samples | The array to populate with frequency domain representations of audio samples. The array length must be a power of 2 (such as 128, 256, 512). Also, the length must not be less than 64 or greater than 8192. |
channel | The channel to sample from. |
window | The FFTWindow type to use when sampling. |
Provides the block of audio frequencies (spectrum data) of the AudioSource that is currently playing.
This method fills the array you pass as the samples
parameter with the spectrum data of the AudioSource.
The frequency domain represents the frequencies and amplitude of an audio signal. Spectrum data provides the raw frequency domain data of the audio, which you can use to create a spectrogram to visualize the data.
Audio frequency bands are ranges of sound frequencies that describe different parts of the audio spectrum (like sub-bass, bass, brilliance). The audio frequency bands are evenly spaced between 0 to half of the sampling rate. GetSpectrumData uses the sampling rate from AudioSettings.outputSampleRate, not the sampling rate in the audio clip.
Use window to reduce leakage or scalloping loss between audio frequency bins/bands.
Note: A more complex window type might be less efficient and worsen resolution (lobe width).
For related information, refer to AudioSource.GetOutputData, AudioListener.GetSpectrumData, AudioListener.GetOutputData.
using UnityEngine;
[RequireComponent(typeof(AudioSource))] public class AudioSourceGetSpectrumDataExample : MonoBehaviour { AudioSource m_MyAudioSource;
void Start() { m_MyAudioSource = GetComponent<AudioSource>(); } void Update() { float[] spectrum = new float[256];
m_MyAudioSource.GetSpectrumData(spectrum, 0, FFTWindow.Rectangular);
// Loop through the populated array // Start the loop from 1 and to 1 less than the length, so the loop can draw lines between adjacent bins.
for (int i = 1; i < spectrum.Length - 1; i++) { Debug.DrawLine(new Vector3(i - 1, spectrum[i] + 10, 0), new Vector3(i, spectrum[i + 1] + 10, 0), Color.red); Debug.DrawLine(new Vector3(i - 1, Mathf.Log(spectrum[i - 1]) + 10, 2), new Vector3(i, Mathf.Log(spectrum[i]) + 10, 2), Color.cyan); Debug.DrawLine(new Vector3(Mathf.Log(i - 1), spectrum[i - 1] - 10, 1), new Vector3(Mathf.Log(i), spectrum[i] - 10, 1), Color.green); Debug.DrawLine(new Vector3(Mathf.Log(i - 1), Mathf.Log(spectrum[i - 1]), 3), new Vector3(Mathf.Log(i), Mathf.Log(spectrum[i]), 3), Color.blue); } } }
numSamples | The number of samples to retrieve. Must be a power of 2. |
channel | The channel to sample from. |
window | The FFTWindow type to use when sampling. |
float[] Returns a block of the currently playing source's spectrum data.
This version of GetSpectrumData is obsolete.
This version of GetSpectrumData is obsolete. Use the other version of GetSpectrumData instead.
This variation of the function allocates a new array each time it is called. Use the other version instead for better performance.
Number of values (numSamples) must be a power of 2. (ie 128/256/512 etc). Min = 64. Max = 8192.
Use window to reduce leakage between frequency bins/bands.
Note, the more complex window type, the better the quality, but reduced speed.