The mastering process starts with listening to the song or collection of songs<\/strong>. As you listen, take notes<\/strong> on what you feel needs fixing. For example, maybe one song needs a bit more top end, or the vocal isn\u2019t loud enough, or track 5 of the album is much darker than all the other tracks. <\/p>\n\n\n\n Keep in mind that we master for a specific context<\/strong>, be it an album, a compilation, or a top streaming service. In an album, the loudness of one song may depend on the song before and after it, whereas the loudness of a single should be competitive with other songs in that genre.<\/p>\n\n\n\n It\u2019s very important to become familiar with and use great-sounding reference tracks<\/strong>. They\u2019ll provide you with an idea as to how finished music sounds on your system. As you work, continually compare your master to commercial releases. For this purpose, I use ADPTR Audio\u2019s MetricAB<\/a>. It lets you open up 16 songs at once, as well as loudness-match them. The plugin also provides you with a few filters to listen to just the lows, just the mids or any section of the frequency spectrum of your reference track. David Glenn of TheMixAcademy<\/a> uses this process to compare his mixes to other records. With MetricAB, it\u2019s a breeze to check the low end of the disco-funk song I\u2019m working on against Daft Punk\u2019s Get Lucky<\/em>.<\/p>\n\n\n\n First, there is something to be said about the order of the EQ and compression<\/strong> plugins. A compressor reacts to the amount of energy that hits its threshold <\/strong>at any point in the song. <\/p>\n\n\n\n If you EQ before the compressor, you\u2019re changing the level and frequency content of the audio that hits the compressor. This will result in the compressor reacting differently than if the EQ happens after the compressor. However, sometimes pushing an EQ into a compressor is great.<\/p>\n\n\n\n Engineers generally use an EQ before the compressor<\/strong> when they\u2019re trying to fix problems with the frequency balance of a song. So try an EQ, a compressor and then another EQ after the compressor<\/strong> to further improve the sound or to clean up some coloration added by the compressor. A compressor, in general, will exaggerate frequency problems that exist in a mix.<\/p>\n\n\n The Master fader on the right processes all the audio sent to the \u201cMIX\u201d bus. The leftmost Aux fader sends the MIX bus to my speakers and the middle fader sends it to my headphones. Each of those faders has the appropriate Sonarworks Reference plugin setting. Mastering is all about making small changes<\/strong> that result in great results. Think in tiny increments, say 0.5dB or even 0.25dB. Listen carefully <\/strong>to how EQ changes impact the overall sound. Many veteran mastering engineers support the idea that if you have to EQ something more than 2-3dB, there\u2019s a serious problem with the mix. Remember that when you EQ a master, you are effectively processing every track that made up the mix, so a small change will be powerful!<\/p>\n\n\n\n An oft-used equalizing technique is known as feathering<\/strong>. This method consists of doing small moves<\/strong> at frequencies neighboring the one you actually want to affect. The end result is a more transparent sound. Instead of adding 1.5dB at 150Hz<\/em>, you add 1dB at 125Hz<\/em> and 0.5dB at 175Hz<\/em>. For more specific EQ tips, see our articles about types of equalizers<\/a>, equalizer curves<\/a>, and linear phase EQ<\/a>.<\/p>\n\n\n\n This is generally the best way to fix tonal problems. The end result of reducing the energy of a frequency band is much more pleasing to the ear than the artifacts of boosting. If you bring down the low end, for example, the result is just like if you would\u2019ve boosted the mids and highs.<\/p>\n\n\n\n If you want your track to sound brighter<\/strong>, for example, a good way to do it is to use a low shelf EQ <\/strong>and bring down the lows<\/strong>. Reduce the low-frequency content until the song sounds brighter, without it having lost too much power. If needed, boost the EQ\u2019s output<\/strong> by 0.5-1dB to compensate for the loss in volume that resulted from this subtractive EQ. The opposite also applies: if your track needs a bit more low end<\/strong>, first try using a high shelf to reduce the highs<\/strong>. <\/p>\n\n\n\n Here are a few commonly occurring issues and how to fix them.<\/p>\n\n\n\n This type of processor behaves somewhere in-between an EQ and a multiband compressor. While similar to a multiband compressor, the dynamic EQ replaces the multiple compressor bands with individual auto-adjusting EQ bands. Dynamic EQ bands raise or lower their gain when the level of a chosen frequency band crosses above or below a threshold<\/strong> defined by the user. <\/p>\n\n\n\n One job that calls for a dynamic EQ is when a song is missing brightness or shine<\/strong> (8kHz and over). You could use a static EQ, but an overall boost in this frequency range can make a mix sound harsh or fatiguing<\/strong>. With a dynamic EQ, the gain of the high boost will auto-adjust to maintain a consistent tone, raising when needed and remaining flat when not needed.<\/p>\n\n\n\n Start by <\/strong>using a gentle high shelf filter<\/strong> to boost the top end<\/strong>. Then, tweak the dynamic settings<\/strong> of threshold, attack, and release. Your goal is that the dynamic EQ backs off<\/strong> a bit when brighter elements appear, such as cymbals, horns or background vocals. <\/p>\n\n\n\n Mastering engineers usually opt for transparent EQ plugins\/units. Some recommendations are:<\/p>\n\n\n\n If you need a refresher on compression, follow this link: <\/em>https:\/\/www.sonarworks.com\/soundid-reference\/blog\/learn\/how-to-hear-compression\/<\/em><\/a><\/p>\n\n\n\n In mastering, compression is optional<\/strong>. Make your final decision with your ears<\/strong>, but if you\u2019re not sure if you need compression, you probably don\u2019t. Listen carefully for tonal changes that you might not like, like reduced clarity, muddy or thin low-end, or harsh highs. Every compressor has its own characteristic tone and you need to learn a few of your compressors well.<\/p>\n\n\n\n Compressors are for bringing up low-level details and adding density. Don\u2019t use a compressor to turn down loud sections or to fix macro-dynamics issues. For example, if the verse is 10dBs quieter than the chorus, use volume automation<\/strong> and bring up the verse to close the gap. <\/p>\n\n\n\n Generally, mastering compression is applied pretty lightly. Most engineers use low ratios<\/strong> (almost always less than 2:1<\/strong>) and aim for around 1 or 2dB<\/strong> of gain reduction<\/strong>. Ideally, when you compress a song during mastering, the end result is felt, not heard. Ratios much above 2:1 will provide more audible and unnatural effects of compression\u2014which may occasionally be appropriate. <\/p>\n\n\n\n The most important controls of a mastering compressor are the attack and release<\/strong>. The transients are affected by the attack control.<\/strong> The release can be though of as the time it takes for the gain to return to normal after the signal is compressed.<\/p>\n\n\n\n A fast attack<\/strong> setting will have the compressor react to the drums, making the gain duck on every beat. A very fast<\/strong> release setting<\/strong> will have the gain return to normal quickly, but this can make the song pump, flutter, or even distort<\/strong>. Therefore, keep your settings moderate.<\/p>\n\n\n\n Start with the attack time<\/strong> around 30-40ms<\/strong>, but go as slow as 100-150ms if you have to. Set it so the transients of the percussive elements of the song are not affected when is pushing 1-2dB of gain reduction. Too fast an attack can make your snare feel weak and farther back in the track. Note: You can only find the best attack time after<\/strong> you\u2019ve set the release.<\/p>\n\n\n\n Too fast a release time<\/strong> will affect the sound in such a way that you\u2019ll get pumping artifacts<\/strong>. On the other hand, if the release time is too slow<\/strong> the compressor will never release <\/strong>the audio, resulting in constant gain reduction. When it comes to getting the best release time<\/strong>, often the compressor\u2019s auto-release setting is your friend.<\/p>\n\n\n\n There are three thoughts on setting your release<\/strong> time. A medium release time, say 100ms, will almost always be transparent<\/strong>. A slower release time will yield a fatter, glued-together feel<\/strong>. A tempo-based release, like an 8th or quarter note time value, can provide a more exaggerated rhythmic pulse<\/strong> that accentuates the groove. <\/p>\n\n\n\n Using multiple compressors<\/strong> in series (one after the other) may let you even out the level of a mix more transparently than relying on just one compressor. Go for very gentle settings and strive for transparency. <\/p>\n\n\n\n \u2018Transparency\u2019 means that the problem is fixed, but there is nothing else noticeably added or removed.<\/em><\/p>\n\n\n\n It\u2019s important to bear in mind that not all compressors are great mastering compressors. <\/p>\n\n\n\n Here are a few plugins that are excellent for mastering:<\/p>\n\n\n\n Multiband compression requires a specialized type of compressor. The benefit of using such a tool is that it separates the frequency spectrum<\/strong> into multiple bands, enabling you to compress different frequency ranges as required. For example, you can tighten up <\/strong>the low end without affecting the rest of the track, or you can tame<\/strong> that terrible-sounding ride cymbal only during choruses. <\/p>\n\n\n\n Like Uncle Ben told Peter Parker (Spiderman): \u201cWith great power comes great responsibility.\u201d<\/em> Note:<\/em><\/strong> if you want to learn more about harmonic distortion, check out Brad Pack\u2019s <\/em>article<\/em><\/a> on the topic.<\/em><\/p>\n\n\n\n Saturation is more than useful when it comes to warming up or brightening a mix. It can add sparkle to guitars and synths, thicken up your drums and give some life to dull, boring sounds. \u201cLo-fi\u201d sounds are becoming more and more popular and even mainstream pop vocals are being treated with distortion more than ever. <\/p>\n\n\n\n The harmonic content<\/strong> of a sound constitutes what we call timbre<\/strong>. Timbre is what makes a voice sound different from any other instrument. Harmonic distortion adds new harmonics that are related to the original sound. The addition of these harmonics alters the timbre of the sound, the end result of which is usually a richer, more defined tone. Where equalizers can bring up or down the level of certain frequencies in a sound, harmonic processors, like exciters, create new frequencies that were not present or noticeable in the original sound.<\/p>\n\n\n\n Let\u2019s say you\u2019re trying to warm up<\/strong> a song, but EQ just doesn\u2019t feel right. Tape saturation may add the low-frequency depth you\u2019re looking for. If the material requires brightening, tube or transistor saturation may be the ticket.<\/p>\n\n\n\n The plugin I generally use for this task is iZotope\u2019s Ozone Exciter. This processor lets you work in mid-side, but it also allows one to excite multiple frequency ranges independently, in parallel. Additionally, the gain match feature can help you decide whether the processing actually improves the material, or it just makes it louder.<\/p>\n\n\n\n Recommended plugins: <\/p>\n\n\n\n Not every song you work on needs a wider stereo image. It\u2019s important you talk to the client about their needs and wants for the project. Panning, reverbs, and delays all contribute massively to the stereo image and they can best be handled during mixing. <\/strong>The dangers of over-using stereo enhancement in mastering include diminishing the song\u2019s groove<\/strong> or weakening the bass foundation of a song. If you spread the groove elements too far apart, they\u2019ll stop sounding and feeling connected to each other and the bass will typically feel less present. Stereo widening can also cause mono incompatibility, so be sure to check your masters in mono<\/strong> if that may be an issue.<\/p>\n\n\n\n Generically, stereo enhancers are referred to as imagers<\/em><\/strong> or shufflers<\/em><\/strong>. When using a tool such as iZotope\u2019s Ozone Imager, try to mono-ize the low end<\/strong> (some engineers go as high as 60Hz). Widen more as the frequencies increase. Imagine the stereo field like a rose in bloom. Top-end can be widened much more than the mids. Since the most important elements of your track reside in the midrange, you might want to leave that 500-1000Hz range alone. In the end, it all depends on the song.<\/p>\n\n\n\n (When it comes to increasing width, using mid-side processors may also be effective\u2014see more about those below.)<\/p>\n\n\n\n Important note: If you need to adjust the stereo imaging of the song you\u2019re working on, do it as the first step in your chain, before you employ other processors. If the stereo image is okay, and all you want to do is widen it a bit, insert your stereo width plugin second-to-last, right before the limiter\/maximizer. Recommended plugins:<\/p>\n\n\n\n First things first: what is mid-side (M-S) processing? This type of processing is an effective way of adjusting either the center image or the sides of a mix. Remember that mono means the same signal in both speakers (appears in the middle), and the stereo means different elements in the left and right speakers. Mid and side are also sometimes referred to as sum and difference<\/strong> signals.<\/p>\n\n\n\n Mid signals typically include the kick, snare, bass, lead vocal, and other lead instruments. Side signals include reverb and effects, drum overheads, background vocals and any instruments that are panned to the left and\/or right.<\/p>\n\n\n\n M-S processing is not a silver bullet. It does have its drawbacks. For example, if you use an M-S EQ on the lead vocal (mid), you may also affect a portion of the guitars and other instruments panned only partially to the sides. M-S processors must decode the stereo field to mid and side components and later rejoin the two back into a stereo signal. All this processing may cause unwanted artifacts, so listen carefully!<\/p>\n\n\n\n If your mix sounds muddy<\/strong>, consider using a low shelf cut<\/strong> on the side channel to clear it up. You could use this to clean up stereo pad sounds without affecting the vocal, kick and bass in the center of the mix.<\/p>\n\n\n\n Sometimes, applying standard stereo compression to your track folds in the stereo image a bit or pumps from the low-frequency content. The solution for this may be to use an M-S compressor. Apply the necessary amount of compression to the center and to the side, independently of each other. <\/p>\n\n\n\n One common use for M-S compression is to correct <\/strong>a mix where the lead vocal is too loud<\/strong>. Gently compressing the mid (perhaps with a high-pass sidechain EQ) can push the vocal slightly back into the mix while retaining a wide stereo image. <\/p>\n\n\n\n Another example would be to control the 808<\/strong> in a trap record without squashing the whole mix. Again a mid compressor, maybe even only the low band of M-S multiband compressor would do the trick. Izotope\u2019s Ozone Dynamics process provides multiband and mid-side processing at the same time.<\/p>\n\n\n\n This might be a new concept to you, but it works in certain cases. Say you\u2019re working on a rock song. The guitars are double-tracked and hard-panned, and they sound great. The cymbals are nice and pleasantly bright. The lead vocal in the center, however, is a bit dull. <\/p>\n\n\n\n If using a mid-side EQ fails you, instantiate a mid-side saturator<\/strong> such as the Ozone Exciter. Switch it to mid-side mode and add that extra character to your vocal by saturating it. <\/p>\n\n\n\n In case the song you\u2019re working on requires it, go ahead and use some creative volume or panning automation.<\/p>\n\n\n\nPlugin Order<\/h2>\n\n\n\n
![\"\"](\"https:\/\/blog-uploads.imgix.net\/2020\/03\/Plugin-In-Order1-406x750.png?auto=compress%2Cformat\")
Here is a list of my plugins and what they are doing:
1. Brainworx bx_digital : EQ and stereo widener
2. Sonnox Oxford Limiter: Only for saturation
3. Plugin-Alliance Elysia Comp: slight compression and soft clipping
4. UAD Manley Vari Mu: Glue compression
5. Kush Clariphonic: EQ for silky top-end
6. Fabfilter Pro-L2: Maximizer and dither
7. Izotope Insight: Spectrum analyzer
8. Waves WLM Plus: LUFS meter
I also automated the Master Fader, which, in Pro Tools, changes the level before<\/em> the plugins.<\/p>\n\n\n\nEQ<\/h2>\n\n\n\n
Subtractive EQ <\/strong><\/h3>\n\n\n\n
\n
DYNAMIC EQ <\/h3>\n\n\n\n
\n
<\/li>\n<\/ul>\n\n\n\nCOMPRESSION<\/h2>\n\n\n\n
\n
<\/em><\/a><\/li>\n<\/ul>\n\n\n\nMULTIBAND COMPRESSION<\/h3>\n\n\n\n
Make sure you don\u2019t overuse<\/strong> multiband compression as it can negatively impact a song in a number of ways. Such a processor has to pass your track through multiple filters before it\u2019s split into the different frequency bands, often adding distortions to your track. Using multiband compression on acoustic instruments can sometimes make them sound artificial or unnatural.<\/p>\n\n\n\n\n
HARMONIC DISTORTION<\/h2>\n\n\n\n
\n
STEREO IMAGING<\/h2>\n\n\n\n
\n
<\/p>\n<\/blockquote>\n\n\n\n\n
MID-SIDE PROCESSING<\/h2>\n\n\n\n
M-S Equalization Tips<\/h3>\n\n\n\n
\n
\n
\n
M-S Compression Tips<\/h3>\n\n\n\n
M-S Saturation <\/h3>\n\n\n\n
AUTOMATION<\/h2>\n\n\n\n
\n
\n
\n