下方圖表顯示人類觸覺感知偵測閾值 (以加速度表示)，以時間頻率做為函數。閾值資料是根據 Bolanowski Jr. 圖 1 中的位移閾值轉換而來，S. J. 等人於 1988 年發表的文章「Four channels mediate the mechanical aspects of touch」。
人類對振動強度的感知 (感知測量) 並不會隨著振動幅度 (物理參數) 線性增加。感知強度以感知音量 (SL) 為特徵，其定義為在相同頻率下，高於偵測閾值的 dB 值。
對應的振動加速度振幅 (以 G 峰值為單位) 的計算方式如下：
$$
Amplitude(G) = 10^{Amplitude(db)/20}
$$
其中振幅 dB 是特定頻率下 SL 和偵測閾值 (相鄰圖表中沿垂直軸的值) 的總和。
下方圖表顯示 10、20、30、40 和 50 dB SL 的振動加速度等級，以及人類觸覺感知偵測閾值 (0 dB SL)，以時間頻率為函式。這項資料是根據 Verrillo, R. T. 等人於 1969 年發表的論文「Sensation magnitude of vibrotactile stimuli」。
valtimings:LongArray=longArrayOf(25,25,50,25,25,25,25,25,25,25,75,25,25,300,25,25,150,25,25,25)valamplitudes:IntArray=intArrayOf(38,77,79,84,92,99,121,143,180,217,255,170,85,0,85,170,255,170,85,0)valrepeatIndex=-1// Do not repeat.vibrator.vibrate(VibrationEffect.createWaveform(timings,amplitudes,repeatIndex))
Java
long[]timings=newlong[]{25,25,50,25,25,25,25,25,25,25,75,25,25,300,25,25,150,25,25,25};int[]amplitudes=newint[]{38,77,79,84,92,99,121,143,180,217,255,170,85,0,85,170,255,170,85,0};intrepeatIndex=-1;// Do not repeat.vibrator.vibrate(VibrationEffect.createWaveform(timings,amplitudes,repeatIndex));
[null,null,["上次更新時間：2025-07-27 (世界標準時間)。"],[],[],null,["# Analyze vibration waveforms\n\nThe most common vibration actuators on Android devices are [linear resonant\nactuators (LRAs)](https://medium.com/@SomaticLabs/what-is-a-linear-resonant-actuator-81cc25f85779). LRAs simulate the feeling of a button click\non what is otherwise an unresponsive glass surface. A clear and crisp click\nfeedback signal typically lasts between 10 and 20 milliseconds in duration. This\nsensation makes user interactions feel more natural. For virtual keyboards, this\nclick feedback can increase typing speed and reduce errors.\n\nLRAs have a few common [resonant frequencies](https://en.wikipedia.org/wiki/Resonance#Examples):\n\n- Some LRAs had resonant frequencies in the 200 to 300 Hz range, which coincides with the frequency at which human skin is most sensitive to vibration. The sensation of vibrations at this frequency range are usually described as smooth, sharp, and penetrating.\n- Other models of LRAs have lower resonance frequencies, at around 150 Hz. The sensation is qualitatively softer and fuller (in dimension). \nComponents of a linear resonant actuator (LRA).\n\n\u003cbr /\u003e\n\nGiven the same input voltage at two different frequencies, the vibration output\namplitudes can be different. The further away the frequency is from the LRA's\nresonant frequency, the lower its vibration amplitude.\n\nA given device's haptic effects use both the vibration actuator and its driver.\nHaptic drivers that include overdrive and active braking features can reduce the\nrise time and ringing of LRAs, leading to a more responsive and clear vibration.\n\nVibrator output acceleration\n----------------------------\n\n\nThe frequency-to-output-acceleration mapping (FOAM) describes the maximum\nachievable output acceleration (in G peak) at a given vibration frequency (in\nHertz). Starting in Android 16 (API level 36), the platform provides built-in\nsupport for this mapping through the `VibratorFrequencyProfile`. You can use\nthis class, along with the [basic](/reference/android/os/VibrationEffect.BasicEnvelopeBuilder) and [advanced](/reference/android/os/VibrationEffect.WaveformEnvelopeBuilder) envelope APIs, to create\nhaptic effects.\n\nMost LRA motors have a single peak in their FOAM, typically near their resonant\nfrequency. Acceleration generally decreases exponentially as frequency deviates\nfrom this range. The curve may not be symmetrical and might feature a plateau\naround the resonant frequency to protect the motor from damage.\n\nThe adjacent plot shows an example FOAM for an LRA motor. \nExample FOAM for an LRA motor.\n\n\u003cbr /\u003e\n\n### Human perception detection threshold\n\n\nThe *human perception detection threshold* refers to the minimum acceleration of\na vibration that a person can reliably detect. This level varies based on the\nvibration frequency.\n\nThe adjacent plot shows the human haptic perception detection threshold, in\nacceleration, as a function of temporal frequency. The threshold data is\nconverted from displacement threshold in Figure 1 of Bolanowski Jr., S. J., et\nal.'s 1988 article,\n[\"Four channels mediate the mechanical aspects of touch.\"](https://pubmed.ncbi.nlm.nih.gov/3209773/).\n\nAndroid automatically handles this threshold in the `BasicEnvelopeBuilder`,\nwhich verifies that all effects use a frequency range that prodcues vibration\namplitudes that exceed the human perception detection threshold by at least\n10 dB. \nHuman haptic perception detection threshold.\n\n\u003cbr /\u003e\n\nAn online tutorial further explains the [conversion between acceleration\namplitude and displacement amplitude](https://www.tangerinex.com/tutorial-1).\n\n### Vibration acceleration levels\n\n\nHuman perception of vibration intensity, a *perception* measure, doesn't grow\nlinearly with vibration amplitude, a *physical* parameter. Perceived intensity\nis characterized by sensation level (SL), which is defined as a dB amount above\nthe detection threshold at the same frequency.\n\nThe corresponding vibration acceleration amplitude (in G peak) can be calculated\nas follows: \n$$ Amplitude(G) = 10\\^{Amplitude(db)/20} $$\n\n...where the amplitude dB is the sum of SL and detection threshold---the value\nalong the vertical axis in the adjacent plot---at a particular frequency.\n\nThe adjacent plot shows the vibration acceleration levels at 10, 20, 30, 40 and\n50 dB SL, along with the human haptic perception detection threshold (0 dB SL),\nas a function of temporal frequency. The data is estimated from Figure 8 in\nVerrillo, R. T., et al.'s 1969 article, [\"Sensation magnitude of vibrotactile\nstimuli.\"](https://link.springer.com/article/10.3758/BF03212793). \nVibration acceleration levels.\n\n\u003cbr /\u003e\n\nAndroid automatically handles this conversion in the `BasicEnvelopeBuilder`,\nwhich takes values as normalized intensities in the sensation level space (dB\nSL) and converts them to output acceleration. The `WaveformEnvelopeBuilder`, on\nthe other hand, doesn't apply this conversion and takes values as normalized\noutput acceleration amplitudes in the acceleration space (Gs) instead. The\nenvelope API assumes that, when a designer or developer thinks about changes in\nvibration strength, they expect the perceived intensity to follow a piecewise\nlinear envelope.\n\nDefault waveform smoothing on devices\n-------------------------------------\n\nFor illustration, consider how a custom waveform pattern behaves on a generic\ndevice: \n\n### Kotlin\n\n val timings: LongArray = longArrayOf(50, 50, 50, 50, 50, 100, 350, 250)\n val amplitudes: IntArray = intArrayOf(77, 79, 84, 99, 143, 255, 0, 255)\n val repeatIndex = -1 // Don't repeat.\n\n vibrator.vibrate(VibrationEffect.createWaveform(timings, amplitudes, repeatIndex))\n\n### Java\n\n long[] timings = new long[] { 50, 50, 50, 50, 50, 100, 350, 250 };\n int[] amplitudes = new int[] { 77, 79, 84, 99, 143, 255, 0, 255 };\n int repeatIndex = -1 // Don't repeat.\n\n vibrator.vibrate(VibrationEffect.createWaveform(timings, amplitudes, repeatIndex));\n\nThe following plots show the input waveform and output acceleration\ncorresponding to the preceding code snippets. Note that the acceleration\nincreases gradually, not suddenly, whenever there is a step change of amplitude\nin the pattern---that is, at 0ms, 150ms, 200ms, 250ms, and 700ms. There is also an\novershoot at each step change of amplitude, and there is visible ringing that\nlasts at least 50ms when the input amplitude suddenly drops to 0.\n\n\nPlot of step function input waveform. \nPlot of actual measured waveform, showing more organic transitions between levels.\n\n\u003cbr /\u003e\n\nImproved haptic pattern\n-----------------------\n\nTo avoid overshoot and reduce ringing time, change the amplitudes more\ngradually. The following shows the waveform and acceleration plots of the\nrevised version: \n\n### Kotlin\n\n val timings: LongArray = longArrayOf(\n 25, 25, 50, 25, 25, 25, 25, 25, 25, 25, 75, 25, 25,\n 300, 25, 25, 150, 25, 25, 25\n )\n val amplitudes: IntArray = intArrayOf(\n 38, 77, 79, 84, 92, 99, 121, 143, 180, 217, 255, 170, 85,\n 0, 85, 170, 255, 170, 85, 0\n )\n val repeatIndex = -1 // Do not repeat.\n\n vibrator.vibrate(VibrationEffect.createWaveform(timings, amplitudes, repeatIndex))\n\n### Java\n\n long[] timings = new long[] {\n 25, 25, 50, 25, 25, 25, 25, 25, 25, 25, 75, 25, 25,\n 300, 25, 25, 150, 25, 25, 25\n };\n int[] amplitudes = new int[] {\n 38, 77, 79, 84, 92, 99, 121, 143, 180, 217, 255, 170, 85,\n 0, 85, 170, 255, 170, 85, 0\n };\n int repeatIndex = -1; // Do not repeat.\n\n vibrator.vibrate(VibrationEffect.createWaveform(timings, amplitudes, repeatIndex));\n\n\nPlot of input waveform with additional steps. \nPlot of measured waveform, showing smoother transitions.\n\n\u003cbr /\u003e\n\nCreate more complex haptic effects\n----------------------------------\n\nOther elements in a satisfying click response are more intricate, requiring some\nknowledge of the LRA used in a device. For best results, use the device's\npre-fabricated waveforms and platform-provided constants, which let you do the\nfollowing:\n\n- Perform clear effects and [primitives](/develop/ui/views/haptics/custom-haptic-effects#primitives).\n- Concatenate them to compose new haptic effects.\n\nThese predefined haptic constants and primitives can greatly speed up your work\nwhile creating high-quality haptic effects."]]
