Benedict Roff-MarshMastering has become this strange sort of urban legend, like alligators talking with aliens in the subway, whilst Ozzy Osborne kisses Gene Simmons under the Ferris wheel in “The Lost Boys”. It just isn’t that way.
In the 90’s I saw the rise of the idea that mastering was some sort of black magic. Suddenly with YouTube everyone is showing how it is done and much of it is crazier than an Alice Cooper show.
As is my way we will look to undo some of the common misconceptions before we look at what is practical and useful.
What is the difference between mixing and mastering?
There is a strong mindset that mastering is what is needed to make a song great. This is very dangerous as it puts all the emphasis in the wrong place. Walking backwards is a great way to trip over.
Looking at the image below, we see that mastering is a tiny bit at the end of a process of many things that need to be working together.
Most of the actions in this graphic are in the sections about creating and performing the work.
Notice how each section slips into the next section, reminding us that we can’t master a song until it is well mixed, we can’t get a mix worth mastering unless the song is well performed, we can’t get a great performance if the song is not well written in the first place.
Each set of activities sets up the next link in the chain.
To answer the question about mixing vs. mastering we need to understand what is happening at each stage of the process. We will assume that we have a well-constructed song (see article + video combos on chord progressions and arrangement):
- Mixing is balancing the separate parts of the performance to a) set the scene and b) tell the story of the song. Mixing is like having a great guitar player adding parts that make your song feel a lot deeper and more powerful. Easy to figure that you don’t need those parts as you have a solid riff, but once they are there, you wouldn’t let them go.
- Mastering is taking that lovely mix of a super song and making sure that it will fit in the box for delivery.
There is a certain amount of technicality to both stages, but I will deliberately not talk much about this as that is exactly what causes people to be off path in what they really need to be achieving at each stage of the process.
What mastering is NOT is a way to take a shoddy song, performance, or mix and magically make it blammin’. While there are stories, they are often like those alien alligators in sewers eating Ozzy Osbourne’s underpants; amusing if you have nothing better to do. Not a map to success.
Which brings me to this other graphic about effort vs. reward. If we want amazing results, we need to invest the amount of energy required to get there.
If we cut corners, there may be ways to get something worthwhile, provided we are aware of the sacrifices we are making, otherwise, we simply set ourselves up for failure when the result is garbage.
The concept that mastering is more important than anything else, if only we had good mastering, we’d have hitz, is in the “Here be Unicorns & Fools” section of the Venn diagram.
To get a bit more technical:
- Mixing is a process of many to one – taking the instruments, voices etc. and making them appear as one song (not lots of things happening at once).
- Mastering is ensuring that song works in the delivery box – this is a lot like a box of tissues where you pull one out and the next one is ready to be pulled out.
Stem mastering
Stem mastering is a relatively new concept. I will be very blunt and say that this is mixing for the lazy and NOT mastering as such. When you go buy a new car, you don’t want/expect it to arrive in a couple of cardboard boxes waiting to be assembled, like Ikea cupboards.
While I get the concept of “What if the mastering person wants to change the…“, this is just not committing to your mix and therefore your craft and music.
Anyone who comes to me asking for stem mastering I simply ask for the whole multitrack pack and charge for a mix, which includes the mastering.
What a mixed and a mastered song sound like
Bearing in mind that this is written word (watch the video please), the easy answer here is that they should sound surprisingly similar. I know!!!
We get told over and over how much difference the “right” master will make to a song. Pfft.
While there are many things online that show huge transformations, some are those dodgy before & after fakes where the after is really the before – before they made it look terrible so they could “restore” it – others are simply poor mixes that someone has made very loud, wide, and shiny. The song and mix are just as poor as they were before, only LOUDER.
From chatter, it is easy to assume that LOUD is the aim of music. If that were the case, we’d all go to airports to hear Boeings take off. ~130 dB is pain inducingly loud; deafness is happening. Best music eva!
And yet there are not crowds of people tossing their undergarments at Boeings, they are off at Justin Beiper shows. This tells us that loud is not where it is at. Story is what moves people.
Commonly an unmastered mix will sound quieter than the mastered version. The unmastered can also seem a bit less dramatic in tone and oomph. The difference though should be minimal, especially if we normalized both audio files.
Because of loudness bias, if the mix is running at -6 dB, pushing it up to 0 dB alone quadruples (double-doubles) the “impact” of that music.
Therefore, many mastering A-B comparisons rely on the impact of the louder seeming vastly better. Put the A & B at the same level and suddenly there is less to be in awe of.
With good mastering – assuming no master bus compression was done – there can/should be a groove compression stage that will really make the song paddle harder (often called glue).
This master bus compression used to be done by the mix engineer but now is often left to the mastering engineer, especially in the lower end of the market. Not because it is better that way but because when mastering, the last thing you need is bad compression that already has destroyed any hope that song has of being listenable.
Better to say “No compression” than get sent a dead duck and be accused of incompetence.
The distinctions between mixing and mastering workflow
Unless pressing to vinyl, any skilled and experienced mix engineer can master for digital release. If the mix engineer is not that skilled, I wonder why they are being used at all. (Remember the Venn Diagram above.)
As noted, mixing is about bringing many into one. Mastering is about preparing that mix to fit in the delivery box.
That used to be a vinyl record (commonly 7″ single + 12″ album) with the understanding that those records would be played at home on systems that ranged from terrible to awesome, and on radio where it would be compressed/limited again to be heard from a plastic box with a 2″ paper cone speaker or in a car from cassette.
Because sound is sound, there are no dramatic differences in mixing and mastering, just an understanding of what is needed at that point in the process.
If the mix engineer has done a good job, the mastering engineer is simply making sure that the final product is ready to fit where it will be delivered.
The tools mixing and mastering engineers use
There is a LOT of talk about different tools for mastering as opposed to mixing. If that makes you feel special, fine. But the reality is that the tools are essentially the same, especially now with most of this work being done digitally (and the better for it).
There is no formula for mastering as such, different people will use different methods. Even different methods from song to song.
A common mastering process may look like:
- Groove or glue compression
- EQ
- Limiting and/or clipping
- Output level
To “spice things up” some saturation devices may be added at key points:
- Groove or glue compression
- Saturation
- EQ
- Limiting or clipping
- Tape emulation
- Output level
Again, there are no one size fits all rules here. Just be aware of the results that are required and work accordingly.
Sum of the parts
Groove or glue compression
Compression at this point is NEVER to control overall levels as that should have been done at the mix. The groove compressor is about making sure that the song moves like one thing (not many things happening at once like strangers at a train station). Please watch the video to see this happen.
Saturation
Is a great way to add a togetherness, depth, and sparkle to things. Especially if saturation has been well used in the mix on tracks & buses. This should never be “Splaturation” (i.e., distortion) but a subtle lift. Saturation absorbs some transients.
Equalization
Commonly EQ is put after the groove compressor as compression will alter the tone a bit. That is because compression squishes the tiny trembling treble more than the big billowing bully bass. This can be ameliorated with a High Pass on the compressor’s side chain to reduce the power of the bass in the Key.
With digital being so detailed, this bit of “muting” of the extreme highs can be very welcome for warming the overall tone. I start by finding a point in the low-mids that is making things feel “muddy” or “honky”.
Once that is clearer, I then assess any other tonal adjustments the song needs. This is not what my ego wants. I don’t automatically add Bass and Treble. Commonly moves at Mastering are under 2 dB, larger moves suggests that something is wrong.
Careful use of saturation at any/all of the points mentioned minimizes the need for treble boosts that tend to make the whole track sound thin.
Limiting or clipping
Is where we manage any final transient leveling. Limiting can push harder (assuming that is actually a good thing) but clipping is kinder and cheerier. Limiting is just a very fast compressor that punches the whole mix down.
Clipping literally chops the spikes off. So long as clipping is light and/or using soft modes, it is a form of saturation that adds sparkle to the final result. If the mix seems quiet even though meters say it is loud, this needs to be addressed at the mix.
Tape emulation
Is another form of saturation, often rolling in tone and time munges. I often use this quite early in the process but show it here (almost) last as this represents where the 2″ Master Tape sat.
Go easy. This is not meant to make your mix sound like it was run over by Kit after The Hoff chugged too much purple neon juice. This is simply about helping the song to sound like one cohesive unit.
Final levels
Is where we make sure that final levels meet practical needs. I only ever have a gain plugin here that sits at -0.3 dB. This is after clipping etc. so nothing should be up there, but if it is, nothing will hit the full 0 dB bit and make the playback go Splat!
I always work to have a Crest Factor of around 12 dB, so my song is breathing nicely. Other people will do less – usually whilst quoting LUFS but I don’t hold with LUFS or overloud mixes (that usually try to cover that the music is poor).
Further reading: The definitive guide to levels in mixing and mastering
Please note that I do not mention stereo wideners. I do not think they are a wise thing and know many working mastering engineers do not use them, at least not to widen stereo.
There can be sense in narrowing low bass but if I think that is a thing, I tend to have that handled at the mix. Similarly, if I don’t like the stereo feel of my mix, I work on my mix.
Stereo widening commonly does far more damage than good. When mixing, focus on depth and stereo takes care of itself very naturally. Hard panning and lots of stereo unison/choruses make a mess that widening merely makes worse.
Mastering chains and presets
While above I show what a common mastering process may look like, in no way do I promote the idea that mastering chains or presets are of any value at all. Quite the opposite as this sort of work must be done by feel from experience based on the songs presented and the intended audience.
Simply applying a preset may seem like it is getting the right numbers, but that is totally missing the point. At least the point of music; it only serves the agenda of plugin sellers.
Analog
Inserting analog devices in the chain DOES NOT automatically make things better. Quality analog tools can sound nice, but no nicer than a nice saturation plugin does.
Most times you hear an A-B it is “unfair” as they used no saturation in the DAW and the analog toy is saturating. That merely is saying they don’t know how to mix (or they know how to sell snake oil to the gullible).
Many versions of the master
I see people talking about needing lots of versions of the master for different purposes. I do not agree with this. These days the standard is the CD which is digital and runs at 16/44. That should translate anywhere/everywhere including streaming services. The only exception being if pressing vinyl as that has more practical limits that must be met (or decided upon).
The only people who should be mastering for vinyl are specialist vinyl engineers. Unless they ask for something else, send them the CD master. If your recording is super bassy, they will either turn that down or suggest that you swap to a double album. If your recording is super wide, they will probably narrow that, so the average record player doesn’t toss its needle out of the groove.
There is some movement towards higher rates like 24/96. This makes some sense but only if your audience care. Some jazz and classical audiences will really care. Some prog audiences care a bit. Most pop audiences don’t at all. Don’t waste time on what makes no difference.
Which brings us to data compressed formats like MP3. You should not need a different master for these formats (or streaming services). If you mix sounds “wrong” on any format, it simply says that a) your mix is wrong, or b) your expectations are wrong.
AM radio through a 2″ paper cone in a plastic box sounded terrible in many ways yet the music when it came out sounded like life. Mixing is about translation. Translation is about presenting the scene and story of the song anywhere.
In 1979 The Pretenders “Brass In Pocket” sounded like great things were happening no matter how bad the radio/tv. It still does 40+ years later. If it doesn’t translate anywhere reasonable then it is not a good song or mix. Go back and fix that.
