Skip to main content
34 votes
Accepted

Synchrosqueezing Wavelet Transform explanation?

Synchrosqueezing is a powerful reassignment method. To grasp its mechanisms, we dissect the (continuous) Wavelet Transform, and how its pitfalls can be remedied. Physical and statistical ...
OverLordGoldDragon's user avatar
29 votes
Accepted

Wavelet Scattering explanation?

Wavelet Scattering is an equivalent deep convolutional network, formed by cascade of wavelets, modulus nonlinearities, and lowpass filters. It yields representations that are time-shift invariant, ...
OverLordGoldDragon's user avatar
17 votes
Accepted

Why LTI system cannot generate new frequencies?

One of the definitive features of LTI systems is that they cannot generate any new frequencies which are not already present in their inputs. One way to see why this is so, comes by observing the ...
Fat32's user avatar
  • 28.3k
11 votes

A case that zero padding increase real resolution and extract more info than naive DFT?

‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎ ‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎ ‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎ ‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎ ‎‎‎‎‎‎‎‎
OverLordGoldDragon's user avatar
10 votes

Why LTI system cannot generate new frequencies?

You can make a simple algebraic argument, given the premise that you provided. If: $$ Y(\omega) = X(\omega) H(\omega) $$ where $X(\omega)$ is the spectrum of the input signal and $H(\omega$) is the ...
Jason R's user avatar
  • 24.7k
9 votes

A system that perfoms Fourier Transform operation - is it an LTI system?

The Fourier transform operator $\mathscr{F}$ is a linear one; i.e., $$\mathscr{F}\{x(t)\}=X(f) ~,~ \mathscr{F}\{y(t)\}=Y(f) \implies \mathscr{F}\{\alpha x(t) + \beta y(t) \} = \alpha X(f) + \beta Y(...
Fat32's user avatar
  • 28.3k
9 votes

Discrete wavelet transform; how to interpret approximation and detail coefficients?

Wavelet transforms can be more difficult to interpret than FFT at face value due to the various representations, nomenclature and output formats. I had to study more than 15 resources to get a good ...
khuang834's user avatar
  • 191
9 votes

Synchrosqueezing Wavelet Transform explanation?

Low-level intuition can be obtained by inspecting the phase transform, visually. Answer complements and is complemented by this one. (-- Answer code) We consider a pure sinusoidal tone; ideas extend ...
OverLordGoldDragon's user avatar
8 votes
Accepted

How to validate a wavelet filterbank (CWT)?

Wavelets isn't just sampling with scales from some min to max - but it is what many implementations do, including scipy and <...
OverLordGoldDragon's user avatar
8 votes
Accepted

Equivalence between "windowed Fourier transform" and STFT as convolutions/filtering

This answer isn't fully developed (but is correct), I may update in the future. Indeed, STFT is strided convolutions, and there's many practical caveats, discussed in this answer. Standard vs improved ...
OverLordGoldDragon's user avatar
7 votes

What's the difference between the Gabor and Morlet wavelets?

For those looking for a compact description, I found this nice docstring while inspecting the kymatio GitHub repository: A Morlet filter is the sum of a Gabor ...
fr_andres's user avatar
  • 170
7 votes
Accepted

Stability of system with poles inside unit circle - conflict with differential equation

What you are missing is that this is about a discrete-time system, because we're talking about poles and zeros in the complex $z$-plane and about poles inside or outside the unit circle. So there is ...
Matt L.'s user avatar
  • 90.5k
7 votes
Accepted

Fourier Transform with both Time Delay and Frequency Shift

If you are ever unsure, just go back to the definition and work out the Fourier Transform property for the specific situation: $$\begin{align*}\mathscr{F}\left\{x\left(t-t_0\right)e^{j2\pi f_0\left(t-...
Andy Walls's user avatar
  • 2,710
7 votes
Accepted

Wavelet Scattering properties & implementation?

Scattering overview provided in this answer. Computational structure Fig 4, Deep Scattering Spectrum In steps: (First order begins) $x$ convolves with $\psi1_i$ --> $W1_i$ Modulus, $W1_i \...
OverLordGoldDragon's user avatar
7 votes
Accepted

Why does a signal with constant frequency have spots that changes colors at a specific value of scale (and so frequency) in the scalogram?

Re: real part There are oscillations because that's what the wavelet transform is - a decomposition into zero-mean, localized oscillations. CWT is convolution (rather, cross-correlation) of signal ...
OverLordGoldDragon's user avatar
6 votes
Accepted

Motivation of time-frequency analysis

What a tricky question to overlook. Indeed I'm one of those who would immedieately press that Fourier transforms do lose time localization of the events as the comments stated. Yet it's certainly (...
Fat32's user avatar
  • 28.3k
6 votes

Motivation of time-frequency analysis

Fourier transforms generally yield complex spectral data. Under some technical conditions, they are bijections. From the Fourier transform, you can uniquely recover one single signal. However, when ...
Laurent Duval's user avatar
6 votes
Accepted

Uncertainty principle - Duration bandwidth principle

An important theorem, known as Weyl's, 1931, is: if function $s(t)$ and related functions $ts(t)$, $s'(t)$ are in $L^2$ (square integrable) with the related $\|\cdot\|$ $L_2$ norm symbol then: $$ \|...
Laurent Duval's user avatar
6 votes
Accepted

How to interpret the effect of different windows in short time fourier transform?

A window $w[n]$ truncates and weights (tapers) an input signal $x[n]$, to produce $v[n] = x[n]. w[n]$., for subsequent spectral analysis of $x[n]$. A windows's effect on the input signal's true ...
Fat32's user avatar
  • 28.3k
6 votes

Comparison of Linear Convolution and N Point DFT

The result {4,1,2,3} is the circular convolution of {1,2,3,4} and {0,1,0,0} which you correctly get by taking the inverse DFT of the product of the DFTs of the two sequences. We can check this by ...
Dan Boschen's user avatar
  • 52.3k
5 votes

Motivation of time-frequency analysis

An FT (being invertible) does preserve all transient event time information, however this time locality information is usually preserved by distributed it as varying changes to the entire phase ...
hotpaw2's user avatar
  • 35.4k
5 votes
Accepted

Infinite extent of spectrum, but also in time in Oppenheim's Discrete Time Signal Processing?

Not at all. The Uncertainty Principle says that a function cannot be both limited in time and limited in frequency. More specifically, the product of the signal's widths in time and in frequency (i.e.,...
Matt L.'s user avatar
  • 90.5k
5 votes

Stability of system with poles inside unit circle - conflict with differential equation

You're conflating the discrete-time definition of a system with the continuous-time representation of a system. Your discrete-time $$Y(z)\cdot\big(z-\frac{1}{2}\big)=X(z)\cdot z$$ does not ...
Peter K.'s user avatar
  • 25.9k
5 votes
Accepted

Wavelet "center frequency" explanation? Relation to CWT scales?

The "frequency" in "center frequency" does refer to Fourier frequency - rate of sinusoidal oscillation - but measures (and interpretations) can vary depending on the wavelet. Said ...
OverLordGoldDragon's user avatar
5 votes
Accepted

Impulse response of Time Varying Channel

In the context of wireless communications, the channel impulse response (CIR) is often estimated indirectly via the time-varying transfer function (TVTF) $H(t, f)$, defined by: $$ H(t, f) = \mathcal ...
Robert L.'s user avatar
  • 2,222
5 votes

A case that zero padding increase real resolution and extract more info than naive DFT?

What zero padding cannot do is help you resolve frequencies that cannot be resolved without it. Without zero-padding, the resolution is 1 Hz and the DFT can in fact resolve the tones at 440 and 441 Hz,...
MBaz's user avatar
  • 15.3k
5 votes
Accepted

"Instantaneous impulse response" in a linear time-varying system

The problem is that there are two common definitions of the impulse response of an LTV system, resulting in the following input-output relations: $$ y(t)=\int_\tau h_1(t,\tau)x(\tau)d\tau\tag{1}$$ and ...
Matt L.'s user avatar
  • 90.5k
5 votes

Amplitude extraction using STFT

Firstly, STFT is fundamentally a time-frequency transform: convolutions with windowed complex sinusoids (i.e. bandpass filtering). You aren't going to "frequency", and "windowed Fourier ...
OverLordGoldDragon's user avatar
4 votes

How to Map CWT to Synchrosqueezed wavelet transform?

Let me explain the intuition briefly. The authors of the paper you've cited assume that the signal $x(t)$ can be written in the form \begin{align*} x(t) &= \sum_{k=1}^K a_k(t) \exp(2\pi\mathrm{i} ...
Alex's user avatar
  • 41

Only top scored, non community-wiki answers of a minimum length are eligible