Time & Pitch [STD & ADV]

Overview

The Time & Pitch module uses the iZotope Radius™ processing algorithm to allow for independent control over the length and pitch of your audio. It is useful for retuning audio to fit in a mix better, or adjusting the length of a selection to account for changes in BPM or timecode.

iZotope Radius

iZotope Radius™ is a world-class time-stretching and pitch-shifting algorithm. You can easily change the pitch of a single instrument, voice, or entire ensemble while preserving the timing and acoustic space of the original recording. iZotope Radius is designed to preserve the natural timbral qualities of the original file, even when applying extreme pitch shifts.

Controls

Algorithm

The Algorithm drop-down menu has three options:

Radius

Designed to work well with polyphonic material such as mixes with more than one instrument, as well as non-harmonic material such as drum loops or rhythmic audio. This is the highest-quality option for most sources.

Radius RT

Good quality, polyphonic, but faster than Radius. If processing speed is important, use the Radius RT algorithm.

Solo Instrument

Designed for monophonic pitched material such as a stringed instrument or human voice. You should use Solo mode only when processing a single instrument with a clearly defined pitch. The human voice is a good candidate for solo mode, as are most stringed instruments, brass instruments, and woodwinds.

In Solo mode, the adaptive window size can significantly affect the quality of Radius’s output.

• If the adaptive window size is too small, you will hear a squeaking noise which sounds like the pitch of the audio is changing very rapidly.
  
• If the adaptive window size is too large then the sound will become grainy as you will begin to hear portions of it being repeated.
  
• A good approach is to start with the default window size of 37 ms. If the results are unsatisfactory, increase the window size until the squeaking noise described above does not occur. If you cannot get the distortion to disappear, switch to Radius mode for processing.
  
• Lower pitched instruments and voices may require a longer adaptive window size than the default, but very long adaptive window sizes can cause audible repeating slices of audio.
  
  

Formant Correction

Formants are the resonant frequency components of voice that tend to be perceived as characteristics like age and gender. You can shift formants independently of pitch and time by enabling Shift Formants.

• Typically you will leave the Formant Shift Strength set to 1 (full strength) and the Formant Shift Semitones set to 0.
  
• If you hear what sounds like an EQ adjustment to your audio, you can try lowering the strength to reduce this artifact.
  
• To achieve special effects, for example to change the perceived gender of a human voice, try adjusting the semitones to a value other than 0.
  
  

STRETCH RATIO

Determines how much the resulting audio will be stretched in time.

• Values between 12.5% and 100% will cause the audio to speed up without affecting pitch, resulting in a shorter audio file.
  
• Values between 100% and 800% will cause the audio to slow down without affecting pitch, giving you a longer audio file.
  
  

BPM CALCULATOR

If you are using Radius to process audio for a tempo change, you can also adjust the stretch ratio with the BPM Calculator.

PITCH SHIFT

Controls the amount of pitch shifting up or down that will be applied to the audio.

TRANSIENT SENSITIVITY

Determines the algorithm’s handling of transient material. Higher values will result in better preservation of individual transients after processing.

• When stretching percussive material, you usually want transient sensitivity set to its default value of 1.
  
• If transients in your audio are being “smeared”, a higher value of 2 will tighten up transience at the expense of incurring heavier processing on non-transient audio.
  
• Bowed instruments such as the violin and cello are especially affected by the transient sensitivity setting. If you hear a stuttering artifact, lower the transient sensitivity to eliminate it.
  
  

NOISE GENERATION (RADIUS MODE ONLY)

Helps noisy material (like sibilance or snare drums) sound more natural when processed.

• This control will generate noise instead of stretching the noise that is already present in the signal and creating new tones. Higher values of the noise generation parameter will cause Radius to generate noise more often, but can cause some phase artifacts.
  
  

PITCH COHERENCE (RADIUS MODE ONLY)

Controls the preservation of the natural timbre of the processed audio.

• The Pitch coherence control in the Radius control panel helps preserve the timbre for pitched solo voices, such as human speech, saxophone or vocals. While traditional vocoders can smear these signals in time and randomize phase, the pitch coherence parameter of Radius preserves phase coherence for these signals.
  
• High values of pitch coherence will avoid phasiness in Radius’s output at the expense of roughness (modulation) in processed polyphonic recordings.
  
• Try turning this up for better results if you’re processing a solo voice or a small group of related instruments.
  
  

PHASE COHERENCE (MIX MODE ONLY)

Preserves the coherence of phase between the left and right channels of the processed audio.

This should be increased if there’s any change in the perceived stereo image after using Radius. It can be decreased when processing a multichannel signal where different channels contain completely different instruments.

ADAPTIVE WINDOW SIZE (ms) (SOLO MODE ONLY)

Adjusts the window size in milliseconds of Radius’ Solo algorithm.

• If the adaptive window size is too small, you will hear a squeaking noise which sounds like the pitch of the audio is changing very rapidly.
  
• If the adaptive window size is too large then the sound will become grainy as you will begin to hear portions of it being repeated.
  
• Increase this if you have trouble getting good results pitching or stretching low-pitched instruments or voices.
  
  

SHIFT FORMANTS

Processes formant frequencies independently of other pitch and time processing.

• When this option is enabled, formant frequencies can be shifted independently of other pitch shifting performed by Radius.
  
• When Radius performs pitch-shifting without Formant Correction, it will shift these resonant frequencies along with the rest of the audio.
  
  

STRENGTH

Adjusts the amplitude strength of the formant correction filter.

SHIFT

How much formant frequencies are shifted. Typically this control can be set to 0, which leaves the formant frequencies unshifted during processing. Adjust this control to fine-tune the formant correction algorithm or for special effects.

WIDTH

Controls the bandwidth of the formant detection filter.

• Smaller values of this control will offer more precise formant correction in the processed audio.
  
• Higher values will include a wider band of formant frequencies.
  
  

More Information

Pitch shifting single instruments (especially bass instruments) can benefit from some adjustments to formant correction. Try enabling formant correction and moving the strength between 0.1 and 0.2. Move the Formant Correction semitones part of the way towards your pitch shift amount. For example, if you’re pitch shifting +4 semitones, move the Formant Correction Semitones between 2 and 3. This can help bring back subtle percussive elements in the original source material.

The formant frequencies of the human voice can actually shift slightly when we sing. You can use the Formant Correction Semitones control to compensate for this. For example, if pitch shifting a human voice by +7 semitones, try setting the Formant Correction semitones between 0 and +2 for more natural results.