Auto-tuning, voice pitch correction (and more)(Letto 85 volte)

---

Autotuning

And here we are at the Autotuning control, the last operational link in the editing process, which will allow us to improve the intonation of the singer and any other monophonic musical instrument that is untempered or poorly tuned.

ATTENTION

Before proceeding with the autotune settings, it is best to check that all editing operations have been completed; autotune is very sensitive and, in some cases, even a small post-production change to a single sound detail could produce unexpected results.

The autotuning function must be managed using a special plugin in two consecutive stages:

1. an initial phase for the general (approximate) setting of the plugin controls, which will remain constant for the entire track;
2. a second phase for setting variables, to be managed throughout the track using the automation features of the plugin itself (an operation known as “autotune editing”)

Regardless of brand and model, all quality autotunes will have roughly the same main controls, while some secondary functions will often be organised with different controls.

In the following explanation, the Antares Autotune plugin was used as a reference. software company which was the first to develop this type of processor and, together with Melodyne, holds the record for its widespread use.

Input Type and Tracking

The Type input allows you to optimise the plugin's response by choosing the most appropriate operating algorithm for the type of source to be processed, whether it is a female or male voice or a high or low extension instrument. Under normal circumstances, the Tracking control should be set to an intermediate position (50%), i.e. halfway between the extreme settings defined as “relaxed” (towards 1%) and ’choosy“ (towards 100%). 

What is the setting criterion?

• If an isolated monophonic track contains disturbing elements, such as various types of noise or interference caused by other sources (as happens in live recordings), which could be misinterpreted by the plugin, or in the case of particularly dysphonic, breathy or guttural voices, it is sometimes possible to obtain smoother results by increasing the value towards a more choosy; in this way, the plugin will react with greater tolerance to disturbances, reducing the risk of abnormal reactions; on the other hand, the retuning it may appear somewhat less precise and effective.
• moving the control to the position relaxed the accuracy of retuning will tend to improve, but you may encounter some abnormal plugin reactions more easily; 

However, bear in mind that once this control has been set to the correct compromise position, it can still be managed by means of automatic systems in order to optimise its “step-by-step” function where necessary.

Key and Scale

To fully understand how these two basic, interrelated functions work, it is necessary to have acquired some musical knowledge which, although elementary for a musician, may be difficult for a sound engineer who does not have musical know-how. In order to make the most of this basic function and thus obtain the best results from auto-tuning, it is necessary to master certain concepts of music theory, including: 

• concept of interval of’intonation
• concept of tone, semitone, paragraph e cent
• concept of colour scale e diatonic scale (or tonal scale)
• concept of shade e mode
• concept of tempered system e untempered system

To determine the best response from the plugin, it will be necessary to skilfully set certain modal-tonal parameters applicable to the entire piece or its various fragments, in order to perfectly define the specific notes of intonation on which the plugin should perform a ’straightening“ action on notes that have imperfect intonation.

Specifically:

• the control Key (Key – Shades) will enable the most appropriate modal-tonal system to be defined and its variations to be managed by means of automation; 
• with control Stairs (Staircase e Mode) we can decide whether to use the chromatic scale (which includes all 12 notes of the tempered system) as a reference for autotune, or one of the 12 diatonic scales that can be set with the aforementioned Key control (each of which can be transposed in the 7 natural modes, including major and minor), or even other more “exotic” scales.

IMPORTANT NOTE

If you do not have sufficient knowledge of the above, it would be advisable to seek assistance from a musician who is sufficiently knowledgeable on the subject.

Alternatively, you can use the Chromatic Scale setting; this will be less effective in terms of correction and is more likely to result in reactive inaccuracies, but it will allow you to use the plugin with a single setting that will be adaptable to any song; when the Scale control is set to Chromatic, the Key control will lose its meaning and any setting will be irrelevant.

With the chromatic scale setting, the process will be simplified, but the results will be less effective and sometimes less natural; therefore, you will need to resort more often to automation and manual editing to compensate for the plugin's inaccuracies.

The use of modal-tonal settings managed through the Key and Scale controls will allow you to manage the autotuning function more reliably, with more accurate and natural results. For this reason, we strongly recommend its use to those who have sufficient musical know-how to be able to use it with mastery. When using the autotune tonal system, they should also bear in mind that many songs contain tonal modulations and often use so-called “chromaticisms”, i.e. “altered” notes that temporarily depart from the rigid tonal structure of the seven notes.

In order to fully utilise the advantages of tonal setting, they must therefore divide the piece into specific fragments, so that they can identify and then set a specific tonality or mode for each of them, containing all the notes of the performance fragment. In fragments where this task becomes too complex and complicated (for example, in the case of long phrases rich in chromaticism), the chromatic mode can be used, which, containing all the notes, will “bypass” the problem (but worsening the accuracy of autotuning in the most critical passages).

To set different tonal settings for the various fragments throughout the timeline, simply assign automation functions to the Key and Scale controls, and then modify the values appropriately in each of the different fragments.

In the figure: Here is a good default setting for a sample song in A minor. After setting the key to A minor (the appropriate key for the song) using the Scale and Key parameters, we have set a default Retune Speed of 24 ms (conservative but effective), with an additional slowdown of 21 ms for long notes, set using the Humanise control, and a slight recovery of the vibrato otherwise dampened by autotune, using the value 2.1 assigned to the Natural Vibrato control. These starting values, during the autotune editing phase, can be better set while listening to the song analytically, and also modified dynamically using automation.

Bypass Notes and Remove Notes

After choosing between chromatic or modal-tonal mode, and in the latter case setting a specific key for each different fragment, the central display will show the list of the seven notes that make up the selected scale, while 12 (i.e. all) will appear when setting the chromatic system. With the Remove control, you can select certain notes from that scale, thereby instructing the plugin to ignore them and align with the nearest note among those that have not been excluded.

With the bypass control, on the other hand, while maintaining the notes always present in the plugin's reference scale, you can exclude the autotuning action only for the specific notes set in bypass; in this case, you can, for example, decide not to apply retuning to notes that accidentally respond poorly to the automatic retuning action.

In the current configuration of Antares' Autotune, however, this function cannot be automated and therefore its setting can only be used manually or as a fixed setting for the entire song. However, it is possible to install multiple autotunes on the track, each with different settings, which can be activated or deactivated alternately in the various steps using the bypass automation.

Retune Speed

This is the most important control in autotuning and therefore the most commonly used in automation. It determines the speed of the plugin's reaction during its pitch correction action. This speed is generally expressed on a scale ranging from 1 to 100 milliseconds (ms).

An extremely slow time may not have any noticeable effect (generally, the action will be insignificant between 100 and 60 ms, while it will begin to be increasingly noticeable as reaction times are progressively accelerated (from 60 ms down to 1 ms).

Faster values will allow for more precise retuning corrections, but sometimes at the expense of the naturalness of the performance, which could be partially or totally compromised, particularly for values below 10 ms.

Below are the most important reference settings for this control:

1. Between 100 and 55 ms – very slow response – the autotune action will often be so minimal as to be inaudible, irrelevant or of little influence, but as it approaches the value of 55 ms, it will begin to improve in effectiveness, while remaining extremely fluid and light.
2. Between 55 and 35 ms – slow response – the autotune action will be moderate and the result will always be very natural, particularly noticeable around 40 ms, a setting that will gently improve intonation while almost always respecting naturalness – this will be approximately the best setting for long sustained notes.
3. Between 35 and 15 ms – the response will gradually become more pronounced – 25 ms will generally be the average value that best expresses the balance between effectiveness and naturalness – around 15 ms, the response will be very efficient but will also result in some sporadic unnatural reactions – however, this is the best setting for fast and passing notes.
4. Between 15 and 5 ms – the action will be very pronounced and often unnatural, with times below 10 ms and approaching 5 ms, some notes will begin to sound slightly “robotic”.”
5. Below 5 ms, the notes of the singing will almost always sound unnatural, becoming completely “robotic” when reaching the minimum value of 1 ms (especially when using the diatonic note detection system, as we will see shortly).

The latter setting is the one that will allow you to achieve a mechanical-robotic effect, which can be heard in some productions, especially in the dance genre and in new rap tracks that are “vaguely” sung.

Important considerations regarding the tonal correction system

It is important to understand that the effectiveness of Speed retune is closely related to the number of ’active notes“ selected using the Key, Scale, Bypass and Remove controls, and to the width of the intervals between the active notes.

In the case of a chromatic scale setting, autotune will have to juggle with 12 semitone intervals (which are therefore very narrow), so even a slight imperfection in the performer's intonation, a hiccup, a vibrato, a noise or other similar causes will be enough to “confuse” the plugin, which will often hesitate in recognising some of the notes, sometimes attracting them towards a wrong intonation alignment; It should also be noted that any error or uncertainty in the plugin's reaction will become increasingly apparent when using faster retuning values.

The reliability of the system will tend to improve when the number of notes to be considered is lower and the width of the intervals between them is greater, i.e. in the diatonic-tonal system which: 

• limits useful notes and therefore “plugin orientation” choices to 7 possibilities instead of 12; 
• It consists of only seven intervals between the seven notes, which will have a semitone amplitude in two cases and a  a tone (i.e. double the semitone) in the remaining 5 cases

Consequently, autotune will work faster and more reliably, with the following advantages:

• faster, more accurate and progressive achievement of correct intonation
• better tolerance and management of vibrato, hiccups, noises and intonation inaccuracies produced by the performer

By further limiting the number of useful notes, increasingly precise and clear results will be obtained (for example, with the Remove control, we can exclude another three notes from the tonal system to better control a specific melodic phrase that is expressed on only four notes).

The use of a tonal system instead of a chromatic one will allow us to use lower Retune Speed values, while still ensuring the same or greater effectiveness in re-tuning, combined with a more natural process. 

Please also note that if you are trying to achieve the typical “robotic” special effect, operating in chromatic mode (i.e. with all 12 notes of the octave enabled) will produce a very bland effect even if you set the Retune Speed to the fastest values; the effect will become much more defined and striking by setting the tonal mode, and will become excellent in cases where the choice is further limited, for example to only the 5 notes of a pentatonic scale: in this case, thanks to the elimination of every semitone interval, the autotune action will be twice as fast, selective and precise, and the robotic effect will be very striking.

It should also be noted that the accuracy and quality of autotuning will be more precise and effective if the source is already naturally in tune; a naturally well-tuned performance processed with autotune set to a tonal scale and moderate retune speed values (e.g. 40 ms) can produce an excellent and extremely natural result, requiring little or no automation. And even the robotic effect, when required, will be better with a performer who is already naturally very in tune or has been previously well “adjusted” by means of manual tuning.

Detune and Transpose, Throat Length, Formant

These controls are particularly useful when the plugin is not used on the track insert, but rather for rendering, i.e. creating a new processed file in place of the original audio fragment. In this case, it will be possible and easy, for example, to transpose the pitch.  of a note or a phrase.

The process of Transpose Unfortunately, when transposing the entire tonal footprint, the dominant timbral resonances will shift to lower or higher frequencies (depending on the case) and in a more or less pronounced manner (depending on the extent of the transposition required), producing the well-known timbral distortions, defined as:

• “devilish voice” when transposed downwards
• “duck voice” when transposing upwards

To correct these anomalies, the button Formant will enable automatic control of the so-called “human mouth” (a sort of adjustment to compensate for the size of the mouth and the length of the vocal cords); this option has the task of preserving the original sound of the voice, compromised by transposition and/or autotuning, by equalising the resonances of the vocal emission and specific control. throat length will allow you to refine the above corrective action manually. Correcting the mouthpiece will yield very appreciable results for transpositions of a semitone, and certainly acceptable results for transpositions of one tone; however, the results obtained will become progressively less accurate as the required transposition moves further away from the original pitch.

IMPORTANT

Even when not using the transpose function, the Formant's automatic function should be set to always active, as it will work very well both during manual rendering and during the automatic process.

Humanise

Ignored by many, Humanise is, on the other hand, a very important control for autotune.

Its values, expressed from 0 ms upwards, will cause a certain delay in the plugin's pitch correction compared to the base values already set with the retune speed.

This delay will be selectively assigned by the plugin only to long, sustained notes in the performance (which the algorithm will be able to recognise), while short notes will continue to align with the faster reaction speed set by means of Retune Speed. This will allow for a more accurate result and a more effective overall pitch correction.

Example

Let's assume we set Retune Speed to 20 ms and Humanise to 20 ms as well: 20 ms will therefore be the reaction speed that the plugin will set for standard short notes. When encountering long notes, the reaction time will progressively delay up to the maximum set value of 40 ms (20 ms + 20 ms). This will save us from having to make numerous adjustments to the retune speed with automation, which can be used for the few notes that escape the plugin's control.

Natural Vibrato

Natural vibrato is another often overlooked automatic control that can save time by avoiding the need to manually set the appropriate Speed Retune variations on the automation timeline. During vibrato, it is necessary to slow down the retune speed values to avoid flattening it and reducing its natural expressiveness. Furthermore, when using a chromatic setting, the pitch modulation of the vibrato may interfere with the plugin's perception, producing unnatural and slightly out-of-tune results (this can also happen with the tonal setting, but much less frequently). With the Natural vibrato control set to the default centre value of 0, it will have no effect; by gradually increasing this value, it will perceive the typical melodisation produced by the vibrato, respecting its range in a fairly natural but controlled way; using the opposite (i.e. negative) values will further dampen the vibrato. In cases where the action of this control is not optimal, it can always be refined with automation.

Vibrato Creator

This control is one of the advanced features of autotune and will allow us to create  new vibratos where they have not been performed by the performer or where they are not very incisive. With this, we can occasionally (i.e. proceeding in stages with automation) add an extra touch of expression to certain sustained notes, using no fewer than eight controls, which may not be easy for the uninitiated to manage.

It should be noted that improper use of this (highly artificial) function could produce unrealistic and completely implausible results, which suggests that it should be used wisely and sparingly after appropriate training and only in cases where it is truly useful.

When used extensively, these controls can also be used to create interesting artificial effects, which may be suitable for certain audio-musical productions.

Autotune operational management

To better clarify how to use it, below is a chronological summary of its operation:

1. Proceed with autotuning only when all other editing operations have been completed. At this point, you will need to install the autotune as the first link in the insert plugin chain, i.e. Before of any other real-time processing plugin; this is highly recommended because, if this is not done, any residual noise that has not been eliminated and any changes made posthumously through plugins (dynamic, tonal or other types of processors) will be detected by the autotune after its adjustment, which could result in unexpected reactions, forcing us to modify its settings again; Therefore, it is important that it is at the beginning of the processing chain and that no subsequent editing operations are performed after its adjustment.
2. Enable automation for at least five basic plugin functions, namely: Key, Scale, Retune Speed, Humanise, and Natural Vibrato. The first two of these (Key and Scale) should only be used if you have the musical know-how to manage them appropriately. The same applies to points C, D and E below (highlighted here in bold italics). If you do not have this expertise, it may be better to set a Chromatic Scale, which will adapt to any tonal-modal system at the expense of some precision and fluidity of the plugin's intervention.

N.B.: > Only for those with sufficient know-how to manage these operations:

3. Detect the basic key of the song and then set the plugin controls to a defined diatonic key, either major or minor, as appropriate. in the case of a modal key other than major or minor, it will therefore be necessary to define the appropriate relative key (major, minor or modal) (for example, the natural key of D is the relative Dorian mode of the natural major key of C (the relativity of two modal scales is achieved when they use the same notes and therefore the same “key alterations”); if you are not a musician, ask for help
4. Analyse the piece and divide it into different episodes (or fragments), each of which requires different diatonic keys to be set, so that all the notes played in each sound fragment are contained within those belonging to the set tonal system; if you are not a musician, seek assistance.
5. In musical passages where chromaticism is too numerous and dense, it will be possible to temporarily set, through the automation of the Scale function, the chromatic (i.e. atonal) function, the use of which should be limited as much as possible so as not to diminish the effectiveness and precision of the autotuning function.

N.B.: > Otherwise, proceed from point B to point F, skipping C, D and E.

6. Set the Retune Speed to a preset value of approximately 20-25 ms, then listen to the entire track, modifying the aforementioned value when short notes are too often poorly reproduced. at this stage, you will need to define a more appropriate basic static setting, trying to obtain plugin responses that offer the best possible compromise between efficiency and naturalness, in order to achieve the best results with short and medium-length notes only, ignoring longer notes for the time being.
7. Now set the Humanise control to 0 ms and listen to the long notes: only where they tend to flatten and sound cold “like an organ” should you try extending the time of this control, which will generally produce the best results with (additional) times of 20 or 25 ms, making the glissandos sound more natural. then set the average value most appropriate for the piece, seeking a balance between the effectiveness of the intonation on long notes and their naturalness.
8. Identify several points of vibrato in the track and select the one with the greatest amplitude: if the original amplitude of the vibrato appears to have been reduced by the autotune action (if in doubt, you can make a comparison using the plugin bypass), you can increase the Default values (which are set to ZERO by default) using the “Natural Vibrato” control until you obtain a sufficiently satisfactory and natural vibrato; then check the other vibrato points and, where necessary, manage the values of this control with automation.
9. Listen to the entire track again, refining the static settings already made.
10. Finally, act on the timeline to automate Retune Speed and:

• decrease the reaction speed in areas that appear unnatural or even adjust it to the maximum value of 100 ms in areas that tend to worsen (or, better yet, you could bypass the plugin in those areas by automating the bypass itself)
• increase it as necessary in areas where corrective action appears insufficient

11. The same refinement in automation can be carried out for Humanise and Natural Vibrato controls.
12. If any notes or phrases are still out of tune and cannot be corrected by increasing the Retune Speed due to the onset of artefacts, consider the following in order: 

• Modify the Choosy-Relaxed control appropriately in automation.
• correct the pitch with a manual tuning operation (already described in a previous chapter), then listen to it again with autotune, setting an appropriate Retune Speed value for it.
• Replace the “defective” take (or single note) with another one from those available.
• As a last resort, if possible: ask the performer to repeat the imperfect musical passage, taking care to recreate the same recording conditions so as not to cause differences in the sound field and/or performance temperament.

13. Listen to everything again to check that everything is OK, preferably together with other people involved in the work and after a sufficient break; unfortunately, autotune is an insidious process, which will highlight any unnaturalness on first listening with “fresh ears”.   

In the figure: At the top, a vocal track and, below it, an example of dynamic adjustment along the timeline of the three main autotune parameters using the editing window, specifically (from top to bottom): Scale (choice of reference scale (12-note chromatic or 7-note diatonic)), Key (choice of key, valid only for the diatonic scale), Retune Speed (pitch recovery speed).