Professional Analog Mastering.
Professional Analog Mastering.
226K+ subscribers
We’re going to start our processing on the vocal channel - once processed, the signal will be routed to an aux track that includes a unique parallel technique. Lastly, both the channel and parallel signal will be routed to the same bus, on which we’ll introduce some final processors.
So the vocal starts on the channel, gets sent to a parallel track, then both the output of the channel and the parallel track meet at the same bus.
We’ll cover the processing used shortly, but let’s take a quick listen to the original vocal and the processed vocal so you can hear what we’re working toward.
Watch the video to learn more >
Our chain starts simple but will get more complex as it goes on.
It starts with EQ, then de-essing, peak compression, wave shaping or upward compression, then saturation. You’re welcome to change the routing or the processor types used, but this is a great starting point.
Also, tuning is optional and doesn’t contribute to whether or not the vocal sounds crisp, but for this track, I’ll insert a tuner after the compressor to avoid amplifying artifacts with post-compression makeup gain.
The first EQ is subtractive - with it I’ll attenuate unneeded lows with a HP up to before the fundamental. I’ll dip the area around 250Hz slightly to reduce masking to high frequencies, and attenuate resonances from the nasal cavity that vary from performance to performance - but are between 800Hz and 1.3kHz.
Right now, I don’t want to amplify anything - I’ll save this for later. Additionally, I’ll leave sibilance alone and let the de-esser handle this in a dynamic way.
With the de-esser, I’ll use this exactly how you’d expect. I’ll isolated the attenuation to the offending frequencies, using a split band setting, and achieve anywhere between 1-4dB of attenuation as needed. I want this range balanced so that I can amplify it later without the sibilance becoming harsh.
For the compressor, this is where some specific settings are helpful. Ideally, find a compressor with lookahead, and enable it - if possible, by 1-3ms.
Use a softer knee with a 3:1 to 5:1 ratio, and reduce the threshold until you achieve 6-8dB of attenuation. Use a very quick attack and a release of 50ms. Lastly, use make up gain.
Let me show you what these settings do - the lookahead and short attack cause attenuation very quickly. Whereas no lookahead results in more gradual attenuation of the transient, the settings we have here compress the full signal.
Subsequently, make up gain will amplify more of the vocal’s quieter details - this contributes significantly to the quintessential upfront sound thats needed to create crisp vocals.
With this free wave shaper we can make these details even more pronounced - to use it I’ll create a variable in the low amplitude range and drag it upward to amplify it. Too much and you’ll hear some ugly intermodulation distortion, but just enough and it can solidify the vocal in the front of the mix.
Last up on the channel strip we’ll saturate the vocal. Since we’ve balanced the vocal with EQ and de-essing, and then brought up a bunch of quieter details, the harmonics should help fill the remaining gaps without emphasizing any unwanted qualities.
The saturator you pick is of course up to you, but I find that tube saturation emphasizes the highs slightly and can even accent some of the transients, which is helpful for creating crisp vocals.
So again, the chain is pretty straightforward for now, but it sets us up well for more advanced techniques.
Let’s listen to what the processing on the channel strip is doing to the vocal.
Watch the video to learn more >
Now that we’ve finished processing the channel strip, let’s create a send or parallel track that has the same amplitude as the main track.
First, we’ll insert a linear phase EQ and with a HP filter, isolate the high frequencies and then boost the air with a high shelf. Make sure it’s linear phase or you’ll have phase interference between this parallel signal and the channel’s signal.
Next, we’ll insert a compressor and heavily compress. For a cleaner sound, and if you have the fab filter pro-c2, the mastering option works well since it introduces very little distortion. However, pick a compressor that adds an enjoyable timbre to the particular performance you’re working on.
For a crisp sound, a compressor type that adds some distortion helps - try using no lookahead, a hard knee, higher ratio, an attack of about 20ms to cut into and distort the transient, and a faster release. Again, experiment to find what works for your vocal.
After compression comes bright reverb. The reverb you pick is also subjective, but try to find something that emphasizes high frequencies or has the option to equalize the reflections.
Blend the reverb in subtly, and then blend these dense reverberated high frequencies in with the original vocal.
For an even brighter sound, you could use an exciter after the reverb - fresh air is a great free option that both equalizes and introduces harmonics to the high range, causing a bright sound.
Let’s listen to what this parallel track sounds like, both soloed and blended in with the original vocal. Notice how the highs become bright, the vocal sounds more detailed overall, and how we have a lot of control over the high range by adjusting the channel fader.
Watch the video to learn more >
The vocal is sounding a lot better, but we still have some processing to add to the collection of the original channel’s signal and the parallel high frequency reflections we created.
I’ll change the outputs of the channel and the parallel track to the same bus.
Now I can control the level of the overall vocal and process everything collectively.
First, I’ll add a short delay and blend it in with the plugin’s wet/dry. This will thicken everything we’ve done so far. 200ms delay is a good starting point.
Then, I’ll add another reverb - this can be a more traditional reverb but again, it’s subjective. Just be sure to blend it in so that the vocal doesn’t become washed out, which would undo everything we’ve been working toward so far.
From here you have a couple of options. With delay and reverb, our originally mono vocal now has variation between the left and right channels - in other words, we have both a mid image and now, a side image.
One great sounding step at this point is side-image transient expansion. It sounds like it wouldn’t work, but adding impact and punch to the side image really helps a vocal sound aggressive and crisp. This free Flash transient expander is a good option.
Alternatively, you could place an EQ here, using M/S filters to emphasize what you want more of. I’ve found that amplifying 2-5kHz on the mid image helps the vocal cut through. Then a boosted bell on the side image in the low mids creates impressive width, while a mild cut to the same range on the mid image adds more clarity again by reducing masking in the highs.
Or you could add an instance of Fresh Air and use subtle settings. Or, depending on the vocal, a combination of these processors to varying degrees. Personally I really like ending the chain with EQ since I can now shape the entirety of the vocal to work with the other instrumentation.
So, let’s listen to what these processors add to the vocal. Notice how they give the vocal a finished and polished quality, help the vocal work well with the instrumental, and create the crisp upfront sound indicative of modern music.