and to keep from getting angry. It almost never fails that somewhere I will expose some rumble, a sustain in a note that starts to sound dissonant, a harsh chord or mic option, something edgy in a vocal, and a high end that could work better for the air in the voice or cymbals on the drums. Then, I decide on whether to surgically remove these elements first, or whether I know some overlap boost/cut is going to correct multiple elements at a later stage. This will come up in a later question, but I created the world’s only corrective filter of its kind in my Mastering Suite for Nebula Pro, and it is what I use for narrow surgical corrective Eq. Once I listen with


terrible digital Eq for where the problems occur, I use my special formula filters to remove these artifacts and you can almost never hear that an Eq was used, but the problem goes away. Another general trick is the use of wide “Q”. Treating the low, mid, high as 3 sections of the song can get you a long distance in making nicely balanced, sweetened final masters. If a recording is great as it is, usually a very small change in one range or the other can make the difference from great to outstanding. I love classic, vintage inductor-based Eq’s with vacuum tube make-up gain for wide “Q” and specifically there is no prettier sound that I know of in recording than an NSEQ-2 in Vacuum Tube Mode at its regular “Q” width for mixing and for mastering. Good monitors and a pretty neutral room are essential for making conscious decisions while mastering, but what about headphones? Michael Angel: Headphones are a mixed bag. You will hear a lot of people state that it is impossible to mix or master inside headphones because it lacks the critical element of your ears making a natural blend of stereo content within a room environment. There is a bit of truth to this, and I would never recommend someone to isolate an entire album process with a single set of headphones, just like I pray that the time comes, sometime in the history of studio recording, that facilities will abandon the “NS” series of near field monitors. I think the end user can attest to the fact that an incredible listening experience can be had by the owner of an amazing vacuum tube headphone amp or an excellent solid state high-end headphone amp, and a top notch pair of headphones. This is not to be confused with an mp3 player and earbuds, which are likely to win you a case of tinnitus in your later years. I personally have, for necessity in live scenarios, and other times when a monitor had to be replaced, done some mixes that I am very pleased with, just using headphones. I would attribute this to consistency in process and the fact that I spend so much time listening critically to sound in general. If I spent weeks at a time away from musical sound and critical listening, I don’t think I could trust the total isolation of headphones as much. The other element is that I believe you can actually appreciate precise stereo placement and distance staging in headphones better than most nearfields and definitely better than most mains, but you do miss the important way that music develops by allowing the natural waveforms to fulfill their maturity. That is to say, 6-15 feet away in a flat, non-reflective room at a triangulated angle is amazing for every element of mixing. If you want absolute idealism, get a pair of amazing, well-cared for used ribbon speakers and a pair of $300 subwoofers barely turned up enough to hear the extra octave. All that you need from these is at least excellent timing down to 40 Hz or so and you can tell the rest by checking against the mains in headphones (rumble, bad low notes that may need high pass). I say this because you truly can harm your speakers AND your ears trying to go full volume listening for the bad stuff that happens in the longest waveforms. These don’t hurt your ears when you are listening, but they cause the most hearing loss of any other waveform. One other thought; I find that good studio mixing headphones like the K-240 are excellent for identifying, isolating, and removing clicks, pops, and mouth noises in vocal tracks. What are your favourite Nebula/Acqua Libraries/Plugins? What other plugins do you think are suitable for mastering?


Michael Angel: I don’t want to come across as if I am using this as an opportunity to only push my own product, but I admit I consider my own work in every product that I have developed, so I do tend to use my libraries and plug-ins for my own mixing and mastering work. I believe that AlexB’s API stereo mastering EQ is an incredible tool - truly for mixing and for mastering. I am certain that AITB, Tim, Henry, and the others have done an excellent job with EQ’s that I have not heard, so let me just state that I have had the honor to help in the earliest of stages with the first commercial releases, testing, listening, advising. This has been many years. The most important thing to me is that when something represents a specific famous piece of hardware, it truly sounds like it. This is an area where Nebula Pro and Acqua absolutely win hands down over any competitive process. Even other companies that have used a convolution process have not compared to the quality I have been able to get in my own work with similar products and I have heard the same thing in testing libraries from other developers. I created the Mastering Suite for Nebula Pro a long time ago, but it remains alone in its category, for me personally, as the world’s finest corrective Eq for any recording that will end up at 44.1 kHz, and I created a 96K filter that allows its use for higher frequencies as well. I will adapt the narrow bands and the harmonic enhancer to an Acqua in the near future. The SongTec, GEQ, and Nice EQ are 3 of the world’s best options for mastering Eq, and since they are inside Nebula Pro you can repeat their settings over and over without the analog concern for the little inaccuracies of redundancy that do occur with their analog use. I recommend one or both of the R2R and Apex for the Studer A-800 at 15 IPS for mixing and 30 IPS for mastering, the Stu A-820 for the same, and the ATR-102 at 30 IPS for mastering. I recommend Tape Booster Plus for making the signal of the recording grow before applying too much compression and limiting. It uses natural harmonics to grow the signal without changing the spectrum or transients. I recommend contacting the other developers and asking them about the Eq’s that they have worked with and have created for Nebula Pro, as many of them will have some great options for use in mastering as well! I can go on, but this is a great starting point, and for a few hundred, including the entry price of the Nebula Pro Plug-in, you can have a literal and honest representation of the real hardware in the tens of thousands, a promise that many other plug-in companies promise, but very few deliver. Speaking of other companies, I do like to recommend things that I have been impressed by, so let me make some honest recommendations that I absolutely do not benefit anything from. First, please do not just go and buy plug-ins that are supposed to represent hardware JUST because they look like the same graphics. Large companies have the advertising budget that I WISH I had. They pay a celebrity handsomely to support a release and give name recognition and it does sell thousands of downloads, but how many of them are truthful representations of those devices? Well, some of them are ok. Personal opinion: I cannot recommend any single plug-in purchase as strongly or sincerely as Nebula Pro. I hope that the future of Acustica Audio proves consistent as the past decade, as I love the deci-


sions they have made thus far and believe in Giancarlo’s work. I believe its strongest asset is for Eq’s and complex high-end signal chains like tubes and preamps, but I believe the future will make them competitive for other areas as well. I have found a great deal of success with the compressors I have chosen to focus on in my own work, and although I believe the technology has some room to mature into more complex compression options that might be easier served with algo-based options, you cannot beat the cost-per-excellence for NebPro based Eq’s. Waves has made a gazillion programs. I think some of them are very useful and others look pretty and cost too much. Personally, I think they nailed the sound of the API compressor while only captured the general usefulness of the Eq’s. They are good, but not as good as Nebula Pro Developers provide for API Eq’s. Waves made a nice iteration of the SSL Buss Compressor, but there are some better sounding compressors and the analog still sounds the best of all of them. I am not as impressed by their SSL EQ, and I once again believe there are good options here in the NebPro world that blow others away. I have long been very impressed by Voxengo’s Polysquasher. As a mastering grade compressor I have compared it to analog and software versions of a lot of devices and it is simply an excellent, well-rounded tool. It is not out to replace a specific device, but is just a great sounding tool that does not sound like algorithms to me. The same can be said for his Elephant as an option for a mastering limiter. It is simply a great sounding tool and gives the timing range necessary for a world of loudness requirements but even with very dynamic uses it is a great tool. The process that I have developed with my coding partner Michael Heiler is something that took the same years of research that went into mastering (pun intended) the complex sound of tape machines. Our process is based upon making complex analog processes come to life without randomization and without linear algo processes and also without the limitations of schematic-based design. The result is a massive amount of processing with relatively little cpu impact. Things like tape, compression, tubes, can be created that complement the natural limitations of Nebula and Acqua, and of course I use my own designs for the exact purposes I developed them, with Michael’s (the OTHER Michael) amazing skills. Har-Bal is an absolutely amazing tool. It cannot replace, in my opinion, the need for a Mastering Engineer, in the same way that a Ferrari replaces the benefit of a great driver, but - man - they have done something incredible with this tool and I love its use for diagnostics and I have run into situations where it was beneficial for fixing complex curves. I don’t mind recommending UAD to anyone looking for the best tools for pretty much any job. They have been among the top developers for a long, long time, and I believe they have applied their own research and have competed even with myself in a smart manner. Again, I wish I had their market presence and ad buys, but even if I became a millionaire I would still be a fan. Sadly, my favorite recommendation is something that is no longer valued in the marketplace, but the LiquidMix was a great tool, and for the money, it was a ridiculous value. I personally found that using a


dsp-powered dynamic convolver still could not match the quality I got with my own work with Nebula Pro, so that speaks volumes to what Giancarlo has done correctly over the years. But, I did find that if you have a decent software emulation of any of the LiquidMix compressors, you can match settings to find the areas where it is the most similar to the original hardware, while other ranges are not as useful. The LA2A and Vari-Mu inside the LiquidMix are worth the price of finding one used. This technology was created by Sintefex and license sold to Focusrite, who have a long history of investing in great individuals and products and then making stupid decisions with them - lol. They took the only technology, in my opinion, that could come near Nebula’s quality and choked the life out of it. The cheaper it got, the more they integrated it into a nice, viable option for recording chain and interface options - perhaps a good market choice for them, but sadly this meant no 64 bit compatible drivers for LiquidMix - how stupid! - lol. The Neve 33609, 1176, and LA2A in the UAD(2) are excellent. I personally find there are very few items left in my massive arsenal that I actually turn to these days outside of NebPro, but when I am working fast and have very little toe shaping to do, I find that tools like the Waves API and Voxengo’s offerings are excellent. If I had a lot of money I would support Voxengo and buy more of their plug-ins because they are unique and sound amazing, but I just do not find that I personally have a need for many things outside of NebPro. In my opinion, everyone serious about recording needs one great microphone (I didn’t say expensive, I said great), one great preamp, one transparent set of converters, good drivers, Reaper and maybe one other DAW, at least 4 Gig Ram, a PC (use Mac for anything besides recording, please!) a massive hard drive, or two, or five (I currently actively work off of 12 hard drives), one good pair of monitors, one good set of headphones, talent, NebPro, and a small investment in the right libraries. Do you use any analog hardware? Michael Angel: Yes! I love analog hardware, and although I have learned to deal with the frustration that goes along with it, I still enjoy moving knobs and tweaking smooth selections and settings. I am relieved at the impact of digitally recreating my most prized devices has allowed for - reliability, redundancy - not wearing out my more valuable analog assets, and more! My favorite piece of analog hardware currently is the Source Plus, which sadly is no longer being created. I designed it with a very talented hi-fi expert engineer with the purpose of being able to impose the precise sound of specific vacuum tubes into the buss and mastering chains. The result was a dual-tube stage design that allows me to tune in only the amount of each tube chamber that affects the sound the way I want. I can change the bias to affect the warmth of the sound, adjusting the voltage of the tube plates, and control the output gain, leaving a wealth of analog options. The purity of the signal at its nominal levels can be heard in the Vintage Tube Collection for NebPro and VST. I have used probably 100 different hardware devices, including everything that you have in NebPro now. I like the flexibility of 500 series items. I have a custom stereo pair of Neve channels that I love for the original transformers. I have the incredible and vastly under-valued Aphex Studio Dominator - if you want a crazy deal on a great analog brickwall limiter - get one!


I have API-based mic preamps, a Chameleon Neve-inspired mic pre/eq that is still ridiculously good compared to most cheapo-Neve designs. The 1084 is still my favorite analog Neve channel/console and I do not have the budget for a full Neve console or even a full sidecar, the 1084 is now my favorite inside the DAW for the Neve sound, where my outboard pair of Melbourn Neve pre’s are wonderful just for shaping and tone. I have one of the very few $700 MXL V77S classic tube mics (much, much nicer than their other tube mics, but their other solid state mics are fantastic). I have all of the expected dynamic mics, a lot of other preamp options, one of the “good” RAMSA consoles that is an absolute steal if you find one. Obviously, I am a true-blue fan of the Vox AD60VTX. I have a true vacuum tube hybrid Marshall that sounds gorgeous (none of that digital effect section Marshall stuff that was truly awful). I have a Brian May guitar and Vox amp that I absolutely love, but you are probably still asking about mastering! :-) There just isn’t much need for all of the outboard analog mastering equipment I have always used now that it is inside Nebula Pro, and even mastering compression and limiting are sounding equal inside the DAW. I enjoy analog for the recording process more than anything, and I still sometimes go in/ out of busses for analog limiting - I mostly like slow, glue-like optical analog compressors and vactrol compression. I have bought and sold various versions of my acoustic drum set which is still in flux, but I also have an electronic pads set hooked to a drum brain, connected to the master control room for MIDI, a wonderful Daking A-Range, an original Trident Series 80 channel + Eq, an original Quad Eight 444x, A Bogen all-tube rectified mixer with EQ, the best sounding Siemens W295B in the world, and of a room and of a closet full of tape machines and the tape stock that was used to create the now famous R2R and Apex collections. I will be bringing a new line of super high-end analog hardware to our community sometime soon. This will include line coloration items that every setting is unique and customized - where its analog stands out the most, and it also includes compression in a classic sense with new features that you would love to have! These will be wonderful for mixing and mastering, but they simply aren’t finished yet! How do you normally set up your mastering chain? Michael Angel: Well, I have more of a process than a chain. I listen to comparable recordings to adjust my listening and thinking towards the project. I load my terrible narrow boost Eq and limiter to identify surgical Eq choices that need to be reduced, and then it is a custom process for each song. I load different Eq’s for decisions I make one at a time. When I feel like the mixes of a full album project are coming together, I will load a chain that often includes TB+ and a mastering tape option and VTM-M2, followed by a brickwall limiter. It depends on what other kinds of comparable correction tools are needed. I fold things down to center mono to check for phase issues and I send things super wide (increase ‘side’ and reduce ‘mid’) to see if they will benefit from any adjustment in that manner. Rarely do I do any hard compression or heavy limiting, and if any of that is needed I wait until I am comparing one track to the next. I serve the client with a list of notations so they get a feel for what was done and it also lets them ask if they need to understand anything in the process.


Is it possible to master at home? What tips would you give to someone who is about to start learning this discipline? Michael Angel: There is a value to the concern from years gone by, comparing mastering to do-it-yourself brain surgery. I really think that anyone who thinks of any of this process, from recording to mixing, mastering to developing products like Eq libraries and the sort, as being an easy field to master (again, that pun!), is likely to miss out on the wonderful complexities that exist. My recommendation for anyone that is interested is not to under-estimate the complexity of the art and the science. Expect the unexpected and assume that there is always something in your blind spot that you did not anticipate. Having said all of this, yes I absolutely believe you can become very skilled mastering at home. To learn the discipline, do not try to adapt the things you read and the tutorials on YouTube without having a minimal high quality listening environment. You simply will never get it and you will think you are either way off course and no good at mastering, or even worse, you will think that you have it all together when truly you have yet to experience that “aha!” moment that is required to find a true love for excellent in sound recordings. Until you have at least sat in a high-end hi-fi listening room and heard your favorite music on the best speakers in a tuned room, you cannot grasp what it means to master a song. But, the good news is that money is hardly your enemy these days. Seriously, 20 years ago, it was impossible to consider mastering with a computer, a few software items, and a $500 pair of monitors. You can almost get by with a good result these days. You truly cannot overdo the choice of monitoring, and the best studio monitors are still not adequate in every way for the best mastering work, but to learn the skill and improve your own recordings, a nice silent room with good padding and scientific decisions in how to treat the room, online tutorials, a lot of good recordings to listen to and practice with, and a lot of patience… you can learn a lot about mastering that just was not possible in the past. The best answer of encouragement I can give you is also this; I love sound and recording and that is what drives me to continue to be interested in music and mastering. I plan to make a lot of information resources available, both for free and for a full online course. I hope that I can be a living example to help you get to that goal, and I have no doubt that the tools available inside Nebula Pro are not just “good enough” but rather are the very best, most ideal tools for the job. All that is missing is your will, determination, patience, humble devotion to improvements, and a reasonable investment in the right monitoring environment. With that said, most people are not cut out to be the perfect fit as Mastering Engineer, but I am certain that if you have the right characteristics, there is no reason you cannot succeed in the right home environment. I hope that these thoughts are a blessing and encouragement to some folks out there, and thank you for the honor of being able to answer these questions.


Ali Zendaki alizendaki What is the true role of a mastering engineer? Ali Zendaki: In my opinion, the role of the mastering engineer is to be the gatekeeper with fresh ears; to ensure quality, bring a sense of cohesion to the project, and remedy any -minor- mistakes that have been made in the recording and mixing process. What are, in your opinion, the myths/clichés surrounding mastering? Ali Zendaki: A myth I hear a lot is that “mastering is easy” and you can “do it at home with a copy of Ozone 6”. This could not be further from the truth. Mastering takes many years of experience, lots of trial and error, and an extremely trained and delicate ear. You are responsible for putting the last mark on the audio before the world hears it, you decide what is a good and bad decision, and taking these decisions lightly could easily result in a negative impact. What genre(s) do you mainly focus on? How does that affect your mastering decisions? Ali Zendaki: I focus mainly on both EDM and Pop/Urban Pop. These definitely have their own characteristics each. I research the currently charting hits in order to see the ‘mix trends’ and do my best to conform to industry standards in that specific genre.


What is your opinion about the Loudness War? Does “loudness” have an impact on actual sales or not? Ali Zendaki: It does have a very small impact on sales in my opinion, since this generation has been groomed to believe louder is indeed better. So when they hear something that maintains a good amount of dynamics, it make come off as inferior, even if in actuality it is not. Let’s speak about compression: are there any tricks you want to share? Ali Zendaki: My main tip about compression is: there are no tricks/rules. For many years, I fell in the trap by reading, researching, and looking for an easy way out. Each project, each song, is its own entity, has its own feel, and requires from-scratch compression. If I did have an actual tip, it’s to be very careful about compressing while mastering, very small amounts are usually what the doctor ordered, no more than 1-2 dB of compression, it’s also just as important to make the compression ‘breath’ with the song, so dynamics and pocket are preserved. What about EQ? Ali Zendaki: The same things goes for EQ. For mastering, use EQ for surgical tasks, making a variety of tracks sound cohesive, or to add polish/finalization. It’s not our job to add our own personal touch, what you receive already has the fingerprint and DNA that should be preserved after the mastering. Good monitors and a pretty neutral room are essential for making conscious decisions while mastering, but what about headphones? Ali Zendaki: I use headphones to master actually. A lot of people frown on this, but hear me out. Once you get to a certain level of headphones, most of the common negative remarks regarding headphones disappear. I use Sennheiser HD800s, along with a SPL Phonitor (which emulates crossfeed), and also a Violectric DA Converter, which ensures my digital bits are converted to analog signal with minimal amounts of artifacts. This combination gives me an amazing sounding ‘room’, with an astronomically high level of detail, and an astronomically small amount of distortion versus traditional speakers, as ‘rooms’ tend to alter the stereo signal. Another device I use is called a SubPac, it’s a very accurate bass generator, that goes on your chair. Rather than a subwoofer, which can be very inaccurate if you don’t shell out a large amount of money for room treatment, the SubPac uses an actual motor to emulate the bass frequencies, and let you ‘feel’ them, rather than hear, creating a more intimate and accurate environment. What are your favorite Nebula/Acqua Libraries/Plugins? What other plugins do you think are suitable for mastering? Ali Zendaki: I can’t pick a favorite, I love them all! Ok, I lied, my favorite EQ has to be the Amber, and my favorite compressor is the Titanium 3 Band! Amazing! I use a whole host of other plugins, for mastering I tend to lean towards plugins by Slate Digital, Eiosis, FabFilter, Massey, Waves, Universal Audio, Tokyo


Dawn Records, and many more, way too many to mention here! Do you use any analog hardware? Ali Zendaki: Not anymore. I used to use a summing box, but that soon imposed limitations, which I did not like. I feel 100% digital, I get complete control of the colorization and distortion I want to impart on the signal. How do you normally set up your mastering chain? Ali Zendaki: To be honest, I just go with my gut; I don’t have a chain or methodology, I just do what I feel the song needs. If I have to say anything, it’s to be surgical first, then impart processes to make the signal ‘sound better’. Is it possible to master at home? What tips would you give to someone who is about to start learning this discipline? Ali Zendaki: Absolutely, refer to my spiel on headphones above as to how this is possible. My advice to anyone mastering at home would be to: learn your room! A/B with monitors, headphones, etc. Find the weaknesses of the room so you can be sure to compensate if your room is not treated sufficiently.


Bob Davodian TranscendingMusic What is the true role of a mastering engineer? Bob Davodian: For me, mastering is not just about processing audio but also about quality control. It’s another perspective within a different environment, one assumed to be superior in acoustics and gain structuring, where a mix can be assessed and the proper steps can be taken to get the best out of it. What are, in your opinion, the myths/clichés surrounding mastering? Bob Davodian: I think that the biggest misconception surrounding mastering is that it can turn a subpar mix into gold. You know, mastering isn’t alchemy! It can certainly make things better, but there is in my opinion a certain overly fervent expectation of what mastering can do for a mix. What genre(s) do you mainly focus on? How does that affect your mastering decisions? Bob Davodian: I mostly focus on metal, rock and pop genres. I have mixed orchestral pieces as well and hope to continue to do so, as metal and orchestral music are my favorite styles of music. I can’t fully say how it affects my decisions, I can only say that for whatever is presented in front of me, I try and bring out the best out of it as much as possible, to the best of my ability. For one metal song, I may do totally different things versus another metal song. If it varies within the same genre, it certainly


holds true that between genres it could vary even more. What is your opinion about the Loudness War? Does “loudness” have an impact on actual sales or not? Bob Davodian: I used to have a blog up, but when I changed hosts a couple of years ago the blog went offline. In this blog I addressed this aspect of loudness in music. My idea spinned off of a polemic thread at Gearslutz. The thread was called “Vote on the Solutions for Loudness” and although I applaud the effort, there is no practical means by which one can thwart the loudness war while making people understand why. I’ve said it time and time again but nobody is understanding the nature of why the contention for loudness exists. Such a contention that compromises musical fidelity is by all means rooted in something else other than what it is apparent. It’s not to be taken at face value. Let me put it this way: any convention or behavior found in cultures is based on a deep psychological force or impetus. Until that impetus is addressed itself, the convention will continue to exist so long as it feeds the underlying will to empowerment, no matter the stupidity involved. If doing one thing either empowers through belief of something assumed to be true or directly creates a brute force result that invokes primitive or visceral excitement, no rule or practical solution will stop it. In other words, devise ways to suppress the loudness wars and you are giving the masses more incentive to defy it. The only thing that can even touch the onset of a solution for loudness is a profound and psychological persuasion. The sense of empowerment used in aiding the detriment of implementing unyielding loudness, which in short is a type of influence really, can also be used to mold the logic (or whatever is left of it) into realizing or at least lauding the prospects of perhaps producing not such loud masters but to show more concern for musicality. First and foremost to achieve this, the most influential and prominent individuals in both the creative avenues of our industry and the most technically proficient must set this bar. The individuals that are looked up to in which young minds would follow suite (because following those at the helm is a sense of empowerment) are the individuals that will create the model in how we must approach music production. A sense of “Man, that’s just dumb making it sound that loud and crushed” coming from the creative and technical leaders would create perhaps a scurvy of implicit shame for any person who would dare revert back to creating squashed, lifeless, harsh sounding music. Incidentally, this is the same recursion record labels/A&R/big name producers had created that started toppling a few pebbles turning it into a loudness avalanche. Why? Again, because following those at the helm gives a sense of empowerment. Rules, methods of enforcing loudness control, and the like will only create more defiance. These types of methods are associated with administration and bureaucracy: precisely the antipodal world to creativity and art. Not a good way of getting people on board. I think at this point in time, things implemented for controlling loudness, such as what we find in iTunes, it may buy some time. As I mentioned above however, I don’t think that will actually teach people why we shouldn’t aim for such loudness. To be frank, the whole debacle with Metallica and their 2008 album “Death Magnetic”, now that was a surprisingly good wake-up call for people and consumers of music. The fact that the Guitar Hero stems, which were not as crushed and pummeled, were circulating around the web enabled for a direct comparison, in turn shaming the louder, crushed version which had lost its musicality in a sense and sounded harsher overall. I think that did the most good for people to understand why it is not such a great idea to crush music for the sake of loudness.


Let’s speak about EQ/compression: are there any tricks you want to share? Bob Davodian: I can honestly say I have no tricks per say. The methods I use for the actual compression and EQ’ing of audio are things you can read or find online. My routing and signal manipulation may be somewhat unique, things I’ve worked on and built upon for years working at my craft but those things I mainly do if I mix music. Overall for mastering, I just listen. I know it’s cliché and a boring answer, but I honestly just use my ears and then tune things to how I want them to sound. In fact, that’s my trick: I practice sweeping EQ’s and setting compressors in extreme ways to understand what NOT to do in a real situation. I train just as an athlete would in the gym, except this is for my ear-brain. Good monitors and a pretty neutral room are essential for making conscious decisions while mastering, but what about headphones? Bob Davodian: Headphones, I think, are a good tool for “zooming in” on blemishes. For editing, they can be great to ensure that one hasn’t left in any bad fades, blips or glitches that may otherwise go unnoticed listening on monitors. What are your favorite Nebula/Acqua Libraries/Plugins? What other plugins do you think are suitable for mastering? Bob Davodian: I really like the AlexB consoles, I use CLC and VBC most. I also really like Tim Pethrick’s libraries, he keeps pushing the boundries. My favorite Acqua’s are probably Silk, Titanium, Green EQ, Amber EQ, and the D361A. I do think all of these are great for mastering, but mostly the Titanium and the two EQ’s are great for broad strokes and sweetening in the mastering stage. Do you use any analog hardware? Bob Davodian: I used to. Everything from multi-track cassette tapes to some reel to reel tape, not much at all and consoles. But around 2009 or so I consciously decided to go all in the box, as they say. I don’t see audio as “analog sound” vs “digital sound”. I see it as “good sound” vs “bad sound”. There are certain effects you can induce in an audio signal and you can essentially reach that effect through either means, analog or digital. Getting there and knowing how to get there may vary between the two different means, but the result is the result. That’s all that matters. Nobody says, “Wow! That song sounds so analog or digital!” They just hear and feel it sounding good or not. How do you normally set up your mastering chain? Bob Davodian: How I set it up is normally as a template. With my tools ready to go. But I don’t necessarily use everything there. They are in bypass until I feel I need to use them. As far as order of processing, I don’t have a given order except for a fundamental process: technical EQ, sweetening EQ, and peak/ISP limiting. In between those anything goes. What actually goes is what the song and situation call for. That’s what dictates it.


Is it possible to master at home? What tips would you give to someone who is about to start learning this discipline? Bob Davodian: I think you can get some results “at home”. But what does it mean to say “at home”? It’s all about how you read into those words. If you don’t have the experience, knowledge, and time behind the craft, like anything else it wouldn’t necessarily be possible to do it or at least not on a consistent basis – which is the most vital thing. Also, to go ahead and simply say “I will master my own mix”, is sort of defeating the purpose of mastering, right? Like above, I mentioned quality control and having another perspective within a different environment, one assumed to be superior in acoustics and gain structuring. Well, of course you can’t have those things if you decide to DIY the mastering process. For anybody who wants to get into the discipline of mastering, I suggest to put time into it. When I was learning, at first in the 90’s there was really no Internet like we have now. Even later on, at the start of the millennium, there wasn’t as much information and options online as there are now, so I went to school for it. Now, forget it. There are simply no excuses. Between the knowledge online, the option for schooling, and the numerous hands-on workshop opportunities by seasoned veterans out there, I’d say you would be hard-pressed not to be able to get the information you need. But any way, discipline yourself to absorb and learn all of this information, then practice and practice and practice. Listen to your favorite commercial tracks and at first maybe try to emulate what you hear. Make mistakes! Just keep doing it. That is truly the only way to get good at it.


Mark Drezzler Drezz What is the true role of a mastering engineer? Drezz: To present the music in the best possible light and make it sound good on all different playback systems. What are, in your opinion, the myths/clichés surrounding mastering? Drezz: Clichés? Myths? … That mastering is some kind of ‘dark art’ and something mysterious. It’s not, the processes are well established and there are many great tools available to achieve that aim. Of course, experience and a track record is important and you should choose a mastering engineer who is compatible with the musical genre or style in question and whose mastering you like. Listening to the work of various engineers is key to finding the right one for you, also looking out for the mastering engineers of albums you like or people who are enthusiastic about your material. What genre(s) do you mainly focus on? Drezz: Electronica, Psy-Trance, Indie, Reggae, Dub, Ska, Funk, Jazz, Folk.


What is your opinion about the Loudness War? Does “loudness” have an impact on actual sales or not? Drezz: Tricky question! It’s a question of perception and completely subjective. For some genres loudness is unavoidable, dance music for instance is very much a loudness/competitive genre. Artists will generally want it as loud as the rest of the genre, for example, if your tunes are going out on beatport or similar and will be in playlists with many artists, then the artist in question will not want their tune to be the quiet track in the playlist, so loudness is a consideration, but of course it’s a skillful engineer who will be able to make it sound loud whilst maintaining as much dynamic as possible. For other genres it can be a little different and perhaps the mastering engineer has a little more flexibility in regards to how the material is to be mastered. For real instruments and more natural sounding music like jazz or acoustic genres, there is a little more scope for retaining dynamics and being a bit more tasteful. If the music is particularly dynamic to start with, featuring intricate and quiet passages alternatively with louder crescendos, then perhaps holding off on the heavy compression or limiting may be preferable. Mastering for vinyl of course will require another set of rules to be applied, for instance retaining a good dynamic and sorting out the bass and making everything below a certain frequency Mono, so as to ensure that the master cut plays correctly and doesn’t jump or distort or display unwanted artifacts of any kind. I am not a fan of balls-to-the-wall slammed and full on maximized loudness!!! It’s much better and musical to have the dynamic in the music. It’s a little sad that we are in this position now, and I must admit I prefer a lot of older records which have the dynamic range and sound sweeter for it, but hopefully we will end the war at some point and we can just concentrate on it sounding as good and dynamic as possible. It’s a very thorny and tricky issue… Let’s speak about compression: are there any tricks you want to share? Drezz: It’s good to know intimately the different types on offer and their strengths and weaknesses, for example the difference between opto, vari mu, vca, fet and so on. Working with a limited arsenal is a good idea, so just using a few specific tools will help one become more familiar with the particular nuances of those particular units and ‘learn’ their secrets. Certain things respond to certain tasks, for instance the vari mu design has always been traditionally employed very successfully as a ‘glue’ on entire mixes, but there really are no hard and fast rules here. If it sounds good it is good, but arming oneself with good knowledge and practice, practice, practice is a good habit to get into. Compression can seem daunting to the beginner, but these devices are awesome tools for shaping the sound and it’s not just about dynamics either as certain units are just as relevant for their unique specific tone and you may want to use a unit not so much for its dynamic effect but perhaps just for its tone. What about EQ? Drezz: Again, so many EQs to choose from, but finding a handful of quality units is a good thing. Different beasts for different tasks. Clean. Coloured. Crisp. Grainy… It’s good to have a palette of choices here and again it depends on the material one is mastering, some things will want a bit of tube color and warmth to really make them sing, whereas other times a clinical and surgical approach is called for. Do not be afraid to cut! It’s not always about boosting and sometimes you may have to turn the bass down (laughs)!


Good monitors and a pretty neutral room are essential for making conscious decisions while mastering, but what about headphones? Drezz: Headphones are an indispensable tool for the mixing and mastering process in my opinion. Not everyone will agree. I know mastering engineers who don’t ever use them, but I think if you are so inclined, they are a very useful addition to your tool kit. There may be sounds revealed to you on headphones that you just don’t hear to the same degree on speakers. Again it’s good to know the market and the choices available. There are many good ones to choose from. Try different ones if possible, and find out online at forums and talks with others, certain phones will be suitable for different genres perhaps but people are usually willing to share their thoughts and experiences on these matters. Speakers are also subjective. One person’s preference would be different to another, it’s all about preference. Knowing your speakers or phones is very important. You have to rack up many hours listening to a lot of different music on your system. Listen constantly, at a sensible volume with lots of different material. Listen to your favorites. Listen to new commercial releases. Listen as much as you can to get used to them. This is key. Obviously the room will need to be treated. This is a complex subject which I cannot delve too deeply into here, but there is so much available info out there. Minimizing the anomalies and frequency peaks and troughs is very important. Bass traps, diffusers, broadband absorbers… there are many ways to tame the room. Building your own treatment can be good for some, but it is important to get the right advice before embarking on such an undertaking, but luckily there are some great people on the forums who are more than willing to give the benefits of their advice. Check out what others have done! What are your favorite Nebula/Acqua Libraries/Plugins? What other plugins do you think are suitable for mastering? Drezz: The new Acqua plugs are absolutely wonderful tools for mastering. Never before have ITB tools been so good for this task. The Green EQ is a truly splendid EQ, clean, transparent, very lush and 3d, this is the one I’d reach for most when mastering, but the Magenta and Amber are also fine tools for the job. Again, it all depends on the source material. I find the Green a good all-rounder, it fits the bill most times for the power and flexibility it offers. The Magenta, I feel, suits certain material more than others. It’s a colored, warm beast, sometimes the material will call for that and sometimes not. The Amber for me falls in between these two, it has some color, but is also useful for precise boosts or cuts. They all sound great to me. As I said before, an unprecedented quality is now available to us and I would not hesitate one moment in recommending any of them for mastering or mixing purposes. They are an authentic replacement for their hardware counterparts, and used with experience, patience and skill they can produce results that rival their hardware inspirations. The Titanium is a fantastic tool. The multi-band is so versatile, but requires some learning if one is unfamiliar with multi-band compressors, you can alter the tonal character of the track quite dramatically with this tool, it’s almost like an EQ as well as a compressor, but caution is recommended, it’s quite easy to overcook!!! Patience and caution required, but the rewards are there for the determined engineer. The new Aquamarine is just awesome. This plug has some serious mojo, and the amount of attention to detail and flexibility it offers is, for in the box, unparalleled and staggering. A future classic and nothing else quite like it out there, the tone and characteristics of the compression are simple great. Again, it’s a beast and requires some tweaking and learning, it’s sensitive, and very, very powerful, but so good (you may tell I really


like this plug and consider it a very special tool)! There are so many great libraries to choose form, where is one to start? Tim P’s offerings are meticulous and very impressive, everything he’s done has a real quality to it and attention to detail, the U76 and Silk EQ are standouts for me. The Cupwise 660 compressor is a great sounding vari mu type which just sounds lush to my ears. The Alex B consoles are also very good. The new Prime bundle is an absolutely awesome suite of plugs, very suitable for mastering purposes, and just sound great… warm, a lovely vintage character, but extremely high quality… a great sound. Do you use any analog hardware? Drezz: I am lucky in the fact that I have some nice vintage pieces of hardware in my studio. Sometimes I will run something through a vintage console to get a bit of ‘hair’ or saturation/harmonic distortion. I like to go through the transformers of the desk, which gives a very pleasant sound. Of course there are transformer libraries for Nebula too, and sometimes I will use these, but occasionally it’s nice to go through the desk, even if it’s going back in the box for additional processing, just for another layer of color or some change in the tone. I also sometimes record stuff down to tape. I have Otari and Studer machines and they sound very nice too. Of course there are the R2R and Apex collections by CDSoundMaster to emulate the tape, and very good they are too, but for me, being a very experienced tape fanatic, there’s nothing still quite like going to tape. It’s my hope that one of the future developments of Nebula/Acqua is that the tape sound is finally nailed as this would be truly revolutionary… we shall see!!! Is it possible to master at home? What tips would you give to someone who is about to start learning this discipline? Drezz: Start small, start slow, get to know the tools. You will soon start to find which tools work for which tasks. Read, talk to others, don’t feel afraid to ask potentially embarrassing questions, everybody has to start someplace, and most people who really know their stuff are willing to give advice. A mastering chain is a very personal thing and comes about after experience. Personally I usually have something like this, but it does change subtlety from time to time and job to job. Console>EQ>Comp>Tape>Limiter… Sometimes the EQ after the comp, sometimes before. I don’t impose certain rules, it’s specific to the raw material.


Gabriel Schwarz Gemini Audio What is the true role of a mastering engineer? Gabriel Schwarz: Nowadays the role of the mastering engineer is to make a finished mix sound as good as it possibly can, while optimizing it for the most commonly used playback systems. In some cases a mix gets mastered for more particular uses like, for example, the iTunes Store. Often it is expected of the mastering engineer to make the mix as loud as possible. This seems especially important for certain music genres, while in other genres it is secondary or even counterproductive. What are, in your opinion, the myths/clichés surrounding mastering? Gabriel Schwarz: The biggest myth is that the mastering engineer can make crap to gold. This is not the case. The mix has to sound as good as it can without processing the stereo buss except maybe a bit of compression. What genre(s) do you mainly focus on? How does that affect your mastering decisions? Gabriel Schwarz: I do all kinds of genres and the decisions vary greatly. Each genre has its own overall sound which makes a record of that particular genre sound authentic.


What is your opinion about the Loudness War? Does “loudness” have an impact on actual sales or not? Gabriel Schwarz: Loudness is part of many genres and they probably would not be what they are without it. But if genres which aren’t expected to be very loud are too loud, it can sound very bad. Generally music has to be produced and mixed in a certain way in order to become very loud at the end without sounding bad. I think if it sounds great, it is not too loud. But every mix has its own optimal level of loudness which should not be exceeded. I do not think loudness is a primary factor for sales. I think it is more a question of “is it loud enough, or too loud for this specific genre, does it sound authentic?” Let’s speak about compression: are there any tricks you want to share? Gabriel Schwarz: Parallel compression can be a great way to make the mix fuller and louder without killing its transients. Also M/S compression can do great things to the spacial perception. For example, it can be used to tighten up the lead vocals, kick etc. in the middle without affecting the sides. What about EQ? Gabriel Schwarz: Sometimes applying a high pass filter a little higher as you would normally do while boosting the frequency right above your high pass. Good monitors and a pretty neutral room are essential for making conscious decisions while mastering, but what about headphones? Gabriel Schwarz: If it’s very good headphones, they can sometimes be helpful for spacial decisions. Also since many people listen to music via headphones, it might be good to check the mastering through some headphones as well. Most mastering engineers however limit themselves to one to three different speakers because they’re experienced enough to know that what they do will translate well on all playback systems. What are your favourite Nebula/Acqua Libraries/Plugins? What other plugins do you think are suitable for mastering? Gabriel Schwarz: AlexB Vintage Master EQ AlexB German Mastering Console CDSoundMaster Mastering Suite Gemini Audio NVE PRE 73 Gemini Audio KultComp Tube Gemini Audio G2500 I used different tape programs from different developers.


Do you use any analog hardware? Gabriel Schwarz: Yes, I use a SPL Kultube and sometimes a Bricasti M7. How do you normally set up your mastering chain? Gabriel Schwarz: High Pass Filter Preamps / Tape stuff EQs and Compressors Brick-wall limiter Is it possible to master at home? What tips would you give to someone who is about to start learning this discipline? Gabriel Schwarz: Yes, but it is definitely harder and you need tools which can compete with the gear which mastering engineers are using. It’s very important to get to know your speakers and room, and to listen to your music on many different systems, so you know how what you are doing will sound everywhere else.


Hubertus Dahlem hubi123 What is the true role of a mastering engineer? Hubertus Dahlem: Polishing a hopefully good mix. He needs to understand where the mixing engineer wants to go sonically. If it is a great mix he needs to polish it, if it is not the best mix he needs to make it better or tell the mix engineer to rework the bits that lack in quality. What are, in your opinion, the myths/clichés surrounding mastering ? Hubertus Dahlem: Give the mastering a shit mix and he will make it a great one! What genre(s) do you mainly focus on? How does that affect your mastering decisions? Hubertus Dahlem: Since I’m a Producer/Writer I work in all genres. In terms of decisions I may reference a track from a certain genre I’m working for in order to know where I want to go. What is your opinion about the Loudness War? Does “loudness” have an impact on actual sales or not? Hubertus Dahlem: I think loudness is more a product of the record companies! If I have a new artist


and I introduce him to an A&R I definitely make sure that I’m in the same ballpark in terms of loudness like my competitors. Loudness tricks your brain and I don’t want someone to think another track is better just because it is louder. It should definitely never be „Metallica Death Magnetic“ loud. Let’s speak about compression: are there any tricks you want to share? Hubertus Dahlem: It took me a very long time to actually understand compression. For me it is one of the most complicated subjects in mixing/mastering. What helped me a lot is to adjust the makeup gain of the compressor you are working with per ear. Just bypass and engage and see if it has the same volume. Some compressor have presets where once engaged it automatically adds 2 dB of gain. An unexperienced mixer would be amazed but in reality it’s just the loudness that tricks your ears and brain. Regarding compression on the Stereo Bus, play the loudest part of the song (the hook) and I would max do 1-2 dB gain reduction. What about EQ? Hubertus Dahlem: Use Acqua EQ’s!! Since I’m part of Nebulas Beta Team my mind set in terms of EQ has changed a lot. Back in the day I may have used a preset here and there and then tweak it. With Acqua plugins you just use your ears. The plugs sound super close to the hardware, you use them like hardware, there are no presets! I love all of them but my favourite is the EQP1, Blue and Green. My friend Zino owns the Avalon 2077 and we tested it against the Acqua Amber EQ and it was pretty hard to distinguish between the real one vs the plugin. Good monitors and a pretty neutral room are essential for making conscious decisions while mastering, but what about headphones? Hubertus Dahlem: At the end of last year my friend Zino told me that he bought a pair of headphones for 1700€. I told him he’s crazy and couldn’t believe that any pair of headphones is worth that price, especially because people say you can’t mix on headphones. I met him in Berlin and heard them. It was mind blowing!! I bought myself a pair of Audeze LCD-X and didn’t look back since then. I’m not kidding, but since using them I became a 20% better mixer. I now mix 70% of the time on the Audezes and do the rest on my NS10’s. What are your favourite Nebula/Acqua Libraries/Plugins? What other plugins do you think are suitable for mastering? Hubertus Dahlem: As with EQ and compression, Nebula is unbeatable with reverb! Reverb: AITB EAR is just beautiful rooms and halls. EMT 140 from VNXT is from another world. EQ: EQP1, BLUE, GREEN Compression: Titanium, Aquamarine Distortion: Vcult I also love all the Henry Olonga’s and many more.


Do you use any analog hardware? Hubertus Dahlem: Yes, I have a lunchbox with Elyisia Xfilter and Xpressor. How do you normally set up your mastering chain? Hubertus Dahlem: It still depends on the track, but in general I have a Nebula console emulation -> tape emulation -> Aquamarine -> Hardware EQ/COMP -> Titanium -> Limiter. Is it possible to master at home? What tips would you give to someone who is about to start learning this discipline? Hubertus Dahlem: Since I’m not a Mastering Engineer, it’s hard for me to answer that. But I think for a mixing engineer it’s normal today to do a “pre-mastered” mix. What I mean by that is that having a good working combination on the Stereobuss is definitely going to benefit your mix. Don’t put too much on your master bus during the mix! It might make you believe that your mix is great where in reality, after turning off your master bus, it falls apart!


Ken Suen Himhui What is the true role of a mastering engineer? Ken Suen: In my mind, the aim of a mastering engineer is to polish and finalize the sound with an appropriate process in the conclusive stage. A mastering engineer has to maintain those songs at a similar level and tone or adjust a definitive version of an individual song. The original character of the mix should be kept as much as possible, unless there are unacceptable mistakes or producers have special requests (I know, producers often are the “mission impossible” kind of guys). Honestly, it’s better to do the mix again if there are big problems, it makes no sense to have a huge change after mastering. Every previous process affects the upcoming processes. As we know, nothing is perfect, we can only push it towards what we think is “perfect” in that moment. The mastering engineer has the responsibility to make sure that the final product is compatible with different audio systems, though it will perform differently amongst audio systems in cars, pc speakers, HIFI speakers, headphones, etc. What are, in your opinion, the myths/clichés surrounding mastering? Ken Suen: Like many mastering engineers advised, get some well mastered CD’s playing before starting your own session. It’s especially good for beginners, home-studio users or even professional mastering engineers. Of course, practicing cannot be avoided. As we all know there is no shortcut, thousands hours of previous mixing experience supports one’s mastering skill. Some problems may


be fixed while mixing and some can only be fixed while mastering. No two songs are identical, we will face new challenges every time we have a new task. We’d better regard any gears as tools. Study what is the good and bad of each gear on hand, we always have to make a comparatively good decision to finish our job. Don’t expect that magic machines in the world might heal all the bad scores within a single process. A good starting point is to write down every problem we detect, fix it step by step until we hit the target. Remember to take a break after an hour or so of listening, refreshing our ears is a smart move. Arguably, compressors and equalizers are the main tools in a mastering chain. We can consider M/S processing in case we want to improve the sense of space in a dense mix. Again, don’t over process with this plugins to avoid phase problems. Another practical plugin called monofilter4 from Nugen Audio is great to make certain frequency ranges from stereo to mono. What genre(s) do you mainly focus on? How does that affect your mastering decisions? Ken Suen: I like all kinds of music with real instruments. Generally, I listen to the song and get the first impression which is most likely what other people do. I then take notes and inspect the frequency response over the song to see whether it matches the problem that I heard. There is some good software to do cleaning job such as Izotope RX or even some free stuff. I would also like to recommend a software called HAR-BAL which helps to understand different styles of music with different spectrum phenomenon, though there is a learning curve to start. We might not finish our job using just that, however it is capable of providing us with an overall picture of different styles of music. Our knowledge assists us to determine the character the song should have. The pictures below demonstrate the spectrum of two different kinds of music, the first one is a song of Rock’n’Roll, while the second one is classical music.


We easily distinguish the different shapes between these 2 songs. We can go deeper to investigate what instruments affect this behavior. What a gorgeous idea to analyze songs like this! What is your opinion about the Loudness War? Does “loudness” have an impact on actual sales or not? Ken Suen: Personally I don’t like to push loudness to the limit. I like to preserve dynamics and space for all instruments to live happily together. Once you squash sounds to its limit, it ruins the sound which finally appears harsh and distorted. As a matter of fact, we can turn up the volume knob and thus we are still able to perceive the relaxation in a higher sound pressure environment. The Loudness War has been continuing for more than 10 years. We listened to a lot of beautiful music destroyed by this practice. I am happy to see that the Loudness War is gradually, although slowly, fading away. Occasionally though, we will still be asked to make a super loud product. Let’s speak about compression: are there any tricks you want to share? Ken Suen: Compressors and EQ’s are daily tools in our life. We can’t help touching them once we see them. Just like a cat never lets go of the fish! Compressor is the tool to control volume/dynamics. We must have the basic knowledge about the attack times, release times, threshold and ratio. Once we understand those concepts, we take all of these parameters as a starting point. Each compressor, especially hardware units, has its own response and character, we just use our ears to adjust the sonic


results. If we want to get more control in a stereo track, the “LINK” mode will be more appropriate, Dual Mono compression reacts more aggressively, and sometimes unpredictably. Obviously, either choice is just a matter of taste. What about EQ? Ken Suen: EQ may sweeten, tighten and clean the sound. One very important rule, don’t abuse it! Apply it when it is necessary. Let the sound perform naturally unless you have a special purpose. We can make it bright but not harsh, we can make it warm but not blurry or muddy. The best way is always to use our ears to decide what sound we need. If it sounds right, it is right! Good monitors and a pretty neutral room are essential for making conscious decisions while mastering, but what about headphones? Ken Suen: Headphones help us zoom on small details like clicks and pops in the track. As I mentioned, because many people listen to music via earphones or headphones, we should always reference our work on a good pair. We should carefully choose our headphones. I like the SONY MDR-V6’s very much for this purpose. What are your favorite Nebula/Acqua Libraries/Plugins? What other plugins do you think are suitable for mastering? Ken Suen: My favorite mastering plugs are Magenta EQ\Green EQ, Titanium Multiband Compressor and Nebula tape programs such as R2R from CDSoundMaster, sometimes I will add German Console (from AlexB) and Vintage Tape machine (VTM-M2) in the chain. I also like the Dynamic Spectrum Mapper (DSM from Paul Frindle). Do you use any analog hardware? Ken Suen: Yes, I still keep some hardware like M7, Elysia Empressor, TFPRO P38ex Compressor and P9 Equalizer. I use it occasionally when I feel that it is suitable. How do you normally set up your mastering chain? Ken Suen: I don’t have any special setup for the mastering chain. I like tweaking sound from the beginning so that I can feel and hear each variation during the process. Is it possible to master at home? What tips what you give to someone who is about to start learning this discipline? Ken Suen: I understand the perfect solution is hiring a room with good acoustic treatment, good far field/mid field/near field monitor and all top notch machines. As a bedroom producer, I realize that there are tight budget projects, sometimes we have to struggle in a difficult environment. Yes, we still have to invest in a pair of workable speakers and some gear. We must check frequency interactions


and response between the room and the speakers in order to get a flat-frequency-response monitoring position. There is some hardware or software products that work well, for example, the ARC2. Fully use what you have, even a pair of PC speakers may do the job. I find that it’s good to check the reverb tails and vocals with most PC speakers. If you intend to master your own mixes, I do not recommend doing it in the same day. The best approach is to let the music rest for a little while.


Franz Mikorey Zino What is the true role of a mastering engineer? Zino: In my opinion the mastering process is a place to communicate and give an objective evaluation of how a track sounds and feels. Most people sit months on their productions and sometimes lose track of sound issues or maybe have a strange listening environment. The mastering engineer on the other hand sits in a highly controlled acoustical environment and constantly works on different projects so he can give a good evaluation to where your music stands. Most mastering engineers also have had a long career in the studio being recording or mix engineers well hitting Gladwell’s 10.000 Hours before they got into mastering. Mastering is able to lift or destroy the song so the key thing is to communicate about what the artist expects to be changed or lifted and what he likes to have untouched. Then it’s obviously about the whole technical aspect of things, but since we’re in 2015 I think every engineer with good credits has it sorted. What are, in your opinion, the myths/clichés surrounding mastering? Zino: There are too many out there - most people mistake the mastering process for the mix process… or they think all mistakes can be fixed while mastering. What genre(s) do you mainly focus on? How does that affect your mastering decisions?


Zino: I don’t focus on any genre - I focus on who to work with. When you meet people who are passionate about their music and have a clear vision, it’s easy to get a connection to their sound. Somehow I never find these people in German folk music… :-) What is your opinion about the Loudness War? Does “loudness” have an impact on actual sales or not? Zino: The Loudness war is a given - it has its roots in the hearing mechanism of humans… But what I can do is teach my clients about the downsides of squashing the shit out of their baby. Sure loud seems better at first but it won’t impact sales and in a couple of years there will be loudness normalization everywhere so you have to choose wisely now for the timelessness of your music. I also think there is a natural loudness every song has which is loud enough most of the time. If you go beyond that it’s always a fight on how to get it there without losing the punch, groove and emotional impact. You also have to mangle or take away some of the low-end which diminishes size and emotional impact. But clipping or limiting is not always bad. It can give a nice glue to things or add sparkle through subtle distortion. Happily most of the time people book me for the three dimensionality of my mastering and when I tell them that the depth of field in the choruses won’t survive, they understand and we find a solution. Let’s speak about compression: are there any tricks you want to share? Zino: Yes - don’t use it for loudness. Use it to add colour, excite the soundstage and control things which fall out of the mix - creatively! And only compress when you know what you want to achieve. What about EQ? Zino: Use it in MS and be careful! There are so many possibilities when you’re in MS - you can shift things around - create depth of field and fix problems without touching things you like. Second tip is to learn your EQs and see where and how their sweet spots are. Also check if the box tone really suits the song. What really sold me to the Acustica ACQUA EQs was the realness of their box tone and how you can easily hear if the EQ is right for the material or not. Good monitors and a pretty neutral room are essential for making conscious decisions while mastering, but what about headphones? Zino: Headphones are a special beast - they can never replace good monitors but they can give you a perspective on things like a microscope. I have found planar magnetic headphones with angled membranes which I use as my second reference. They excel when it comes to limiter distortion and damping resonances in the mix. What are your favourite Nebula/Acqua Libraries/Plugins? What other plugins do you think are suitable for mastering?


Zino: I love the Green and Amber EQs - I have the hardware units of those and they are spot on! Automation is king so whenever a static setting does not work I switch to the plugin and start writing automation. Also the Titanium has a very adorable color to it and the compression action is superb. A lot of time in stem mastering I use the Titanium to even harsh or boomy frequencies and add punch before I hit the analog chain. Do you use any analog hardware? Zino: I do 95% of my work analog. Over the years I collected my dream setup through a crookwood mastering console. Before the Acqua plugins came out, I only used digital limiting and dynamic EQ. Nowadays the realistic color of the Acqua plugins is just right on some projects and since I can automate them they’re a perfect match to my analog chain. How do you normally set up your mastering chain? Zino: GML 9500 and Knif Soma are my workhorse EQs. They are in the chain most of the time complementing each other nicely. When it’s about real dynamic work (adding punch or taming overly dynamic elements or sibilance) the Maselec MLA-4 and MPL-2 are insane tools - they’re like an extension to my brain. You have a vision - lock the settings - done. When it comes to glue, the Roger Foote P3S ME has all the options without ever getting in the way of the song. Its RMS mode is king. The rest of the chain is for special colors - The Neve MBP with its silk control can do almost every color I need from a VCA compressor. On the other end, the HCL Varis Tube compressor either excites in pentode mode or smoothes in triode while always giving a nice 3D sheen. Neumann w495 and API 5500 are special colors which suit maybe 20% of the time but when they hit it’s almost unlistenable without them. :-) Is it possible to master at home? What tips would you give to someone who is about to start learning this discipline? Zino: Yes and no. If you invest in acoustics for your bedroom and fiddle with them until your speakers give you a good representation, you can achieve acceptable results which sometimes are really good. But if you need to constantly work on the highest level and put out a very good product every day, the room and speakers are no.1 to invest in. Most of the time it’s about the psycho acoustics of the room more than about the frequency response. The brain can get used to frequency unbalances quite fast but psycho acoustic problems make you judge wrong without a clue why. I just invested in a FTB room by Thomas Jouanjean of Northward Acoustics and it was the best decision I ever made. Tips: - Learn your gear - use a handful of tools but know them very good - Take breaks and wash your ears by listening to great sounding music - Master creatively without making it loud and save that for the end - Always gain match to the unmastered version - Always gainmatch your EQs and compressors - Never create preset chains - only use what you really need - Never mix into limiters or heavy compression


All content included in this document, such as texts, graphics, logos, button icons, images, sounds, videos, digital downloads, data compilations, and software, is the exclusive property of Acusticaudio S.r.l. or its content suppliers, and is protected by international copyright laws.

Other company and product names mentioned herein are trademarks of their respective companies. Mention of third-party products is for informational purposes only and constitutes neither an endorsement nor a recommendation. Acusticaudio S.r.l. assumes no responsibility with regard to the performance or use of these products.

Every effort has been made to ensure that the information in this manual is accurate. However, there are a chance that we have made mistakes. Please let us know about the mistake, and we’ll fix it in the next version of this manual.


Mastering With Acustica 1.3

Recommend Documents