Skip to content

envelopes

Envelopes for Noise Modulation

The envelope functions in the Noise Module modify the amplitude of noise signals over time, simulating real-world non-stationary noise patterns. These functions introduce gradual or abrupt variations in noise intensity, making the generated noise more realistic for applications in signal processing, machine learning, and system testing.

Each envelope function follows a unique pattern: - Linear: Gradual increase or decrease in noise amplitude. - Sine: Periodic oscillations in amplitude. - Random Walk: Stochastic variations mimicking unpredictable changes. - Blockwise: Sudden step-like changes in amplitude.

These envelopes allow dynamic control over noise behavior, enhancing the realism of simulated noisy environments.

Envelope Functions Overview

envelope_linear

Location: noise/envelopes.py

Description

envelope_linear generates a linearly changing amplitude envelope, transitioning smoothly between two defined noise power levels. Depending on the param value, the envelope either ramps up or ramps down over time.

Parameters

  • num_samples (int):
    Number of samples in the envelope (must match the signal length).
  • npw (tuple of float):
    (min, max) range of noise power values.
  • param (float):
    Controls the direction of the transition:
  • > 0.5 → Increasing envelope.
  • ≤ 0.5 → Decreasing envelope.

Returns

  • envelope (numpy.ndarray):
    Linearly increasing or decreasing amplitude envelope.

Example

import numpy as np
import SigVarGen as svg

env = svg.envelope_linear(num_samples=1000, npw=(0.1, 1.0), param=0.7)
print(env[:10])  # First 10 values of the envelope

envelope_sine

Location: noise/envelopes.py

Description

envelope_sine applies periodic modulation to noise amplitude, creating fluctuations that mimic oscillatory environmental effects. This envelope is particularly useful for simulating periodic disturbances such as wind noise or cyclic variations in signal intensity.

Parameters

  • num_samples (int):
    Number of samples in the envelope.
  • npw (tuple of float):
    (min, max) range of noise power values.
  • param (float, optional, default=0.005):
    Frequency of oscillations (lower values result in slower oscillations).

Returns

  • envelope (numpy.ndarray):
    Sine wave-modulated amplitude envelope.

Example

env = svg.envelope_sine(num_samples=1000, npw=(0.1, 1.0), param=0.01)
print(env[:10])  # First 10 values of the envelope

envelope_random_walk

Location: noise/envelopes.py

Description

envelope_random_walk generates a stochastic amplitude envelope, where the noise power follows a random walk pattern. This mimics unpredictable fluctuations observed in real-world signals, such as biological noise or varying environmental conditions.

Parameters

  • num_samples (int):
    Number of samples in the envelope.
  • npw (tuple of float):
    (min, max) range of noise power values.
  • param (float, optional, default=0.01):
    Standard deviation of the random step size (higher values cause more variability). Higher value results in higher amplitude, which might be clipped.

Returns

  • envelope (numpy.ndarray):
    Stochastically varying amplitude envelope.

Example

env = svg.envelope_random_walk(num_samples=1000, npw=(0.1, 1.0), param=0.02)
print(env[:10])  # First 10 values of the envelope

envelope_blockwise

Location: noise/envelopes.py

Description

envelope_blockwise applies stepwise variations in noise power, creating block-like intensity changes. This is useful for simulating sudden shifts in noise levels, such as network congestion bursts or machine noise variations.

Parameters

  • num_samples (int):
    Number of samples in the envelope.
  • npw (tuple of float):
    (min, max) range of noise power values.
  • param (int, optional, default=100):
    Block size (number of samples per step).

Returns

  • envelope (numpy.ndarray):
    Stepwise amplitude envelope.

Example

env = svg.envelope_blockwise(num_samples=1000, npw=(0.1, 1.0), param=50)
print(env[:10])  # First 10 values of the envelope