The processing needed to achieve an amazing-sounding vocal can be organized into 3 main stages.

First is editing, second tuning, and lastly, introducing effects.

Advanced Vocal Editing.

• Editing is like the black sheep in audio - nobody really wants to talk about it, but if you start editing your vocals, you’ll notice a massive difference in the quality.

• Editing can be basic, such as adding fades where needed, isolating segments and adjusting the level with clip gain, or maybe realigning multiple vocal performances. All of these tasks can be performed in your DAW.

But for greater control over the vocal, you’ll likely need an FFT editor - like this RX platform.

When the vocal is imported, the audio is divided into thousands of bins, each with amplitude, frequency, and phase values.

This means I can adjust much smaller regions of the vocal.

For example, say I want less sibilance from the performance. If I used a de-esser, a wide array of frequencies would be affected - even if I set a narrow band. Additionally, the compressor will introduce a fair amount of distortion, changing the timbre of the region.

Or if I used basic editing and isolated the sibilance to a clip and reduced the clip’s gain, I’d avoid the distortion of a de-esser, but now I’ve attenuated all frequencies equally in that segment, not just the sibilance.

In an FFT editor, I can find the sibilance indicated by its higher amplitude, select only the relevant frequencies, account for any frequency variation over time, and then lower the gain of that specific instance.
If I lower it too much or accidentally affect the wrong range, I can always undo the change and try again.

So let’s listen to sibilance controlled with a de-esser or frequency-specific compressor, then sibilance controlled with basic editing, and finally, sibilance controlled within an FFT editor.

Controlling sibilance is really just the start of what you can do - say the vocalist had a dry mouth when singing, and now these small clicks can be heard in between vocal phrases.

Without FFT editing, you’d be stuck with this sound, and it would only get louder the more you process the vocal.

Again, find the section and highlight, reduce the amplitude with the gain module or function, and repeat for all instances. If attenuating it sounds unnatural, copy the same frequency range from another part of the vocal and paste it over the needed area.

Use the fade tool to create a natural blend between the original and the pasted range.

The same could be done for any electrical pops, clicks, plosives, or basically anything that needs to be attenuated.

Does it take a lot of time? Absolutely, but it will sound much better in the long run.

Let’s listen to a vocal I edited to have these electrical pops. I’ll attenuate and copy and paste sections in this editor to get rid of them - notice that they’re almost inaudible after editing.

Last up in the editor, let’s attenuate unneeded frequencies and then check the signal’s phase.

Odds are you’ll use an EQ with an HP filter to remove unneeded lows. But if you’re already editing the vocal, you might as well do this much more accurately with the editor.

Find the vocal’s fundamental or lowest frequency from the lowest sung note. Highlight the region below it and attenuate it. This increases headroom, gives other instruments in this space more room, and can even help loudness-based processers, for example, an RMS compressor, measure and subsequently process more accurately.

If you plan on heavily compressing or saturating the vocal, consider attenuating frequencies above 21kHz to avoid aliasing distortion. Again, select the range and use the gain module to reduce the amplitude.

Some of you may think this is strange, but it is incredibly similar to what an oversampling function does—except with FFT processing instead of a less accurate linear phase filter.

Lastly, check the phase of the recording - it’s possible that the peaks and troughs are offset, which could affect how a processor measures the signal. The phase module will suggest the rotation and can make the needed changes.

This doesn’t change the frequency response or the amplitude; it only changes the phase angle to remove any offset.

So, let’s listen to the original vocal and then all of the editing we’ve done so far. This includes changes to the sibilance, electrical clicks, frequency response, and phase.

Advanced Vocal Tuning

In some genres, you want to hear the tuning. For example, if you’re working on a Hip-Hop track and you want the vocals to have that familiar tuned sound, then an automatic tuner, like autotune or Metatune, is a great choice.

For almost every other genre, though, you don’t want to hear it.

You want the vocal to sound as if it was performed in tune, not like it was tuned.
To achieve that, a platform like Melodyne is a lot more suitable.

The quick method is to import the vocal, highlight all notes, and use the macro tuning tools to control the pitch center and the drift.

But for a more natural sound, try using a small amount of the macro tool before going in and fine-tuning each note.

Because the pitch typically shifts up or down in correlation with the next note, the relationship between notes is just as important as the tuning itself.

This means you’ll need to listen to a passage a few times, making minute adjustments as you go, to find the most natural-sounding transitions between notes without compromising too much on the tuning itself.

It’s difficult to explain, but say we only have 2 notes. One note is in tune while the other isn’t—the in-tune note has a pitch bend that leads naturally into the out-of-tune note, but the note itself sounds out of tune.

Do you tune the out-of-tune note and make the transition less natural? Do you leave it as is? Or do you try to find a happy medium between in-tune and natural?

In my opinion, the 3rd option sounds the best. With a little tweaking of the note’s pitch drift and vibrato, as well as tuning both notes to prioritize the relationship between them, you can achieve a natural but in-tune sound.

With that said, let’s listen to the previously edited vocal that’s been aggressively tuned using Melodyne.

Then, we’ll listen to the same vocal but carefully tuned to prioritize note transitions, as we discussed a moment ago.

Notice how the second option sounds just as in tune but refrains from sounding over-tuned.

Advanced Vocal Mixing

I’m not going to tell you how to compress, eq, or anything like that - I’m assuming if you’re watching this, you have a method that works well for you.

Instead, let me share a couple of advanced tips that should help elevate your vocal production.

If you want incredibly clean compression—compression with little to no artifacts, harmonic distortion, or alterations to the timbre—use this Compressor V4 by Toneboosters.

Set it to the punch algorithm, and use an attack above 13ms and a release above 50ms.

It’s by far the cleanest compressor I’ve found - If you want more info on this processor, I made a detailed video about it, which I’ll link in the description.

Next, if you want incredibly warm saturation, and by that, I mean saturation that introduces even-order harmonics so that perfect octaves of the fundamental are generated, you need asymmetrical waveshaping.

Most saturators don’t do this - instead, try this free Shape Lite plugin by Audec.

Select Flex Clip Asymmetrical and decrease the input to the lowest amount.

Set the N Curve to a value of 1, the P curve to a value of 0, and the P Liner to a value of 0.2.

Then, increase the input subtly until you achieve some saturation. If the vocal is peaking around -3dB, distortion will start once the input hits roughly -10 to -9dB.

Blend the distortion in with the mix dial and compensate for the lower gain with the output dial.

Alternatively, if you don’t mind having a little less control, keep the input and output at 0 and blend in the mix to about 1%-2%.

Just about anything you can do with a typical saturation plugin can be accomplished with this plugin. It doesn’t matter how well a saturation plugin’s interface is designed or how well it’s marketed, etc. They all need to follow the same framework for achieving harmonics.

Use the values I gave here as a starting point.

Lastly, if you can’t find the right volume for a vocal in relation to the mix, send the lead vocal to a bus and the instrumental to a bus.

Put LUFS meters on both. If they’re the same LUFS, you’ve established the same loudness relationship between the instrumental and vocal as almost 1,000 top charting singles over the past 60 years.

I’m not just saying that - a 2014 AES paper found that for 927 #1 singles, the loudness of the lead vocal matched the loudness of the instrumental.

Let’s listen to a vocal that’s compressed with the clean settings I showed you before being saturated with the free Shape Lite plugin.

Then, I’ll match the loudness of the lead vocal and the instrumental.

The vocal in this example isn’t done, but these tips should make a good addition to the current methods you’re using.