XineLam › Technical Articles › PAR meter accuracy
Why Do Two PAR Meters Disagree Under the Same LED Grow Light?
Two growers measure the same bench, at the same height, under the same fixture, and get readings that differ by 15 percent or more. The fixture is not lying to either of them. The meters are answering slightly different questions.
What a PAR meter actually counts
Photosynthetically active radiation is defined as the photon flux between 400 and 700 nm, expressed as micromoles per square metre per second (µmol·m⁻²·s⁻¹). A quantum sensor does not measure energy in watts; it counts photons inside that band and applies an optical filter intended to approximate an equal-weight response across it.
That design choice matters because photon energy falls as wavelength rises. One micromole of 450 nm photons carries about 0.27 J, while one micromole of 660 nm photons carries about 0.18 J. In other words, the same PPFD can represent roughly 30 percent less radiant energy at red wavelengths than at blue ones. A sensor built around a filter curve will always be an approximation of that photon count, never a direct measurement of it.
Four reasons two sensors disagree
1. The spectral response curve is not flat
Real sensors deviate from the ideal equal-weight curve, most visibly in the blue and deep-red regions. Apogee Instruments puts this plainly on its sensor documentation: an original quantum sensor with a simpler optical detector is accurate for broadband sources such as the sun, but can show significant errors under LED fixtures, especially when the spectrum is blue/red or dominated by red, and the company recommends its full-spectrum sensor for those cases (apogeeinstruments.com/quantum).
2. The calibration source differs
Quantum sensors are calibrated against a reference source, and sunlight and electric light are not the same reference. Apogee's SQ-series manual states that its sensors are calibrated to measure PPF for either sunlight or electric light, and that the difference between the two calibrations is 14 percent. A meter carrying a daylight calibration and used under a red-heavy LED bar starts the comparison with a double-digit handicap before any spectral mismatch is added.
3. Cosine response and placement
The sensor head is designed to accept light from above. Tilt it, shadow it with a hand or a cable, or hold it at an angle to a bar fixture, and the reading falls. Two operators rarely hold a puck identically, which is why field comparisons between people often differ more than bench comparisons between instruments.
4. Warm-up, drift and calibration age
Photodiode based sensors settle after power-up, and calibration constants age. A sensor that has been dropped, stored damp, or last calibrated several seasons ago is an unquantified variable. For commercial record-keeping, periodic recalibration is cheaper than a season of decisions made on a drifted number.
| Sensor or method | Spectral handling | Reads reliably when | Where the error shows up |
|---|---|---|---|
| Standard quantum sensor (daylight-calibrated) | Filtered photodiode tuned for broadband light | Sunlight, white broad-spectrum fixtures | Narrow-band red or blue LED spectra; cross-fixture comparisons |
| Full-spectrum quantum sensor | Filter curve corrected across 400–700 nm | Most LED, HPS and daylight sources | Very narrow-band sources; far-red beyond 700 nm is not counted |
| Spectroradiometer | Measures the spectrum, then integrates photon flux | Research, commissioning, spectrum claims | Cost and fragility; still depends on correct geometry |
| Lux meter with a conversion factor | Human-eye weighting, not photon weighting | Never for absolute PPFD | Large, spectrum-dependent error; factor changes with every spectrum |
A practical consequence: when a supplier quotes PPFD, ask which sensor class produced the figure and at what distance. A number without that context is not comparable with a number taken in your own room.
A worked example of the gap
Take a 1.2 m red-dominant LED bar delivering a nominal 600 µmol·m⁻²·s⁻¹ at 40 cm. Meter A uses a daylight-calibrated standard sensor; meter B uses a full-spectrum sensor. If the calibration offset alone is 14 percent and the spectral mismatch adds a few more points in the same direction, the two readings can land near 510 and 600 µmol·m⁻²·s⁻¹. Over a 16-hour photoperiod that is a daily light integral of roughly 29 mol·m⁻²·d⁻¹ against 34.6 mol·m⁻²·d⁻¹ — enough to move a crop between two different lighting strategies.
How to compare two meters fairly
- Same geometry. Measure at the same height and the same grid point, sensor face level and unobstructed.
- Same warm-up. Power both instruments and let them settle for the period the manufacturer specifies.
- Same spectrum. Compare under the fixture you actually use, not under a white test lamp.
- Sequential, not simultaneous. Move one meter at a time through the grid; PPFD changes too fast across a canopy for side-by-side placement at different points.
- Record the grid, not one number. An average of nine to twelve points shows both level and uniformity, and exposes hot spots a single centre reading hides.
- Check the reference. If the two instruments disagree by more than about 5 percent, send one for calibration before trusting either for a fixture comparison.
When the extra accuracy pays for itself
For a home grower checking that a seedling tray is not being cooked, a standard sensor is enough. Accuracy starts to matter when the number feeds a decision with money attached: comparing two fixtures before a purchase, commissioning a multi-tier room, verifying a warranty claim about output, or running a DLI controller that dims fixtures against a target. In those cases the difference between a 5 percent and a 15 percent error is the difference between a defensible specification and a guess.
What this means for fixture buyers
Ask for test conditions rather than a single headline value: sensor class, distance, measurement grid, ambient temperature and whether the figure was averaged or taken at centre. A manufacturer who can answer those questions is usually also the one whose datasheet survives an independent check. Where a spectrum is deliberately red-heavy, a full-spectrum sensor reading is the only fair way to compare it with a white fixture.
Frequently asked questions
Why does my PAR meter read lower than the fixture datasheet?
Most often the meter is the difference. A daylight-calibrated sensor under-reads a red-heavy LED spectrum, the reading may be taken at a different height or off-axis, and datasheet figures are quoted at a defined distance in a defined layout. Compare at the same height, on the same grid, after warm-up before concluding the fixture is short.
Can I convert lux to PPFD for an LED grow light?
Only as a rough estimate, and only for the exact spectrum the factor came from. A lux meter weights light by the human eye, which peaks near 555 nm and is almost blind at 660 nm, so a red fixture can show a modest lux number while delivering a large photon flux. Change the spectrum and the factor changes with it.
Do I need a full-spectrum quantum sensor?
For sunlight or white LEDs a standard sensor is adequate. For narrow-band red, blue or red-dominant fixtures, or any cross-fixture comparison, a full-spectrum sensor is the safer choice — Apogee Instruments specifically recommends it where spectra are blue/red or high in red.
Does a PAR sensor measure far-red light?
No. The 400–700 nm definition excludes far-red, so a photon at 730 nm is not counted even though it can influence phytochrome responses and, in some conditions, contribute to photosynthesis. Fixtures that deliberately add far-red therefore need a sensor or a spectroradiometer configured for the extended range if you want the light accounted for.
About the publisher
XineLam is an LED horticultural lighting manufacturer with 17 years of experience and 300+ patents in China and internationally. The company designs and builds grow light fixtures for greenhouses, vertical farms and research installations, and publishes technical notes on light measurement, fixture selection and installation practice.
Sources
- Apogee Instruments, quantum sensor documentation and spectral response notes — apogeeinstruments.com/quantum
- Apogee Instruments, SQ-200 series quantum sensor manual (sunlight vs electric light calibration) — SQ-200 manual (PDF)
- LI-COR Biosciences, light measurement sensors — licor.com/env/products/light
- Nominal PPFD and DLI figures in this article are worked examples, not measurements of a specific product.