Skip to main content

What is the Step Size in FFT? (with MATLAB)

 

In FFT (Fast Fourier Transform), the step size refers to the spacing between consecutive points in the output data after performing the transform. It's often determined by the sampling rate of the signal. The step size is crucial for accurate frequency representation, and smaller step sizes provide finer frequency resolution in the resulting frequency domain representation.


Step Size of a Signal in the Time Domain (in general)

Suppose you have a signal sampled at 1000 Hz (sampling rate) for a duration of 1 second. The step size, or the time difference between consecutive samples, is then given by the inverse of the sampling rate:

Step size = 1 / Sampling rate = 1 / 1000 Hz = 0.001 seconds

 

General Frequency Resolution:

Sampling frequency fs​=1000Hz

Duration T= 1 second

Number of samples N=fs⋅T=1000⋅1=1000

 Î”f=1 / T

Δf=1 / 1 second = 1 Hz

 

Frequency Domain Step Size in FFT

Step Size in the Frequency Domain

The step size in the frequency domain refers to the spacing between adjacent frequency bins in the FFT output. It is determined by the signal's sampling rate and the size of the FFT:

Δf = fs / N

Where:

  • Δf: Frequency step size (frequency resolution).
  • fs: Sampling rate (Hz).
  • N: FFT size (number of bins).

Total Bandwidth

The total bandwidth covered by the FFT is determined by the sampling rate and the Nyquist theorem:

Total Bandwidth = fs / 2

Frequencies above the Nyquist frequency (fs/2) cannot be represented due to aliasing.

Frequency Step Size after FFT

Combining the above, the frequency step size (bin width) in the FFT output is:

Δf = fs / (2N)

 

Suppose:

  • Sampling frequency: Hz

  • FFT size:

  • Then:

    Δf=1000512≈1.953 Hz

So, your FFT bins are spaced about 1.953 Hz apart.


Key Observations:

  • Smaller Δf results in higher frequency resolution.
  • To achieve smaller Δf, increase the FFT size (N) or the signal's duration (T).
  • Total bandwidth is inversely proportional to the number of bins (N).

Time Domain Step Size in FFT

 Time step (seconds) = ​Hop size / fs (samples)
 
Suppose:

    Sampling frequency fs ​= 1000 Hz

    FFT window length = 512 samples

    Hop size = 256 samples (i.e., 50% overlap)

Then:

    Each FFT is calculated on a 512-sample window

    The window shifts forward by 256 samples

    Time step size = 256 / 1000​ = 0.256 seconds

So, a new FFT is computed every 0.256 seconds of the signal. ​
 

MATLAB Code

% The code is developed by SalimWireless.Com


clc;
clear all;
close all;


% Parameters
fs = 1000; % Sampling frequency (Hz)
T = 1; % Duration (seconds)
N1 = 256; % FFT size for coarse resolution
N2 = 1024; % FFT size for fine resolution
t = 0:1/fs:T-1/fs; % Time vector


% Signal with multiple frequency components
f1 = 50; % Frequency 1 (Hz)
f2 = 60; % Frequency 2 (Hz)
f3 = 200; % Frequency 3 (Hz)
signal = sin(2*pi*f1*t) + sin(2*pi*f2*t) + sin(2*pi*f3*t);


% FFT with coarse resolution (N1)
fft_coarse = fft(signal, N1);
frequencies_coarse = (0:N1-1)*(fs/N1); % Frequency vector
magnitude_coarse = abs(fft_coarse);


% FFT with fine resolution (N2)
fft_fine = fft(signal, N2);
frequencies_fine = (0:N2-1)*(fs/N2); % Frequency vector
magnitude_fine = abs(fft_fine);


% Plotting
figure;


% Coarse Resolution Plot
subplot(2, 1, 1);
plot(frequencies_coarse(1:N1/2), magnitude_coarse(1:N1/2));
title('FFT with Coarse Resolution (N = 256) where step size is 3.906');
xlabel('Frequency (Hz)');
ylabel('Magnitude');
grid on;


% Fine Resolution Plot
subplot(2, 1, 2);
plot(frequencies_fine(1:N2/2), magnitude_fine(1:N2/2));
title('FFT with Fine Resolution (N = 1024) where step size is 0.977');
xlabel('Frequency (Hz)');
ylabel('Magnitude');
grid on;

Output






Copy the MATLAB Code above from here

Return to Fourier Transform Main Page →

Further Reading

📊 MATLAB & ONLINE SIMULATIONS (Web-Based)
🎓 GATE & UGC - NET SOLUTIONS
📡 WIRELESS & SIGNAL PROCESSING


Contact Us

Name

Email *

Message *

Popular Posts

Hybrid Beamforming | Page 1

Beamforming Techniques Hybrid Beamforming... Page 1 | Page 2 | Hybrid Beamforming: Hybrid beam formation was developed to address some of the limitations of digital pre-coding approaches. Every antenna element is connected to an RF chain in digital pre-coding (beam forming) method. We also know that each RF chain is in charge of providing a separate data stream between the transmitter and the receiver. We know that a larger number of independent data streams leads to higher data rates. It has a spatial multiplexing feature for MIMO. As a result, we may assume that switching from MIMO to massive MIMO will benefit us more in terms of spatial multiplexing in massive MIMO, where each antenna is coupled to a single RF chain. We'll proceed with a definition of hybrid beam forming. Overview of hybrid beam forming with example: Unlike digital beam forming, more than one antenna element is connected to a single RF chain in hybr...
Read more

UGC NET Electronic Science Previous Year Question Papers with Solutions

Home / Engineering & Other Exams / UGC NET 2026 PYQ ⬇️ Download Papers and Solutions 📋 Exam Pattern 💡 Preparation Tips ❓ FAQs 📊 Exam Highlights: Electronic Science (88) Feature Details Junior Research Fellowship (JRF) ₹37,000 + HRA per month Eligibility M.Sc/M.Tech in Electronics (55%) Validity of Certificate JRF (3 Years) | Lectureship (Lifetime) 📥 Download UGC NET Electronics PDFs Complete collection of previous year question papers, answer keys and explanations for Subject Code 88. Start Downloading 📂 View All Question Papers June 2025 - Question Paper Download PDF June 2025 - Solved Paper + Explanation ...
Read more

MATLAB Code for Rms Delay Spread

RMS delay spread is crucial when you need to know how much the signal is dispersed in time due to multipath propagation, the spread (variance) around the average. In high-data-rate systems like LTE, 5G, or Wi-Fi, even small time dispersions can cause ISI. RMS delay spread is directly related to the amount of ISI in such systems. RMS Delay Spread [↗] Delay Spread Calculator Enter delays (ns) separated by commas: Enter powers (dB) separated by commas: Calculate   The above calculator Converts Power to Linear Scale: It correctly converts the power values from decibels (dB) to a linear scale. Calculates Mean Delay: It accurately computes the mean excess delay, which is the first moment of the power delay profile. Calculates RMS Delay Spread: It correctly calculates the RMS delay spread, defined as the square root of the second central moment of the power delay profile.   MATLAB Code  clc...
Read more

Comparisons among ASK, PSK, and FSK (with MATLAB + Simulator)

Modulation ASK, FSK & PSK Constellation MATLAB Simulink MATLAB Code Comparisons among ASK, PSK, and FSK 📘 Comparisons among ASK, FSK, and PSK 🧮 Online Simulator Bandwidth 🧮 MATLAB Code BER Analysis 📚 Further Reading 📂 View Other Topics on Comparisons among ASK, PSK, and FSK ... 🧮 Comparisons of Noise Sensitivity, Bandwidth, Complexity, etc. 🧮 MATLAB Code for Constellation Diagrams of ASK, FSK, and PSK 🧮 Online Simulator for ASK, FSK, and PSK Generation 🧮 Online Simulator for ASK, FSK, and PSK Constellation 🧮 Some Questions and Answers Comparisons among ASK, PSK, and FSK Comparison among ASK, FSK, and PSK Parameters ASK FSK PSK Variable Characteristics Amplitude ...
Read more

How Windowing Affects Your Periodogram

The windowed periodogram is a widely used technique for estimating the Power Spectral Density (PSD) of a signal. It enhances the classical periodogram by mitigating spectral leakage through the application of a windowing function. This technique is essential in signal processing for accurate frequency-domain analysis.   Power Spectral Density (PSD) The PSD characterizes how the power of a signal is distributed across different frequency components. For a discrete-time signal, the PSD is defined as the Fourier Transform of the signal’s autocorrelation function: S x (f) = FT{R x (Ï„)} Here, R x (Ï„)}is the autocorrelation function. FT : Fourier Transform   Classical Periodogram The periodogram is a non-parametric PSD estimation method based on the Discrete Fourier Transform (DFT): P x (f) = \(\frac{1}{N}\) X(f) 2 Here: X(f): DFT of the signal x(n) N: Signal length However, the classical periodogram suffers from spectral leakage due to abrupt truncation of the ...
Read more

What is - 3dB Frequency Response? And applications

📘 Overview & Theory 📘 Application of -3dB Frequency Response 🧮 MATLAB Codes 🧮 Online Digital Filter Simulator 📚 Further Reading Filters What is -3dB Frequency Response?   Remember, for most passband filters, the magnitude response typically remains close to the peak value within the passband, varying by no more than 3 dB. This is a standard characteristic in filter design. The term '-3dB frequency response' indicates that power has decreased to 50% of its maximum or that signal voltage has reduced to 0.707 of its peak value. Specifically, The -3dB comes from either 10 Log (0.5) {in the case of power} or 20 Log (0.707) {in the case of amplitude} . Viewing the signal in the frequency domain is helpful. In electronic amplifiers, the -3 dB limit is commonly used to define the passband. It shows whether the signal remains approximately flat across the passband. For example, in pulse shapi...
Read more

Phase Shift Keying (PSK) Modulation & Demodulation (with Simulation)

Phase Shift Keying (PSK) Theoretical Framework: Phase Shift Keying (PSK) is a digital modulation paradigm where information is encoded by modulating the instantaneous phase of a continuous-wave carrier. In Binary PSK (BPSK), the most fundamental implementation, the bitstream is mapped onto two discrete phases separated by $\pi$ radians—typically representing binary '1' and '0'—resulting in an antipodal signaling scheme where each state change induces a phase reversal in the carrier. BPSK Transmitter Synthesis: The analytic representation of a BPSK signal $s(t)$ is defined as: \[ s(t) = \begin{cases} A_c \cos(2\pi f_c t), & \text{for symbol } m_1 \\ A_c \cos(2\pi f_c t + \pi) = -A_c \cos(2\pi f_c t), & \text{for symbol } m_0 \end{cases} \] NRZ-L m(t) × ...
Read more