Skip to content

Transient Analyzer (Transient Response & Wavelet Analysis)

Transient Analyzer

Overview

A tool for detailed analysis of "transient sounds," such as instantaneous impact sounds or sounds whose frequency changes from moment to moment. It uses the Continuous Wavelet Transform (CWT) to visualize sounds while maintaining temporal resolution—capturing "when and at what pitch a sound occurred"—which is often difficult for standard FFT analysis to achieve.

☕ Coffee Break: How is Wavelet Transform different from FFT?

There is a magical tool for analyzing sound called "FFT" (Fast Fourier Transform), but FFT has one major weakness: its "timing (time) tends to get blurry." Because FFT analyzes a "block of sound over a certain period," it is great at handling long, continuous sounds like a steady "beeeeep," but it struggles to accurately pinpoint exactly "when" a momentary sound (transient sound) like a sharp "click!" occurred. To use an analogy, it's like trying to take a picture of a fast-moving object with a slow shutter speed—the image ends up blurry.

This is where the Wavelet Transform comes in! Wavelet (small wave) transform analyzes sound using "short wave puzzle pieces" that can stretch and shrink in length and width. When measuring high-pitched sounds (short waves), it shortens the puzzle piece to accurately capture "time." When measuring low-pitched sounds (long waves), it lengthens the puzzle piece to accurately capture "frequency." Thanks to this clever "zoom lens" like mechanism, it can clearly draw everything from an instantaneous click to a deep, rumbling bass onto a single image (scalogram) with perfect clarity!

Operation

Recording

  • Record: Pressing this button starts the recording. It stops automatically after acquiring data for the set time (Record Time).
  • Trigger: Like an oscilloscope, recording can be started at the moment a sound volume beyond a certain Level is detected. This is useful when you want to automatically capture a sound exactly when you clap your hands.

Analysis

  • After recording is complete, press the Analyze button to perform the Wavelet Transform.
    • Note: This process is computationally intensive and may take several seconds to show results.

Reading the Charts

  • Transient Waveform (Top): The waveform of the recorded sound (Time axis).
  • Wavelet Scalogram (Bottom):
    • Horizontal axis: Time
    • Vertical axis: Frequency (Logarithmic display)
    • Color: Intensity at that moment.
    • Similar to a spectrogram, but it analyzes low-pitched sounds broadly in the time direction and high-pitched sounds sharply. This allows for a well-balanced simultaneous display of instantaneous click sounds (time information) and low hums (frequency information).

Settings

Controls

  • Channel: Select the channel to record and analyze (Left / Right / Average).
  • Wavelet: Selection of the waveform (mother function) type used for analysis.
    • Options: cmor1.5-1.0, mexh (Mexican Hat), morl (Morlet), cgau1, gaus1.
    • Generally, the default setting cmor1.5-1.0 (Complex Morlet wavelet) is the most well-balanced and suitable choice.
  • Min / Max Freq: Specifies the frequency range (lower and upper limits) to be analyzed. By narrowing it down to the band you want to see, you can eliminate wasteful calculations and get clearer results.
  • Record Time: Sets the recording time (in seconds). The default is 0.5 seconds. Be careful not to make it too long, as the calculation time will become significantly longer.

Use Cases

  • Impact Sound Analysis: Examining the response when a pulse is input to a speaker (impulse response) or changes in sound components when something collides.
  • Evaluation of Instrument Attack: Detailed observation of how harmonics appear at the moment a musical instrument starts sounding. For example, you can clearly separate and observe the difference between the impact sound of the hammer the moment a piano key is struck, and the resonance of the strings that follows.