Formula for Frequency Resolution (in general)
The frequency resolution is the smallest frequency difference between two adjacent frequency points in your sampling range. It is determined by the total frequency range and the number of frequency samples N. The formula for the frequency resolution (or step size) Δf is:
Δf = (fmax - fmin) / (N - 1)
Where:
- fmin is the minimum frequency in the range (in this case, -50 Hz).
- fmax is the maximum frequency in the range (in this case, 50 Hz).
- N is the number of frequency points / frequency bins.
Using the Given Values:
From the function:
- fmin = -50 Hz
- fmax = 50 Hz
- N = 1000
The frequency resolution is:
Δf = (50 - (-50)) / (1000 - 1) = 100 / 999 ≈ 0.1001 Hz
Understanding Frequency Resolution in Signal Processing
Alternative Formula Using Time Duration
Another common way to define frequency resolution, especially in time-domain signal processing, is:
Δf = 1 / T
Where:
- T is the total time duration of the sampled signal.
FFT and Frequency Bins
When applying the FFT to a signal sampled at a rate fs, the spectrum is divided into N frequency bins:
Δf = fs / N
This version assumes you’re analyzing the signal in the range of 0 to fs (for one-sided spectrum) or -fs/2 to +fs/2 (for two-sided spectrum).
Why Frequency Resolution Matters
A fine frequency resolution enables better separation of closely spaced frequency components. It is critical in applications such as:
- Biomedical signal processing (e.g., ECG, EEG analysis)
- Spectrum sensing in wireless communications
- Speech recognition and audio filtering
- Fault detection in rotating machinery using vibration data
Common Misconception:
Many believe that increasing the number of FFT points alone increases frequency resolution. While it improves the frequency axis detail, true resolution is governed by signal duration or sampling span, not zero-padding.
Further Reading