How Does a Pulse Oximeter Work? The Two-Light Measurement Explained

A pulse oximeter is one of the most familiar pieces of monitoring equipment in any clinical environment, yet the physics behind the number on the screen is often taken for granted. For biomedical engineers specifying replacement sensors, educators teaching waveform interpretation, and distributors supporting end users, a solid grasp of the measurement principle makes it easier to troubleshoot readings, explain limitations at the bedside, and choose the right sensor format for a given clinical setting. This article walks through the light-absorption method step by step, from the two wavelengths used to the averaging that explains why the displayed value lags behind rapid physiological change.

The core idea in one paragraph

At its simplest, a pulse oximeter shines two wavelengths of light through a perfused section of tissue, detects how much light arrives at the other side, and compares what was emitted with what was received. Blood absorbs light, and the amount absorbed depends on which wavelengths are used and whether the haemoglobin in the blood is carrying oxygen. By measuring absorption at two carefully chosen wavelengths and separating the changing (pulsatile) part of the signal from the steady part, the device estimates the proportion of haemoglobin that is oxygenated. Everything else in a pulse oximeter design — the sensor optics, the signal processing, the calibration — exists to make that basic two-light measurement accurate and robust in real clinical conditions.

Oxyhaemoglobin vs deoxyhaemoglobin absorption at red (~660 nm) and infrared (~940 nm) wavelengths

The choice of wavelengths is the foundation of the entire method. Haemoglobin changes its optical behaviour when it binds oxygen, and the two states can be distinguished because their absorption spectra differ across the red and infrared regions.

• Red light (~660 nm): Deoxyhaemoglobin absorbs more red light than oxyhaemoglobin does. Oxygenated arterial blood therefore appears brighter red and lets more of this wavelength through, while desaturated blood absorbs more of it.
• Infrared light (~940 nm): The relationship reverses. Oxyhaemoglobin absorbs more infrared light than deoxyhaemoglobin, so a more oxygenated sample attenuates this wavelength to a greater degree.

Because the two states pull in opposite directions at these two wavelengths, measuring absorption at both gives the device two complementary pieces of information. A change that increases oxygen saturation will reduce red-light absorption and increase infrared absorption; a change in the opposite direction produces the mirror-image pattern. This differential behaviour is what allows a single sensor to estimate saturation without needing a separate reference measurement of total haemoglobin concentration or path length. It is also why the accuracy of the emitting components and the spectral characteristics of the sensor matter — an accessory must emit at the intended wavelengths for the monitor's algorithm to interpret the returning light correctly. When specifying compatible SpO2 sensors, matching the sensor's optical characteristics to the monitor's expectations is part of ensuring reliable readings.

The pulsatile trick: separating arterial blood from tissue and venous blood

If a pulse oximeter simply measured total light absorption, the result would be dominated by everything in the light path that is not arterial blood: skin, bone, connective tissue, venous blood, and the non-pulsatile portion of arterial blood itself. The method solves this by exploiting the fact that arterial blood volume changes with each heartbeat.

The detected light signal can be decomposed into two components:

• DC component: The large, steady portion of the signal. It represents light absorbed by tissue, bone, venous blood, and the constant volume of arterial blood. It carries no information about oxygen saturation changes with each pulse.
• AC component: The small, rhythmic variation superimposed on the DC level. It corresponds to the surge of arterial blood with each cardiac cycle, which transiently increases the light path and therefore the absorption.

By subtracting the steady DC level and analysing only the pulsatile AC component, the device effectively isolates the light that has passed through freshly delivered arterial blood. This is the "pulsatile trick" — it turns a difficult absolute measurement into a relative one, and it is the reason a pulse oximeter requires adequate pulsatile flow to function. It also explains why the method is sometimes called "pulse" oximetry: without a detectable pulse, there is nothing to isolate. This isolation step is central to answering what is SpO2? in practical terms — the reading reflects the oxygen saturation of the pulsatile arterial compartment, not of venous blood or tissue as a whole.

The ratio of ratios and the empirical calibration curve

Once the AC and DC components are known for both wavelengths, the device computes a normalized measure of pulsatile absorption at each wavelength — the ratio of the AC to the DC signal. It then takes the ratio of these two normalized values, commonly referred to as the "ratio of ratios." This single number captures how the pulsatile absorption differs between red and infrared light.

What the device does next is important to understand: it does not calculate oxygen saturation from first principles. Instead, it applies an empirical calibration curve that maps the ratio of ratios to arterial oxygen saturation. This curve is established during validation testing in which the ratio measurements are compared against arterial blood samples analysed by a co-oximeter — the reference method for measuring saturation. The relationship between the ratio of ratios and true saturation is not perfectly linear, so the calibration curve is typically a polynomial or look-up relationship derived from that validation work. Human and animal volunteer studies under controlled desaturation protocols are the standard approach for building these curves.

Two practical consequences follow. First, the accuracy of any pulse oximeter depends on the calibration curve being appropriate for the population it is used on and for the sensor type it was validated with. Second, because the curve is empirical, different device designs may handle borderline conditions differently. This is why clinicians are advised to treat the displayed number as an estimate with a known accuracy range rather than an exact laboratory value. For distributors and educators, it is also a reminder that sensor compatibility is not merely a mechanical or electrical question — the monitor's algorithm expects signals with particular optical and timing characteristics.

The plethysmograph waveform and perfusion indication

Most modern pulse oximeters display more than a number: they show a plethysmographic waveform, often called the pleth or plethysmograph. This waveform is a real-time representation of the pulsatile component of the light signal, essentially a plot of the AC amplitude over time. It rises with each arterial pulse as blood volume in the tissue increases and falls as it drains.

The pleth waveform serves several purposes in clinical practice:

• Perfusion indication: A well-formed, stable waveform suggests adequate pulsatile blood flow at the sensor site. A flattened or irregular waveform may indicate poor perfusion, a poorly positioned sensor, or motion artifact.
• Signal quality assessment: Educators frequently teach waveform interpretation as a way to judge whether the displayed saturation value can be trusted in that moment.
• Timing information: The pleth can provide a visual cue of the cardiac cycle, useful when correlating with other monitored parameters.

It is worth noting that the pleth waveform is not a calibrated measure of blood volume or flow; it is a qualitative indicator. A normal-looking waveform does not guarantee an accurate saturation reading, but a poor one strongly suggests the reading should be interpreted with caution and the sensor site reassessed.

Transmission vs reflectance sensors (finger/ear vs forehead)

Pulse oximeter sensors come in two optical configurations, and the choice between them depends on the clinical setting and the patient.

Transmission sensors

In a transmission sensor, the light emitters and the photodetector are positioned on opposite sides of the tissue. Light travels through the tissue from one side to the other. This configuration requires a relatively thin, well-perfused site, which is why finger clips and ear probes are common transmission designs. Finger clip and silicone soft tip sensors are familiar transmission formats across monitor and handheld devices.

Reflectance sensors

In a reflectance sensor, the emitters and detector are placed side by side on the same surface, and the device detects light that has scattered and reflected back through the tissue. Reflectance sensors are typically used on sites where transmission is impractical, such as the forehead. Wrap and forehead-style reflectance sensors exist across monitor and handheld devices. Site selection affects both signal quality and the practicalities of securing the sensor, and it can influence how quickly the reading responds to change.

For biomedical engineers, the key point is that the same measurement principle underlies both configurations, but the optical path and tissue characteristics differ. A sensor validated for one site should not be assumed to perform identically at another.

Why the measurement lags and averages

Clinicians sometimes notice that the displayed saturation value changes more slowly than the patient's actual condition, and there are good reasons for this. The pulse oximeter must extract a small pulsatile signal from a large background, and individual pulses contain noise and variability. To produce a stable, usable number, the device averages the ratio-of-ratios measurement over several seconds — often a window of a few seconds to around ten seconds, depending on the device and its settings.

This averaging has two effects. It smooths out beat-to-beat noise and motion artifact, improving readability. It also introduces a delay: a genuine change in arterial saturation will take some seconds to be fully reflected in the displayed value. In addition, there is a physiological component to the lag — the transit time for blood to travel from the lungs to the monitored site and for the tissue signal to respond. When teaching rapid-response scenarios, it is useful to remind users that the pulse oximeter is a trend and verification tool, not an instantaneous readout, and that clinical assessment should not rely on the number alone. Understanding the averaging window also helps explain why brief desaturation events may be attenuated on screen.

Where the method reaches its limits

No measurement principle is universal, and pulse oximetry has well-recognized boundary conditions that every user should understand.

• Very low saturation: The empirical calibration curve is typically built from validation data spanning a range of saturations, but accuracy tends to degrade at very low values. Below the validated range, the displayed number should be treated with particular caution.
• Dyshaemoglobins: The two-wavelength method assumes that the haemoglobin present is either oxygenated or deoxygenated. Abnormal haemoglobins such as carboxyhaemoglobin and methaemoglobin absorb light in ways that can mislead the ratio-based calculation, producing readings that do not reflect true oxygen saturation. Co-oximetry is required to detect these conditions.
• Motion: Patient movement introduces artifact that can corrupt the pulsatile signal, leading to unreliable readings or alarms. Signal processing helps, but motion remains a common source of error, particularly in active patients.
• Poor perfusion and signal quality: As noted, the method depends on a detectable arterial pulse. Low perfusion, sensor malposition, and ambient light interference can all degrade performance.

For manufacturers and distributors, these limits are not reasons to distrust the technology but reasons to support users in interpreting it correctly. Selecting a sensor format suited to the patient and site, verifying that compatible accessories match the monitor's requirements, and educating clinical staff on waveform assessment and averaging all contribute to safer monitoring. The pulse oximeter remains a workhorse of patient assessment precisely because its underlying principle — two wavelengths, a pulsatile signal, and an empirical calibration — is elegant, non-invasive, and well understood when its boundaries are respected.

Signal quality: why the waveform is part of the measurement

Because the calculation isolates a pulsatile component that may be a tiny fraction of the total light signal, anything that corrupts that component corrupts the result. This is why oximeters display the plethysmograph waveform and, on many devices, a perfusion or signal-quality indicator: a clean, regular pleth with a strong amplitude means the AC component the device is ratio-ing is genuinely arterial; a ragged or tiny waveform means the displayed number is built on noise and averaging. Clinically, the habit worth teaching is to read the waveform before the number — check that the pleth pulse rate matches the palpated or ECG rate, and distrust any SpO2 value whose waveform you would not trust. Sensor fit feeds straight into this: the right size and format from the sensor range for the patient in front of you is an accuracy decision, not a comfort one.