Common Misconceptions About the Mastering Process
A very big part of my job as a mastering engineer is educating clients before, after, and during their project.
Prior to mastering, this involves things such as explaining what kind of files I need from them in order to give them the best results, what can and can’t be achieved in mastering, and what formats they’ll need for production and distribution masters based on their release plans. Gone are the days where one digital master works well for all digital releases although things could be coming full circle, especially if CDs officially die.
Back in the 90’s and early 2000’s, when a project was done it was common to only receive a master CD-R and a few reference masters from your mastering engineer and that was that. Common end formats now include a DDP image (for CD replication), 24-bit/Native Sample Rate WAV files for some online distribution services, 16-bit/44.1k WAV for others, mp3 files tagged with lots of metadata for personal/promo use and some digital download card services.
Those are just the main formats, all of which can benefit from some fine tuning to some degree. There are also sub-formats like vinyl pre-master, cassette pre-master, Mastered For iTunes, etc.
During the mastering process, I inform clients about what can and can’t happen in the mastering process. There’s no doubt that loudness is the #1 concern from clients.
My #1 goal as a mastering engineer and business owner is to make clients happy, fulfill all their requests, and produce the best sounding master possible with what I’m given to work with. Most of the time this is fairly easy, but sometimes I get into a situation where I must express some concerns about fulfilling a client’s request to continue making their master louder and louder with each revision.
I think it’s been settled that people have this fear of their song or album appearing to be too quiet when randomly played before or after something from another artist. This has caused the loudness war to spiral out of control. There was a loudness war in the days of vinyl too, but there were different limitations, variables, and obstacles to overcome with vinyl than we have now in the digital realm.
What many people don’t consider is that there is loudness that can be measured technically using various meters and analyzers, and then there is perceived loudness which is how loud your ears and brain process loudness. It’s possible for a song to have a loud reading on a digital meter or software analyzation, but it may not strike the listener as being very loud or intense.
Despite popular belief, this is not always because of the mastering. Things that can greatly affect the perceived loudness of a song can vary, but all of the following are related to some degree:
- The musical performance itself
- Recording techniques & styles
- Mixing techniques & styles
- Frequency content/response
- Transient peak quality
Low frequencies tend to cause higher readings on meters, but don’t exactly translate to more loudness to the human ear. A song with a great drummer, hitting drums very hard, and that are well recorded and mixed could have more perceived loudness but technically lower RMS (average) levels than another song that was not performed and recorded as well (or in the same manner), but actually reads louder when the RMS and other levels are analyzed.
In other words, your master may be technically louder than another master, but it may not sound that way because of the things previously mentioned.
Often times, these things can’t be solved in mastering and need to be addressed in the mixing stage. Attempting to push a master to a greater perceived loudness than another master can lead to an over-processed sounding master that is fatiguing and hard to listen to.
I’ve found that every song eventually reaches a certain loudness point in the digital realm where going any louder starts to have negative artifacts and negative impact on the overall sound quality. This breaking point can vary slightly depending on the person doing the mastering and the tools they are using, but in the digital realm, there is a fixed digital ceiling that cannot be exceeded, and it’s the same for all mastering engineers.
Some mastering engineers have tricks and tools for getting more perceived loudness out of a master even though we’re all working with the same fixed digital ceiling, but it’s all relative to the source material. Matching the RMS level of one song to another is not hard, but in mastering, making a poorly recorded song have the same perceived loudness as a well executed mix can often times be impossible.
My point is, I suggest avoiding temptation to push your master so loud that it starts to do more damage than good, just for the sake of not sounding a little quieter when comparing to material that was produced with a different set of musicians, engineers, standards, tools, budgets, etc.
A good mastering engineer has the skills and every intention of meeting your mastering needs, but I’ve seen and also personally been pushed to make a master so loud that it really starts to lose impact and destroys the material, just for the sake of not sounding too quiet when compared to the flavor of the month. In most cases you’re comparing apples to oranges.
We’re used to having our digital audio masters so loud these days that it doesn’t take much for parts of a song to start reaching the digital ceiling. From there on out, going any louder comes with artifacts. Up to a certain point, these artifacts can be positive and sound good if it’s skillfully done. Eventually a loudness point is reached where most would agree that more harm than good is occurring.
Another major factor is mastering for an album vs. a standalone single track. Mastering processing that works for one song on it’s own, may not work for the same song when it’s part of an album. This can be due to the various frequency content of some or all songs, the style and intensity of a song, or a combination of both.
I’ve had cases where I do a test master of one song for a client. They like the test master and have me do the entire album. As much as I try to use that test master as a benchmark or goal for the entire album, sometimes an album project contains songs that have substantially different dynamic ranges, styles, or other audio qualities from one another. This means that the loudness of the test master, might not work for that song in the context of the entire album.
If I am mastering a typical rock album, and the album ends with (or otherwise contains) a sparse acoustic song with no drums or major percussion, it takes some subjective listening to determine how loud the oddball acoustic track should be and still make sense in the context of the album. If it’s as loud as the rock songs, it wouldn’t sound natural in my opinion. If I were mastering the acoustic song on it’s own, I might be inclined to master it a little louder than what I would have in the context of the entire rock album.
I’m currently in the middle of a project in which a client is sending me screenshots of what he wants his master waveforms to look like, as opposed to how they should sound. This is generally a bad idea, and a bizarre way think about your music.
Don’t be afraid to adjust the output level on your listening device before you push your mastering engineer to squash the life out of your music.
Mastering is not a rubber stamp process. It’s an art in which compromises must be made in order to get a group of songs to work together, and translate to as many formats and playback systems as possible.
Stay tuned for part two coming soon.
Missing our best stuff?
Sign up to be the first to learn about new tutorials, sales, giveaways and more.