Professional Analog Mastering.
Professional Analog Mastering.
226K+ subscribers
A foundational chain follows a strict format to keep things as easily controllable as possible.
In other words, as the signal travels from one processor to the next, there’s very little conflicting processing - which is to say, no processor is undoing what another processor did before it.
For a vocal, here’s what I’d recommend.
Start with subtractive EQ to balance the vocal. That means an HP filter below the fundamental frequency range to cut out unneeded and unmusical frequencies; a bell filter to adjust the relationship between the fundamentals and low-order harmonics, usually to reduce some of the extra energy in the low mids.
Then, a bell filter in the mids to high mids to reduce some nasally tones, which are also higher order harmonics, or multiples of the fundamental.
Sibilance is unrelated to the fundamental frequencies and can be controlled better with a dynamics processor, either a de-esser or a dynamic EQ bell filter.
With the frequency response controlled, control the dynamics with a peak-down compressor.
Attenuate the peaks by 3-6dB, up to 8dB for aggressive genres, and introduce makeup gain to bring up quiet details. Then insert an upward processor like the Oxford inflator to bring quieter details up even further and create a dense, upfront sound. This plugin has gotten pretty affordable, but there is a free, nearly 1:1 remake on GitHub.
A processor like this will also introduce harmonic distortion that fills the spectrum and adds to the musical content of the frequency response. So, so far the frequency response, and the dynamics both from the top down and lower amplitudes up have been controlled.
This can now serve as the foundation for your vocal chain. After it, do whatever you see fit, and add any creative effects that work well for the song.
This can be reverb, delay, chorusing, pitch and amplitude modulation, heavy distortion or amp emulation - any creative effect will work well if built on top of this foundation.
Once you have your creative effects introduced, and you’re happy with them, a final EQ at the end of the chain lets you shape everything that came before it.
Let’s take a quick listen to this foundation so to speak. Then I’ll add a couple of different combinations of creative effects after, so you can hear how you can branch off from it.
Watch the video to learn more >
A little music theory can go a long way when mixing - especially when the vocal is the main focus of a song.
Personally, I know very little music theory, but I always make it a point to learn the root note of the vocal performance.
Once I know it, it affects a lot of decisions I make.
To find it, I typically just use a frequency analyzer and see what fundamental frequency occurs most often. The fundamental is easy to spot, it’s the first high-amplitude spike - again, it’ll move around during the performance as the vocalist changes notes, but if I pick the most frequently occurring one, odds are I’ve found the root note.
The Pro-Q 3 has a piano roll at the bottom which makes it easy to find what the note is, but if you’re using a different frequency analyzer, just look up the frequency.
For example, if the frequency is 110Hz, you could just Google, ‘What note is 110Hz?’, and you’ll have an answer.
Now, remember back to what I said in the first section of the video - the nasal tones will be higher-order harmonics. If I picked the note that’s occurring most often, then the nasal resonances will most often be multiples of the fundamental frequency.
So, if the fundamental is 110Hz, odds are the nasal resonances occur at 770Hz, 880Hz, 990Hz, 1100Hz, and maybe 1210Hz. These are the 7th, 8th, 9th, 10th, and 11th order harmonics respectively. Again, these harmonics are just whole number multiples of the fundamental, so 110Hz x 7, 110Hz x 8, and so on.
These are fantastic starting points for finding what to attenuate - the best choice, in my opinion, is to find the harmonic highest in amplitude, center a bell on that, and vary the Q value so that the area around it is attenuated but to a lesser extent than the primary nasally harmonic.
Of course, use your ears, the vocal will interact uniquely with the surrounding instrumentation, which may make attenuating a different frequency more effective, but knowing this offers a lot of guidance as to where to start.
The same could be said for frequencies that I want more of.
If I want a super full-sounding vocal, I could amplify one of the fundamentals and shape the filter with the Q to amplify the other fundamentals as I see fit.
If I want more clarity, then I could amplify a multiple of the fundamental between 2kHz-5Khz or the vocal’s 3rd formant. This way I’m boosting the area with a lot of vowel and consonant information while emphasizing the musical aspects of the performance.
This is a lot like what saturation does - but instead of generating overtones through waveshaping, I’m amplifying the overtones that exist naturally in the vocal.
You could even do this with a vocal’s air - at that point there probably aren’t any natural harmonics to adjust, but if there’s signal occupying the frequency that relates to the root note, it can be boosted to introduce a small amount of extra musicality into the performance.
Aside from EQ, this could be used creatively. For example, say I use reverb as a send.
I could follow it with a linear phase EQ and amplify in-key frequencies in the ranges I want more of.
The same could be done to the output of any time-based or temporal effect.
So, it’s information that can help you make more musical decisions, as well as guide you to finding the most likely culprits for issues in the frequency balance.
So, let’s take a listen to the vocal affected with filters that target the root note and its overtones. Notice how effective the EQ is, since the bells are centered on frequencies that are present in the vocal, and not just in their general area.
Watch the video to learn more >
Most of you probably know this, but if this keeps one person from making the time-consuming mistake I did, then it’s worth covering.
When I first started making records I remember using automation to try and control the volume of my vocal at various points in the song.
Then, I’d add a new effect or maybe change a setting on one, and suddenly everything sounded different or the effect I expected to happen didn’t happen.
This seems obvious to me now, but at the time I had no idea that all volume automation was occurring after all of my effects.
Nothing I was doing was affecting the original vocal, only the processed vocal. When I found a section that was hitting the compressor too hard and automated it, that did absolutely nothing to solve the problem. When the vocal changed amplitudes during a different take and now the processing sounded different, volume automation did not help.
Routing is truly the most important aspect of all things audio-related, and mixing is no exception.
The routing goes:
The file stored on your computer,The waveform or track as you see it in your DAW,All effects and parallel sendsThe volume faderA bus if the vocal is routed to one, and if not, then directly to the stereo output
Clearly, by adjusting the fader with volume automation, I misunderstood the very strict cause-and-effect relationship that’s inseparable from the signal path. What I needed to do was adjust the clip gain. This directly adjusts the gain of the signal prior to processing.
All this to say, don’t touch volume automation, or even effect automation until you have your chain established. If you want more or less signal going into your effects chain, use clip gain.
If you want to make a volume adjustment to the already processed vocal - maybe you want the vocal up 1dB during the chorus or something, that’s when you use automation.