Professional Analog Mastering.
Professional Analog Mastering.
226K+ subscribers
The primary role of compression is to attenuate either peaks, or the average loudness depending on the compression type, and increase headroom.
With additional headroom, the overall signal can be amplified. In turn, quieter details are made louder, which reduces masking, and makes the signal sound denser, more impressive, and of course more detailed.
But with compression, peak or rms attenuation is the first step in the process. To increase quieter details, you first need to attenuate the peaks or rms to create more headroom. Then gain is applied post compression.
Maximization is similar in the end result; however, it works in reverse.
The quieter details are first amplified; meanwhile, a very high threshold with a soft knee and moderate ratio is used to ensure peaks don’t exceed a specific limit.
So amplification occurs until gradual compression is introduced.
By wave shaping quieter aspects of the signal while leaving the peaks alone for the most part, maximization can achieve this differently than a clipper would. If we observe an oscilloscope while using this Oxford Inflator, and compare it to a clipper, notice that both increase the quieter details; however, the peaks are shaped less with the inflator.
All this to say, if you want to increase quieter details, but want an alternative to traditional compression, limiting, clipping, and so on, then maximization or upward processing, is a great option.
Let’s take a listen to a compressor with post compression make up gain compared to an upward processor, set to the same loudness.
Notice that both increase quieter details; however, the timbre is different due to how the peaks are treated.
Watch the video to learn more >
This is one of my favorite ways to use compression when mastering. Controlling the dynamics of either the mid or the side image means you control the width.
If I compress the mid and not the side image, then whenever compression occurs to the mid, the side image is higher in amplitude relative to the mid image.
Subsequently, the master would sound wider.
If I compress the side but not the mid, then whenever compression occurs to the side, the mid image is higher in amplitude relative to the side image.
As a result, the master would sound more centered.
This is useful in an of itself, but if I add make up gain, especially automatic make up gain, it becomes more interesting.
When auto make up gain is introduced, then whichever channel is being compressed is also amplified, and made more detailed similar to how we discussed in the first part of this video.
So, if I compress the mid channel with auto make up gain, then it becomes more detailed and impressive, while the side remains the same. So the track sounds more centered. And of course opposite is true if I compress the side with auto make up gain.
This does something really interesting. Other forms of stereo expansion are static - for example if I use a side image bell filter to boost the highs, this area is constantly expanded throughout the track.
But with this method, the expansion would occur dynamically and in tandem with louder passage of the song.
It makes for a very natural sounding way to increase width, or to center the track, depending on which image you compress and if you use make up gain.
So, let’s listen to a master being expanded with this method. I’ll compress the side image and use make up gain. Notice that whenever compression occurs, the track dynamically expands.
Watch the video to learn more >
If vocal sibilance is too aggressive during mastering, ideally you could go back to the mix and adjust it there. Sometimes, that simply isn’t an option - when that’s the case, frequency specific compression, or de-essing, is the best alternative.
Weiss De-ess is the best option for this, but I’ll show you how to do it with a dynamic eq in a moment.
In short, I’ll switch its processing to mid and side, and de-link the channels.
Channel 1 is now the mid image, and channel 2 is the side.
On the side image, I’ll ensure that no compression is occurring.
On the mid image, I’ll center the bands on the sibilant range or ranges, and attenuate by up to 2dB. Notice that the attenuation is only occurring to the mid image - this is what we want.
Since the lead vocal is likely centered, the sibilance from the vocal will also primarily if not solely occupy the mid image.
By compressing the mid image whenever sibilance occurs, we can control the sibilance in a master transparently.
If you don’t have this plugin, you can do something very similar with a dynamic eq.
I’ll find the range responsible, then center a mid-image bell on that area. Next, I’ll make it dynamic, and carefully set the threshold so that attenuation only occurs when sibilance is present.
The point being, if you’re having to de-ess during mastering, you’re obviously dealing with a lot more signal than the simple lead vocal. By isolating the processing to the most problematic area, that is the relevant frequency range in the center of the stereo image, you can achieve de-essing without changing too much about the other instrumentation.
Let’s take a listen to mid-image de-essing, and notice how it controls the sibilance without affecting unrelated instrumentation.
Watch the video to learn more >
If you just want to control some dynamics with compression, so that later on when you’re limiting, you have more headroom, here’s what I’ve found to be the best option.
The subtle blend between peak and RMS compression, ideally with a mastering algorithm, will result in clean compression with increased headroom.
The FabFilter Pro-C2 is a great option, but I’d actually recommend this free plugin Kotelnikov instead.
It makes blending peak and RMS compression much easier and lets you use an internal side-chain to control what’s triggering the compression.
Set a low ratio, and gradually reduce the threshold until you achieve 0.5dB to 1.5dB of attenuation. This compressor isn’t for noticeably changing the sound, it’s for reducing dynamics as transparently as possible for the sake of increasing headroom.
Once you’ve achieved that, open your limiter or clipper depending on what you’re using, and see if you’ve reduced the amount of peak attenuation.
Adjust the balance between RMS and peak detection until you’ve controlled the peaks as much as you need, while still retaining the original timbre of the mix.
Then, increase the gain of your limiter to take advantage of the additional headroom.
So, it’s nothing fancy here, just balancing average loudness and peaks for the sake of using less limiting and making your attenuation less noticeable.