Music 265B Week 13 Mastering
PDF version available HERE
Mastering is the final phase in the music production process. During this phase, we turn our finished mix into a Master audio file, from which all mass-produced copies of CD’s, MP3’s, vinyl pressings, and digital streaming files will be made. In other words, this is the copy that we will sell, distribute, or otherwise present to the general public. Because of this, our master copy needs to meet a few “quality control” standards in order for the public to consume it. This step may be as simple bouncing down our mix into a stereo audio file, but the process is often more complicated than that.
People will listen to our song on a variety of devices, like a pair of cheap ear-buds plugged into a cell phone, their car stereo, or inside a club on a massive sound system. This means our one file needs to sound fine across all of these devices. When you mixed, you may have done all of the work through a pair of headphones, or in a home studio on a pair of decent studio monitors. The mix may have sounded fine through those speakers, but if you play it back on your car’s stereo, or have your DJ friend play your track through the nightclub’s PA system, things may suddenly sound out of balance: the extreme low end of the kick drum & bass guitar are overpowering the mix, the splashy high-end of the vocals & cymbals are piercing, and so on. When we listen to a track through a pair of tiny earbuds, headphones, laptop speakers, or through a small pair of studio monitors, these devices usually aren’t capable of accurately representing the full sound spectrum: they usually aren’t large enough to produce the extreme low end frequencies. Beyond that, everything from the configuration of the speakers to the size, shape, and surfaces of the room you’re in will add their own unique “color” to the things you’re hearing: some frequencies will be boosted, others will be cut.
To compensate for all of this, we need to use a trick called Reference Monitoring: we need listen to our mix through all of the devices that our potential fans would use to hear our music, in different environments as well. In other words, we need to actively listen to our mix through headphones, a boom box, a hi-fi system with a subwoofer, and so on. Ultimately, we want a mix that will sound fine on all of these devices. For example, we want our low-pitched instruments to be loud enough to sound clear through a pair of headphones, but not so loud that they blow up a subwoofer. Our highs need to be loud enough to sound clear, without being too piercing across all of these devices, and so on.
Most audio interfaces allow us to listen to our mix through multiple devices (or A/B our mix). Our “main” monitors may be a pair of studio monitors, and our “alternate” monitors may be a set of computer speakers. We probably have a headphone port as well. Some even let us listen to our mix in a mono format. In this case, we can do most of our mastering work right away. If we need to listen on our car stereo, or on another device, we need to Bounce our song down to a stereo audio file. We will cover Bouncing near the end of the lesson.
Less is More
We have already covered a lot of the same tools that we will use during the mastering phase: equalization, compression/limiting, automation, fades, and so on. When we experimented with EQ & compression, we usually made extreme changes: drastic cuts ands boosts in narrow frequency ranges, heavy compression & limiting on certain instruments, exaggerated fades, and so on. When we master a track, we usually make broad, light changes to the overall mix. The idea is that our mix is “close enough” to what we want the public to hear, so it should just need some light polishing, if anything. If you find that the song needs some extreme changes during the mastering phase, then you might just have to go back and remix the tracks. Then again, if it’s not broken, don’t fix it.
More often than not, we need to make our mix better prepared for the mastering phase. If it is necessary, we might have to go back and alter our EQ, compression, panning, and other plug-ins to fix our trouble spots in the overall mix. For example, in recent years vinyl records have been making a comeback. If we’re planning on cutting our music to vinyl, we need to make a few adjustments to account for the physical limitations of that technology. In fact, we may need to make a specialized mix just for the vinyl format. Since vinyl records make their sound when the needle moves along grooves on the record, sounds that make the needle move faster will compromise the sound quality: the record my distort, and the needle might skip. This includes louder volume, high frequencies (like those crisp sibilants in the vocal track), a wide stereo image (and phasing effects), deep bass, a wide dynamic range, and so on. On top of all this, as the record plays, the grooves get smaller, and the sound quality gets worse. If you mix for vinyl, keep those limitations in mind, and tailor your tracks to account for that. When we mix for a digital format, we don’t have these limitations. All things equal, in the digital world what we hear is what the fans will hear.
The Mix Bus
Since mastering involves making changes to the overall mix, we can work in one of two ways. We can either Bounce our song down to a stereo audio file & master it by itself in a new Pro Tools session, or we can do our mastering work on our main mix track in Pro Tools. If your system has trouble managing all of the plug-in processing, it will be easier to bounce the track down and master it in a new session (See the section on Bouncing to Disk, just be sure to bounce the track into the same bit depth & sample rate as our session).
If we don’t already have a mix channel, then select Track > New and create 1 – Stereo – Aux Input track. This mix track should receive its Input from an empty stereo bus (like Bus 1 & 2, or whatever is available). This track should Output to the system’s main output (like the Master Fader, on Out 1 & 2, or something similar). Next, route the outputs of all of our other tracks into that mix channel’s stereo bus. If we routed all of our instrument groups, like our drums, to their own aux track, then just reassign the output of those aux tracks. The signal flow should look like:
Instruments (audio tracks) output to instrument groups (aux tracks) input.
Instrument groups (and reverb/other effects) output to the Mix track (aux track) input.
Mix track & Master Fader output to the main system output.
With this technique, we will use our mastering plug-ins on the Mix track, and keep our Master Fader set to zero. That way, the volume level on the Master Fader is the true level of our mix. In the end, we never want our Master fader to peak into the red: this will create digital distortion.
By now, our mix (or remix) should be relatively complete. As we listen to our mix on different speaker systems, we might need to make a few small adjustments to the overall song’s balance. We might find that the overall low end is still too loud or soft when we listen with a subwoofer, or the midrange is too muddy, tinny, the highs are too piercing, or too dull, and so on. Instead of making the kinds of extreme adjustments we may have used to warp the tone of our instruments, use a multiband EQ plug-in (like EQ3 7-Band) on our mix track to gently boost or cut the troublesome frequency ranges as it is needed.
We can use compression, limiting, and other techniques to alter the overall dynamic level of the song. Compressors and limiters function just like they did during the mixing phase, but we have a few more tools we may use. A Multiband Compressor/Limiter can affect several different frequency ranges. For example, we may want to control some of the pumping low frequencies, without affecting the mids or highs. With a multiband compressor, we can select just the frequency range we want, while ignoring the others. Otherwise, we can use different compression settings on other frequency ranges to get a different effect.
Generally, we may need to control some of the loudest peaks in the songs. We still want to preserve the band’s overall dynamic contrast, but all of the instruments combined may “pump” on the downbeat a little too much. We can Normalize or Maximize a track. Normalization aims to set the track’s peaks to a uniform volume: all parts of the song from the softest to the loudest will be raised or lowered to match the same general level. A little bit of normalization can help a mix sound consistent, but too much will be boring: the loud parts sound more exciting when there is a soft part to create contrast. Maximizing tries to make the overall track as loud as possible, without distorting. A maximizer works just like a compressor/limiter, with an added feature. The Maxim plug-in has a Ceiling setting, designed to prevent our levels from going past a certain level. For example, we can set the Ceiling to -0.1 dB (just below the limit of distortion at 0 dB) to smash the volume down below that ceiling. This can be handy when we need to prevent our tracks from peaking into distortion, but too much maximization can cause ear fatigue. Again, we want to preserve a sense of dynamic contrast.
Some engineers use some kind of Enhancer to add some high frequency “sparkle” to the mix. Enhnacers work in different ways. Some, like an Exciter will add some phase-shifting to the higher frequencies. Others will delay certain frequency ranges and blend them back into the original signal. Plug-ins like these can add some interesting effects to your mix, but they will create some phasing issues if you don’t know how to use them properly.
We can also use EQ and Compression in extreme or unconventional ways. We can filter our entire mix through a very narrow EQ band to make the mix sound like it is playing through a telephone or old record player. To do this, use a high pass & low pass filter to remove all frequencies except for the midrange. A Filter Sweep uses a similar effect, but the EQ range shifts around: this sounds like a guitarist’s wah pedal. Again, we filter the mix down to a narrow bandwidth, but we automate the frequency knob to “sweep” from low to high, and back down again. When dealing with repetitive music, we can tastefully apply these special effects to a small section to create some contrast.
We can still use a bit of Automation if the track calls for it. If we plan on creating a long studio fadeout, it is best to do this in the mastering phase. Otherwise we might want to use some fader automation to control some of the loudest peaks in the song, as needed.
Dithering can be a confusing technical process for most beginners. Just know that even if you don’t fully understand how it works, you still need to use dithering when you’re converting a file from a higher resolution to a lower one. When we first started this project, we probably created our session using a high bit rate, like 24 Bit or 32 Bit Float. This higher bit rate gave us a more detailed dynamic range, but our final product will most likely be at a standard CD quality: 16 Bit. Remember, Bit depth refers to the potential loudness of any given sample in our song. 32 Bit & 24 Bit aren’t necessarily louder than 16 Bit, they just have a higher resolution: more potential levels between silence and the same maximum volume. Think of it like money. Let’s say you have $10. You could have a single $10 bill (1 Bit), 10 $1 bills (10 Bit), or $10 worth of change. They all add up to the same total value, but there are more potential places to “round” to the nearest amount as your resolution goes higher.
When audio gets converted from a higher resolution down to a lower one, the loudness of any given sample gets “rounded” to the nearest bit in the dynamic range. This might cause errors, or distortion in the track. To compensate for this, we use a Dither plug-in to reduce this noise. A dither plug-in uses Noise Shaping, kind of like noise cancelling, to prevent these “rounding” errors in the track. Even though we have a fancy Pro Tools rig at home, capable of recording and playing back extremely high resolution files, our fans are still going to listen to our music on extremely dated technology. The average CD player & other digital media devices, process audio at 44.1 kHz, 16 Bit. We should always use the Dither plug-in as the final processor in our signal chain. Place the Dither plug-in on the Master Fader.
Bounce to Disk
Bouncing to Disk lets us convert our multi-track session into a single track that other devices can understand. Before we bounce our tracks down to a single stereo file, we need to make sure our Master fader is not overloading at any given point in the song. If the meter ever peaks into the red, the track will overload and cause some unwanted distortion. To fix this, either lower the overall volume, or use compression/limiting, EQ, automation, and other tricks to make sure we never overload the Master fader.
When we’re ready to bounce, select and highlight a section in the session’s timeline: this represents the section that will be rendered into a stereo file, from start to finish. Typically, this means selecting the first & last clips in the session. If we have reverb and other effects that “ring out” after the last clip has played back, then be sure to highlight this “ringing” space as well. Next, select File > Bounce To > Disk to bring up the Bounce menu. In the Bounce menu, make sure that the Bounce Source matches the Output on our Master fader: this should be the system’s main stereo output (e.g. Out 1-2, Analog 1-2, and so on: every system may be different).
Let’s make a bounced file in Standard CD Quality:
File Type: Wav or Aiff
Bit Depth: 16 Bit
Sample Rate: 44.1 kHz
SPECIAL NOTE: If you plan on mastering your track as a stereo audio file, then bounce your track using the same sample rate, bit depth, and file type as your current session (e.g. Aiff, 48kHz, 32-Bit Float). You can go to Setup > Session to see this information.
File Name: Give the track a unique name, like the title of the song and today’s date. You will most likely bounce down multiple versions or mixes of these songs, so choose a name that will make this easy to find.
Directory: This is the location where the file will be saved. Press the Choose button to change this location. It is a good idea to save these bounced files in a separate folder inside our main session folder. In other words, press the Choose button, find the Pro Tools session file, create a New Folder called “Bounced Files” if one doesn’t already exist, and choose this Bounced Files folder as the directory destination. Don’t save into the Audio Files folder. This way, the session data, and our bounced tracks are all in the same place on our hard drive.
We can create an MP3 version of this bounced file if we check the Add MP3 box. MP3 files have significantly lower audio quality than standard CD quality Wav & Aiff files, but they are also much smaller files.
Before we press the Bounce button, we have the option to use Offline bounce mode. During a normal bounce, the session will play the track and count down in real-time. If the song is 5 minutes long, Pro Tools will play back the track from start to finish as it counts down to the end of the song, and renders it into a finished audio file. This is our last chance to catch any mistakes, stop the bounce process, fix our mistakes, and start over. Otherwise, we can use Offline bounce to quickly bounce the track down to a finished stereo file. In Offline mode, we won’t hear the track playback, but it will bounce and render much faster than real-time bouncing.
Now that our Master file is finished, we should play it back to make sure everything is ok. Load the file into a media player like iTunes, and press play. If everything sounds good, burn a CD, transfer it to your phone, or upload it to a streaming service: you’re done.