Autotune is a great example of signal processing that can be used as an effect in overdrive or in transparent ways to enhance a performance. This tool is used by more artists than you might think. However, some people make it obvious.
Think of artists like Cher, T-Pain or even Lil Wayne rapping. They almost sound like robots when they sing. I believe that the decade 2010 will be referred to as a “dark era” in music. This is partly due to the regrettable future of many artists.
Consider all the artists who aren’t using it. I can assure you that they are all using it, but their mixing engineers do a great job. A great vocal performance is the main factor in an “invisible autotune”.
To push the effect into overdrive, you must intentionally have a poor performance to a certain extent. This is why it works so well for rap artists when they sing at the right pitch.
As far as I can see, there are three levels to autotune:
- Too much – Cher, T-Pain and Lil Wayne
- Too much – Country and Pop music
- The perfect – Every other genre you don’t see it in
Only the quality of the singer is the difference between the second stage and the third. Pop artists are chosen for their marketability, which is based on how well they sing.
I will show you how to get all three stages from sounding robotic to sounding synthesizer-like to undetectable perfection.
What is Autotune?
Autotune, an audio signal processor tool, was originally designed to correct off-key pitch in vocal performances or any other instrument. It can also be used to apply pitch correction so aggressively it sounds artificial and synthetic. Steinberg PitchCorrect
The term Autotune has come to be synonymous with pitch correction. However, it is a trademarked term and tool that Antares audio Technologies created. There are many DAW plugins that exist, each with their own names. However, the effect is commonly called “autotune”.
Cher’s 1998 single Believe changed pop music forever. While it was initially intended to disguise off-key pitches, she slightly adjusted them to lock on key. The effect is used in a dramatic way and was a big inspiration to musicians such as T-Pain who made it their own sound.
How to Autotune
This section will explain how autotune works, but it will be done from the perspective that polishing a great vocal performance is important. To make the plugin transparent and musical, you must create a great vocal track.
This means that you will sing the same song multiple times, creating a “best-of” track. You can correct any errors in pitch correction software only after you have compiled the best performance from multiple takes.
Please note: You must use pitch correction only on mono tracks. Stereo tracks don’t work out. You should also be free from reverb, delay and other audio effects.
1) APPLY THE PLUGIN to THE TRACK AND FAMILIARIZE WITH IT
I assume that you are familiar with digital audio software. It is not possible to predict which plugin you will use. I will be showing you the Logic Pro Pitch Correction plug-in. They all look the same, but they work differently.
Although I will show you Logic Pro’s default plugin, I will use terminology that is more likely to be found in all the plugins. Because I have several names for every parameter that you can adjust, it will be easy to identify what is which.
If you don’t know what the settings are, scroll down to Autotune Settings. Each setting is described and I also suggest starting points for each of the three levels mentioned above.
2) SELECT THE SCALE, KEY, AND INPUT TYPE
There will be two options. You have two options: you can choose a key (like G Minor), and a scale (Melodic major scale), or you can use the keyboard to manually enter it. Sometimes they will ask for the root note of the scale instead of a key. This would be G if you are using G Minor or G Major.
You will still need to adjust the notes available for tuning the keyboard if you select a key or scale from the list. Sometimes vocals don’t use all the notes in a scale. If this happens, you might not want them available.
You can also choose the chromatic scaling if you want to go for the robot rap vocal sound. This means that all possible notes are available in pitch correction targets.
3) CHOOSE YOUR RETRO PERFORMANCE
Once you have defined the effects, it is time to actually make it work. Now, we’ll be focusing on the slider or retune knob. You won’t be able to detect a slower speed than 400 ms or half a second. In most cases, it is too slow to do anything.
You can loop the vocals in solo mode so you can hear only the singer. Then, slow down the retune. You’ll start to see the meter indicate which note is being detected, and how flat or sharp it is in cents. Slowly increase this. This is the amount of pitch correction that must be applied.
The speed at which the pitch correction is applied to the note is known as the retune time. For example, if you have to change a note by 50 cents, it can happen instantly at zero milliseconds or over 100ms. The amount you choose will depend on what sound you desire.
You may wish it to be a little noticeable for electronic music. Set it to zero for the robot effect. You can play with the effect to make it transparent, but it will be slower than normal. It is impossible to determine the correct amount. You need to listen!
4) DECIDE HOW PERFECT YOUR VOCALS SHOULD BE
You have many options about how perfect and precise you want your vocals to sound, regardless of how fast you can “snap” to the right pitch. A human cannot hold a perfect note. Even the best can sag by a few cents. It will sound unbelievable if you get rid of that waving.
Flex Tune is a great option to bring back the human element in your vocal track. Logic Pro’s pitch correction does not have this option. The processor can adjust the pitch of the vocals by how much you specify. It basically tells the processor not to change the pitch if the pitch is “close enough”. This can help bring your track back to reality by adding a small amount.
If you have it, or another similar knob, now focus on the Humanize knob. The knob tells the plugin not to change to an unintended note, but to keep to a certain note if vocals begin to falter. It’s like a compressor release. This is useful if your singer keeps on holding onto notes at the end phrases.
Natural vibrato is an alternative that’s not always possible. It can detect when the vocalist uses purposeful vibrato, and allow you to adjust the intensity. This is not something I would recommend. This decision must be made while recording vocals .
5) UNSOLO VOCALS & TWEAK INSIDE CONTEXT OF FULL MIX
This is the final step. This is your last step. Now, listen to the whole mix and your work. The only thing that you will likely need to adjust is the retune speed.
You may want to use a robotic sound for electronic music such as dubstep and EDM. You’ll need it to be less obvious for pop, classical, jazz, rock or pop music, which have more natural sounding genres. Therefore, you might choose a slower retune speed.
You may find that your vocal pitches sound fantastic, but there are still issues with the tracks. To improve your quality, check out our How to Mix Vocals article. These principles should be applied to every track of vocals.
Two things I’d like to accomplish now. I need to describe the settings of your plugin and provide starting settings that correspond to the three types of pitch correction I have already mentioned.
PITCH CORRECTION OPTIONS
You can adjust the parameters of each pitch correction plugin. You will most likely be presented with five options.
- Key, Scale, and Input Type
- Retune Speed (Response Time)
- Flex Tune (Smoothness).
- Humanize (Release time)
- Natural Vibrato (Vibrato Adjustment)
These are the names they are on the official Autotune plugin. They are usually the same as what you will find on the other Autotune plugins, even if they are renamed to some degree.
Key and Scale Without telling the software what key or type of scale it is, it will not know which pitches to adjust. These can be selected from a list, such as major scale, harmonic minor, or other. You can also type it yourself using the keyboard layout.
There will be many input types available. You can choose from Normal (higher vocal registers), or Low (bass and baritone). Some input types will let you choose from terms such as Tenor, Alto, Soprano, and so on. This is because the software can expect certain octaves to be processed faster and more accurately.
Response Time (Retune Speed) This is measured in milliseconds. It will allow you to go to zero which means instantaneous (how quickly you get the robot sound). This is how fast a bad note can be retuned to the right one. It should not be set too fast. This will cause it to sound obvious and can also lead to glitches.
You will most likely want subtlety so use a slower value to ensure the listener does not notice the changes. This will give you a natural sound. If the performance is too fast, it will sound artificial.
Flex Tune (Smoothness). This option is not always available. This option adds smoothness to a processed song that was slower due to the singer’s pitch. This option smoothens out transitions to make them more subtle and less noticeable.
This may seem to be a way to prevent retuning, but it is more like allowing for a margin of error. This allows you to decide if a note is close enough in pitch to allow it to be processed and left alone. This helps to preserve the human element of the track, rather than making it perfect.
Humanize – This is similar to the release time on a compressor. This tells the pitch correction how to keep long notes in tune. This may not be necessary if your vocalist isn’t able to hold on for long sustained notes at the ends of phrases.
It’s used to correct accidental vibrato in long notes. It won’t sound “pitch perfect” if the singer’s voice is not pitch-perfect. It will snap to another note if the note is too weak. This prevents that from happening.
Natural Vibrato (Vibrato Adjustment ) – This option is not always available. This option can add vibrato to an otherwise empty space, which is kind of the opposite of Humanize. To sing vibrato well, you need talent. This option is available if your singer cannot do it, but the song requires it. You can also lower the vibrato’s intensity.
AUTOTUNE SETTINGS – SINGING
The three levels of pitch correction I have mentioned are “completely overkill”, “too much is required”, and “perfect and noticeable”. These are just some suggestions. You’ll need to modify them to suit your vocal tracks. Each time, the requirements will vary.
You can use the effect in a dramatic way like Cher or T.Pain by setting the flex tune speeds and retune speed to zero. This will instantly snap the notes to the correct pitches. You can increase the volume of the humanize knob and decrease natural vibrato.
The vocals will sound more like a synthesizer. It is possible to make the recording more dramatic by deliberately singing at a different pitch. To make the plugin work hard to re-pitch each note, you should be close to perfect but not quite perfect.
Try setting the retune speed at 5 ms to 15, with very little flex, to match the current expectations for country and pop music. To keep the track realistic, it is worth adjusting the retune speed just a little. You’ll have a track that is pitch-perfect but doesn’t “snap to the notes quickly” so it sounds more like a live performance.
A slower retune speed is best for natural sounding autotune. (Set it so that you can barely hear it, then back it off slightly), and perhaps 10% flexibility tune to allow for some waving. That’s it. You can leave the rest of the settings as is.
Pro Tip: If you don’t need it often, but certain parts of the performance are absolutely necessary, you can apply pitch correction. You can add the plugin to an automated track with a slow retune speed. Then, on the parts that you need to fix, increase the retune to make it correct the pitch.
What are the differences between Autotune plugins?
There are many pitch correction plugins. This may have led to some confusion about which one is best and which ones are better. All of them do the same job, but some have more options while others offer a surprising number of additional choices.
Here’s a list with some of the most popular options for pitch correction plugins:
- Celemony Melodyne
- Antares Auto-Tune (various editions)
- Wave Tune (various variations)
- Zynaptiq Pitchmap
- Melda MAutoPitch
- Synchro Arts Revoice Pro
- Izotope Nectar
- Logic Pro Pitch Correction
- Steinberg PitchCorrect
- Cakewalk/Roland V-Vocal
- Mu Technologies Mu Voice
There are many more options, but these are the most widely used. Antares Auto-Tune is the most common. Celemony Melodyne, a great but expensive choice that offers incredible options, especially when paired with the visual user interface, is another excellent option. This is what it looks like: Celemony Melodyne
There are two main differences: the paid versions have more options, while the “Pro” versions offer more options. Most users will not need the most advanced options, especially if they aren’t able to add pitch correction in real-time.
Pro Tip: Melodyne has a very transparent interface. You can adjust the pitch and tweak the cents to make the track sound perfect. Then bounce it. You can then use Autotune on the track to get a natural vocal performance.
This is how to use Autotune
It is easy to follow. Because the software is complex, the names are difficult to understand. The rest of the process is easy once you know what each knob does.
You should be able to use autotune just as easily as a professional thanks to our explanations and walkthrough. It’s as easy as turning the knobs slowly and listening with your ears to find the right settings. This plugin is truly revolutionary in its power and simplicity.