A wrist PPG/ECG-style sensor samples an electrical/optical signal from the body; motion artefacts add noise, and a filter smooths the trace to recover the true heart rate.
signal(t) = ECG(t; HR) + noise(t)·motion
filtered(t) = signal(t) * LowPass(k=filter)
SNR = P_signal / P_noise
- Heart rate — the true underlying cardiac rate driving the periodic waveform.
- Motion noise — artefact amplitude injected by wrist movement during sensing.
- Noise filter strength — how aggressively a digital low-pass filter smooths the raw signal.
- Sensor sensitivity — gain applied to the raw optical/electrical signal before filtering.
The 3D ribbon traces the filtered waveform in real time above the wrist sensor model; watch signal quality drop as motion noise rises, and recover as filter strength increases (at the cost of a slightly delayed trace).