Nyquist Sampling Theorem

Definition

The Nyquist sampling theorem states that a band-limited analog signal can be perfectly reconstructed from its samples if it is sampled at a rate at least twice its highest frequency component.

Mathematically, if the highest frequency in the signal is fmax, then the sampling frequency fs must satisfy:

fs ≥ 2 fmax

The quantity 2 fmax is called the Nyquist rate, and half the sampling frequency (fs / 2) is called the Nyquist frequency.

This theorem assumes:

• the signal is band-limited,
• sampling is ideal,
• and reconstruction is done correctly.

Main Content

1. Sampling Rate and Nyquist Rate

• The sampling rate is the number of samples taken per second from an analog signal, measured in Hz.
• The Nyquist rate is the minimum sampling rate required to avoid loss of information, and it equals twice the highest frequency present in the signal.

For example, if an audio signal contains frequencies up to 10 kHz, then the minimum sampling rate required is:

Nyquist rate = 2 × 10 kHz = 20 kHz

So, to represent that signal accurately, the sampling rate must be 20 kHz or higher.

Why this matters:

• A lower sampling rate cannot capture all the variations of the signal.
• A higher sampling rate gives more accurate representation, though it increases storage and processing requirements.

Example:

• Human hearing typically extends up to about 20 kHz.
• That is why standard audio CDs use 44.1 kHz sampling, which is safely above twice the highest audible frequency.

2. Aliasing and Signal Distortion

Aliasing

• occurs when a signal is sampled below the Nyquist rate.
• In aliasing, high-frequency components of the signal incorrectly appear as lower-frequency components after sampling.

This creates distortion and makes the recovered signal different from the original.

ASCII illustration of the idea:

Original signal:      /\/\    /\/\    /\/\
Samples too slow:     *    *    *    *
Reconstructed view:   /----\/----\/----\

When samples are too far apart:

• the true peaks and troughs are missed,
• the signal shape is misinterpreted,
• and different analog signals may produce the same sampled sequence.

Example:

• Suppose a signal contains a 900 Hz component.
• If it is sampled at 1000 Hz, the sampling rate is less than twice the frequency.
• The resulting digital samples may falsely appear as a much lower frequency signal.

Practical importance:

• Aliasing is one of the most common problems in digital systems.
• To reduce aliasing, engineers often use an anti-aliasing filter before sampling.

3. Reconstruction of the Original Signal

• If the sampling condition is satisfied, the original analog signal can be reconstructed exactly from its samples.
• Reconstruction is done using an ideal low-pass filter or mathematically through interpolation.

The key idea is that sufficient sampling preserves all the information required to rebuild the original waveform.

ASCII illustration of ideal reconstruction:

Analog signal:      ~~~~~~~~ ~~~~~~~~
Sample points:       *   *   *   *   *
Reconstructed:      ~~~~~~~~ ~~~~~~~~

How reconstruction works:

• The samples act like reference points.
• A reconstruction system fills in the values between the samples smoothly.
• If the sampling was above the Nyquist rate, the reconstructed signal matches the original.

Important note:

• Exact reconstruction is theoretical and assumes perfect sampling and perfect filters.
• In real systems, reconstruction may be very close to ideal but not always mathematically perfect.

Example:

• In digital audio, the sound waveform is sampled and later converted back using a digital-to-analog converter.
• If sampled properly, the listener hears the original sound faithfully.

Working / Process

1. Identify the highest frequency component of the analog signal

• First, determine the maximum frequency present in the signal.
• This can be from measurement, specifications, or signal analysis.
• Example: if a voice signal contains frequencies up to 3.4 kHz, that is the highest frequency to consider.

2. Choose a sampling frequency at least twice that value

• Apply the Nyquist condition: fs ≥ 2 fmax

• Add practical safety margin in real systems to avoid filter imperfections.

• Example: for a 3.4 kHz voice signal, the minimum theoretical sampling rate is 6.8 kHz, but telephone systems often use 8 kHz.

3. Sample, store, and reconstruct the signal

• Take measurements at equally spaced time intervals.
• Convert the samples into digital data.
• Later, reconstruct the signal using interpolation or an analog output stage.
• If the sampling rate was sufficient, the reconstructed signal remains faithful to the original.

Advantages / Applications

Accurate digital representation of analog signals

• It allows continuous signals to be converted into digital form with minimal information loss.
• This is fundamental for modern computing and signal processing.

Widely used in communication and multimedia systems

• Audio recording, video capture, telephony, radar, biomedical signals, and broadcasting all rely on proper sampling.
• Examples include microphones, cameras, ECG machines, and mobile communication systems.

Prevents distortion and supports reliable signal recovery

• By following the theorem, systems avoid aliasing and preserve signal quality.
• This improves the clarity, correctness, and reliability of digital processing.

Summary

• The Nyquist sampling theorem says a signal must be sampled at least twice its highest frequency for perfect recovery.
• Sampling below this limit causes aliasing and distortion.
• It is the basic rule behind converting analog signals into digital form.

Important terms to remember

• : sampling rate, Nyquist rate, Nyquist frequency, aliasing, reconstruction, band-limited signal