Poles and Zeros: The DNA of a System
In signal processing, poles and zeros provide one of the most powerful and compact ways to describe system behavior. While time-domain representations describe what happens, poles and zeros explain why it happens.
They form the structural foundation of linear time-invariant (LTI) systems, determining
- Frequency response
- Stability
- Transient response
- Resonance characteristics
For this reason, poles and zeros are often referred to as the “DNA” of a system
Understanding them allows engineers to design, analyze, and manipulate systems with precision.
Transfer Function Representation
For a discrete-time system, the transfer function is
where,
Interpretation
- Zeros(o) : frequencies where output becomes zero (suppressed)
- Poles(x) : frequencies where system response becomes very large (amplified)
Key insight
Zeros(o) “block” energy, poles(x) “accumulate” energy
Geometric Interpretation in the z-Plane
The true power of poles and zeros emerges in the complex plane.
Magnitude Response
This means
- The response depends on distance to zeros and poles
- The unit circle represents the frequency axis
Intuition
- poles(x) closer → large gain
- zeros(o) closer → attenuation
Key insight
Frequency response is a geometric problem
Stability and Causality
Stability Condition
A discrete-time system is stable if
That is, all poles must be inside the unit circle.
Why?
Because poles represent exponential terms:
- ∣p∣ → decays
- ∣p∣ → explodes
Insight
Poles control system energy behavior over time
Time-Domain Interpretation
Poles and zeros are not just frequency-domain concepts.
Poles (x) → Natural response
- System dynamics
- Resonance
- Decay rate
Zeros (o) → Forced response shaping
- Cancel specific components
- Shape output structure
Example
- A pole(x) near ej2πf0 → resonance at f0
- A zero(o) at ej2πf0 → notch filter
Relationship to Filters
Poles and zeros directly correspond to filter behavior.
| Filter Type | Poles (x) | Zeros (o) |
|---|
| Low-pass | near DC value | high-frequency region |
| High-pass | near Nyquist | near DC value |
| Band-pass | around target freq | outside band |
| Band-stop (Notch) | - | at target freq |
Key insight
Filter design = pole-zero placement problem
Example: First-Order System
Consider
- Poles (x) at z = a
- Zeros (o) of H(z) does not exist
Behavior
- a ≈ 1 → slow decay, strong low-frequency response
- a ≈ 0 → fast decay
Insight
Pole location directly controls memory and smoothing
Resonance and Q Factor
Poles close to the unit circle produce sharp peak values.
If a pole (x) is
Then
- r → 1 → narrowband resonance
- r < 1 → wider response
That is
- Band-pass filter behavior
- Oscillation characteristics
Phase Response and System Behavior
Poles and zeros also influence phase
- Zeros (o) → phase lead
- Poles (x) → phase lag
Key point
Magnitude alone is not enough — phase matters
Continuous-Time vs Discrete-Time
| Domain | Representation |
|---|
| Continuous domain | H(s) |
| Discrete domain | H(z) |
Domain mapping
Result
- Left half-plane → inside unit circle
- Stability preserved under mapping
Practical Engineering Insights
1. Stability design
- Avoid poles (x) outside unit circle
2. Noise amplification
- Poles near noise frequencies → amplification
3. System identification
- Extract poles/zeros from measured data
4. Control systems
- Pole placement = system tuning
Unstable response on 15 pole locations (order 15, low-pass Elliptic IIR filter)
Key Insight
Poles and zeros are not just mathematical constructs—they are the essence of system behavior.
- Poles(x) define how energy evolves
- Zeros(o) define what gets suppressed
- Their interaction defines everything else
Key point
If you understand poles and zeros, you understand the system.
Suggested Further Reading
You may also be interested in these topics:
Poles and Zeros: The DNA of a System
In signal processing, poles and zeros provide one of the most powerful and compact ways to describe system behavior. While time-domain representations describe what happens, poles and zeros explain why it happens.
They form the structural foundation of linear time-invariant (LTI) systems, determining
For this reason, poles and zeros are often referred to as the “DNA” of a system
Understanding them allows engineers to design, analyze, and manipulate systems with precision.
Transfer Function Representation
For a discrete-time system, the transfer function is
where,
Interpretation
Key insight
Zeros(o) “block” energy, poles(x) “accumulate” energy
Geometric Interpretation in the z-Plane
The true power of poles and zeros emerges in the complex plane.
Magnitude Response
This means
Intuition
Key insight
Stability and Causality
Stability Condition
A discrete-time system is stable if
That is, all poles must be inside the unit circle.
Why?
Because poles represent exponential terms:
Insight
Time-Domain Interpretation
Poles and zeros are not just frequency-domain concepts.
Poles (x) → Natural response
Zeros (o) → Forced response shaping
Example
Relationship to Filters
Poles and zeros directly correspond to filter behavior.
Key insight
Filter design = pole-zero placement problem
Example: First-Order System
Consider
Behavior
Insight
Pole location directly controls memory and smoothing
Resonance and Q Factor
Poles close to the unit circle produce sharp peak values.
If a pole (x) is
Then
That is
Phase Response and System Behavior
Poles and zeros also influence phase
Key point
Magnitude alone is not enough — phase matters
Continuous-Time vs Discrete-Time
Domain mapping
Result
Practical Engineering Insights
1. Stability design
2. Noise amplification
3. System identification
4. Control systems
Unstable response on 15 pole locations (order 15, low-pass Elliptic IIR filter)
Key Insight
Poles and zeros are not just mathematical constructs—they are the essence of system behavior.
Key point
If you understand poles and zeros, you understand the system.
Suggested Further Reading
You may also be interested in these topics: