Get a free year on Tuts+ this month when you purchase a Siteground hosting plan from $3.95/mo
Quick, what is the first thing that comes to mind when you think of mastering? Was it make the tracks sound better? Louder? Punchier? Or something along those lines? If so then congratulations you know a part of the mastering process!
Also available in this series:
- Mastering the Mastering Process Part 1: Overview
- Mastering the Mastering Process Part 2
However, there is a lot more to it than making things sound bigger; there may even be circumstances when you wouldn't want to make a track sound bigger! In this tutorial we are going to examine what exactly the mastering process is and how to become more effective at those areas.
Some parts of this tutorial series can be done in any DAW while some sections can only be done by a select few DAWs; this is why some DAWs are referred to as Mastering DAWs.
This first tutorial will cover the Mastering Process up until the Loudness War and getting ready for pressing. If your Ps and Qs are not in line or your track timing isn't landing on a down beat then step inside and get ready to master the Mastering process. Oh and yah we will definitely play with compressors and EQ too!
Ready? Let's get Mastering!
What is Mastering?
Mastering is and has always been somewhat of a black art amongst musicians and even some audio engineers; and usually audio is a black art to the rest of the world so that says something about Mastering in particular! In order to achieve a better understanding of how we master today, we must first understand where mastering came from and how it has changed. Time to put on your thinking caps because history is about to hit you upside the head!
Originally mastering as a separate process did not appear until we had the advent of magnetic tape. Before this time audio engineers would set up a mix and this mix would be played directly to a disk that would be cut with grooves using a mastering lathe; the recording, mixing, and mastering processes were essentially all one giant process. Once engineers began to use tape however they were able to separate these processes into distinct parts. While the recording and mixing processes involved the sound the Mastering process was left to control how the disks were cut. If you cut too far into the disk then you would end up with a hole in the disk or you would interfere with the opposite side of the disk. In addition you had to also keep in mind the running time of the disk and how much material you could fit on one side of a disk.
Since these master discs would be used to print all other disks for consumer consumption they had to be done right. Mastering engineers would use compression, EQ, etc. to control the audio content a little bit better so they could more easily cut the discs. Overtime engineers and producers realized that this post processing on the final mix could actually enhance the overall sound of the mix. Remove the disc cutting and replace it with CDs and all of a sudden you have Mastering acting as it does today.
Today Mastering Engineers have a much easier time than they once did. Many do not need to cut disks and truth be told many modern Mastering Engineers have not; only specialized analog Mastering will contend with this aspect in today's world. The Mastering Engineer is responsible for sequencing the track order, applying post-production to the audio material, control the overall volume, add appropriate meta-data, and finally create the master that the CD pressing facility will use to generate the CDs. In the following sections we will look at these different jobs the Mastering Engineer must perform and how to effectively perform these jobs.
Often times sequencing is one of the more forgotten aspects of the mastering process in today's world. Sequencing for a Mastering Engineer is not adding MIDI data (that is MIDI sequencing!) but the act of organizing the tracks into their final order for the CD. But that is not the whole story now is it?
If I handed you a collection of random songs and told you the order they should go in then you could probably do it no problem. But would your timing make sense? The artistic aspect to sequencing is in the time in between when one track ends and the other begins. CDs are supposed to sound like a cohesive performance that was played straight through from track one to the final track. However bands do not always blitz from one song straight into another, there is a natural sense of space. When watching an orchestra perform it is the conductor who controls that empty space and as a Mastering Engineer you must be the conductor. Here are some guidelines to follow in order to achieve a better flow in your sequences...
- Count like a conductor. When the song ends keep counting and feel for that natural release of when the song is truly over; the song isn't over until the conductor puts down his hands.
- Slower songs will almost always have a longer lead out time than faster songs for that natural release. It is on slower songs you really need to be deliberate on when you will actually end the track because a sense of timing is more noticeable in slower songs.
- Be careful of transition times from a fast song into a slower song. You do not want to add excessive amounts of lead out time on a fast song to compensate for a slow song that will be played next. What happens if they rip the CD into MP3s or play it on shuffle? No one wants to hear a fast song, a long pause, and then a fast song. Instead add more lead in time to the slower song.
Another issue to keep in mind with sequencing is how long or short the fade ins and fade outs will be. No matter what the case you will need a fade in or fade out of at least a few samples (that's a fraction of a second) to prevent any audible clicks or pops from occurring. Remember your lead in and lead out times are not necessarily the same as your fade in and fade out times. Also consider whether you want the song to end on definitively or to just fade away (more of a 70s thing but hey you never know when it may sound cool!)
Here are some short examples of sequencing between a slow song and a fast song. The first example will be fast to slow without lead times, second is with lead times, third is a slow song to fast song without lead times, and finally the fourth is the third but with lead times.
Ah post processing, what most people take mastering to be. This is where as an audio engineer you get have the most fun and be artistic with your work. However before we dive into the techniques we should cover what exactly goes into post processing for mastering.
When you process for mastering your job is to enhance the mix and make the album sound like a cohesive performance. Now unfortunately you will probably have to fix things to make it that way as opposed to just enhancing but such is the life of the Mastering Engineer. You cannot focus on just one song at a time but keep in mind the sound of the entire album and make all the songs fit that mold. Why would a solo guitar and voice on one song be just as loud as the entire band playing on another? Now obviously for the listener's sake you will need to make that voice and guitar a little louder but it in no way should be quiet as strong as the full band. The more the material sounds similar the easier your time as a Mastering Engineer will be.
Post Processing - Restoration
When we do processing for a Master we first need to make sure the tracks are up to snuff. There should not be any hissing, clicks, pops, etc. because if they are there all we are going to do is accentuate them when we try to enhance the rest of the track. If they are there then you need to get rid of them as best you can. While we all wish we could work with audio engineers who know how to mix and edit properly we do not live in a perfect world; you will probably end up Mastering a CD mixed in some guys basement at some point! Here are some guidelines for removing this unwanted junk...
- For hiss your best bet is to get a denoiser or dehisser that uses a sample to remove your particular hiss. By using a noise sample from your track you are ensuring that it will adequately remove the noise and leave behind your audio relatively unscathed. Most all denoisers and dehissers will work on a sample basis.
- When dealing with clipping distortion (aka too loud!) you will have different options depending on your situation. If you have very minor spikes that you want to smooth out then in many DAWs you can actually redraw the waveform to smooth it out; do not apply this technique liberally as it can become quite obvious. For more extreme cases of distortion you will need a declipper of some form or another. The declipper essentially will reduce the volume of the whole track (or selected region) and attempt to correct (educated guess) the waveform. I personally have had a lot of luck with poorly recorded vocals and removing the distortion with a good declipper.
- With clicks and pops you can usually attempt to redraw them if they are only a few samples long. Remember clicks and pops occur when a waveform drastically changes directions (square wave for a few samples).
While not directly related to anyone restoration issue another very helpful tool can be spectral editing. A spectral editor will show your audio on a graph with time as the Y-axis and frequency as the X-axis with color intensisty showing volume at a particular frequency. The benefit you have with these editors is you can usually see abnormalities and they do a good job at removing them without leaving artifacts when used in small doses. They are not so hot at removing hiss and distortion but if there is a persistent click that keeps popping up on quiet sections then this is the tool for you.
Finally if worst comes to worst you will have to ask the mixing engineer to fix their problems. As the Mastering Engineer you are responsible in the clients eyes for making the final product shine and if the mixing engineer screwed up that bad they will have to fix it. Its never a joyous occasion to tell a fellow audio engineer that they screwed up that bad but sometimes you must.
Post Processing - Balancing and Track Compression
Now that we have made sure our tracks are clean and mixed well we can finally begin to add that extra shine to the album. Things you are going to want to keep in mind are...
- What do these tracks need to make them sound cohesive?
- Is the overall mixing balanced or is weighted to one side? (Bass heavy, harsh high end, etc.)
- What is the sound the album is going for?
- Your job is to enhance not change the album; if you have to use judicious amounts of EQ and compression then something is wrong.
A lot of these questions can be answered simply by asking the producer or artist. However if you do not have access to either of these then you will have to take your best shot at it. Ideally you will have a song or two that are very definitive in their sound and is an obvious choice for the overall sound of the album. Another option you have is to load in a recent well produced album that sounds similar to the one you are working on to have as a reference. Once you figure out which track this is (whether by choice or by asking the client) you can begin to work on the album.
- Start by try to match the overall sound and tone of the tracks to your reference track with subtle adjustments (unless you are working with a Object Oriented DAW you will have to have separate tracks for each song in order to do this). If it is a tone issue then use EQ, if it is a fullness issue (NOT A VOLUME ISSUE) then some compression would work. Remember to use these tools subtly. If you have more surgical sounding EQs and compressors now would be the time to use them (anything that is Linear Phase will be helpful). Remember that this stage is for matching sound, not adding more musicality.
- If you find that after controlling the tone across the album that some tracks that sound like they should be at the same volume (two acoustic tracks for example) but are not at the same volume then use the faders to adjust the gain; remember you can always turn the loud song down to match the others. I would not use compression for this as you probably already have some compression that you applied in addition to the mixing engineers compression and there will be some final master compression; let us not compress it anymore shall we? You have gain and faders for a reason!
- Next I would highly recommend checking the volume of all the songs and not just the ones that sound similar. As I pointed out earlier an acoustic guitar and singer isn't as loud as a full band and shouldn?t be quiet that loud. However there is a 'competitive level' volume wise that should be met. Use your ears to be the judge of what still sounds right on its own (as a single) and in context of the CD.
- With everything matched and even sounding you can begin to enhance or 'sweeten' the tracks as some engineers refer to it as. If you are going to use any compression for sweetening then do it now as you should only be using subtle amounts of EQ and compression and putting the compression after subtle EQ will negate the EQ. I prefer to use multiband compressors for mastering because since you are applying the compression to the whole track your kick and bass might end up compressing your whole track if you are not careful! I compress very slightly (no more than 2:1 ratio) and adjust the attacks and release relative what band I am working in. Try using a quicker attack and release (between 10-20ms) for the low end to keep it nice and tight. For the mid range I would use a 20-30ms attack and release to keep it controlled but with room to breathe. Finally for the high end a slower attack and release will help maintain a more natural top end while applying subtle compression. These are only tough estimates, as different pieces of music will require different ratios, attack and release times, thresholds, etc.
- Finally we can move onto applying some final EQ to the tracks. Emphasize the ranges that you believe need to come out more in the mix. I personally like to emphasize the 2-5 kHz range just a tad as a lot of definition for a lot of different instruments can be found in that range. Another range to consider is the low end, does it have too much boom, not enough? Remember subtly enhance with your EQ with no more than a few dbs and use fairly wide Q widths as well.
Here are the two tracks before post processing and after post processing...
Conclusion Thus Far
So as of now where do we stand? We have sequenced our tracks, ensured their cleanliness, and gave them that extra shine for final production. However we still need tackle the dreaded issue of loudness, the mystery of meta data, and finally how to get it to a CD plant. In addition, we still need to talk on what is appropriate Mastering equipment is and cover some other tricks to make an album really cohesive (especially for Jazz and Classical recordings).
I hope you have learned something thus far and that you will stick around for round two of Mastering the Mastering Process. In the mean time really hone in making those tracks gel together and get that professional sound. Thanks for reading!