Why M/P Ratio Is the Metric That Matters

The M/P ratio is the melanopic-to-photopic ratio: how much circadian-weighted light a source delivers for each unit of visual brightness. It is the single number that lets you compare a 2700 K warm white against a 5700 K daylight white at the same lux. It is also a ratio, not a measurement, and this article shows exactly where it helps and where it stops being useful.

Key Facts

1. SUMLINK platform M/P range: 0.45–1.06. Measured across the 2700 K–6500 K tuning range of the BY001 linear platform, following the melanopic DER definition in CIE S 026:2018. Source: BY001 catalogue page (labelled Design target — engineering values from the LED and driver platform specification, not yet confirmed by an independent laboratory report).
2. Melanopic EDI at 500 lx horizontal: 225–530. The same curve expressed as melanopic equivalent daylight illuminance (m-EDI, D65). Source: BY001 catalogue page (Design target).
3. WELL v2 Feature L03, daytime tier 1: ≥150 EML [136 melanopic EDI]. WELL writes both units; 150 × 1/1.103 = 136. Source: WELL v2 Light concept, Feature L03.
4. CIE S 026:2018 is the calculation basis, not the threshold source. The standard defines the melanopic action spectrum, the α-opic metrics and the reference conditions under which they are reported. It does not set an indoor exposure target. Thresholds come from WELL, from research recommendations, or from your own brief. Source: CIE S 026:2018 (metric standard; full text not read).
5. EML ↔ m-EDI conversion: ×1.103. EML = m-EDI(D65) × 1.103, so 136 m-EDI = 150 EML. This is the only arithmetic you need to move between a WELL clause and an α-opic calculation. Source: definitional, derived from the two metrics' normalisation.

Terms and units (six definitions, one table)

Source: CIE S 026:2018 (metric definitions). Unit names verified against the WELL v2 Feature L03 text; the standard's own full text was not read.

The M/P ratio is a division, and that is the whole idea

Take a white light source, measure its spectral power distribution, weight that SPD twice — once with the photopic luminosity function V(λ), once with the melanopic action spectrum — and divide the second result by the first. The number you get is the M/P ratio, which CIE S 026:2018 calls the melanopic daylight efficacy ratio (melanopic DER).

Because it is a ratio, M/P has no units and no illuminance in it. A 2700 K source at 200 lx and the same 2700 K source at 800 lx have the same M/P ratio. That is the point: M/P describes spectrum, while lux describes quantity. Together they give the absolute melanopic dose — which is the number that actually matters.

The reaction `m-EDI ≈ lux × M/P` is the working relation. At 500 lx, a source with M/P 0.45 delivers about 225 melanopic EDI; a source with M/P 0.92 delivers about 460. Same brightness, twice the circadian signal. This is why a spec sheet that lists only lux, CCT and CRI cannot tell you whether a light will do anything circadian at all.

Source: definitional (CIE S 026:2018 metric definitions); the arithmetic relation m-EDI = lux × M/P is internal (SUMLINK internal estimate).

The curve across a tuning range: 2700 K to 6500 K

M/P is not constant across a tunable-white fixture's range, and it is not linear either. Warm white has very little short-wavelength energy, so it carries little circadian signal; as you shift toward daylight white, melanopic content climbs faster than photopic content.

Source: BY001 catalogue page, melanopic DER table (all values labelled Design target). Internal consistency check: 500 lx × M/P reproduces every m-EDI figure exactly.

Two things to notice. First, read the two WELL tiers separately, because they cross at very different points. Tier 1 (150 EML = 136 m-EDI) is cleared by every row in the table at 500 lx — even the warmest, 2700 K at 225 m-EDI, sits well above it. Tier 2 (275 EML = 250 m-EDI) is the demanding one: the 3000 K row at 255 m-EDI is the first step that clears it, with the 3500 K row at 290 m-EDI the first to clear it with real margin. Second, the ratio between the warmest and coolest setting is 2.4× in M/P (1.06 / 0.45). A tunable-white system that can only make a small spectral shift will never move the circadian needle very far, no matter how wide its CCT marketing claim is.

One source defect worth flagging: the live BY001 page annotates its 3000 K row "below the WELL daytime threshold" and its 3500 K row as the first step that clears it. Neither annotation is consistent with either published WELL tier in its stated unit — 3000 K clears tier 2 (255 > 250 m-EDI), and every row clears tier 1. The annotations are only self-consistent if the threshold is read as 275 m-EDI rather than 275 EML — the unit mix-up this article exists to correct. The table above resolves it the defensible way; the catalogue page should be amended to match.

Three time bands, three different thresholds — and who set each one

Most confusion around M/P comes from mixing up three separate questions: what the target is in the morning, during the day, and in the evening. They come from different sources, and only one of them is a certification threshold.

Note on the evening row, and a correction to previous copy on this site: our own warmest platform setting is 2700 K, which measures M/P 0.45. That is above the M/P 0.40 figure some briefs quote as an evening cap, so an evening target written as "M/P < 0.40" cannot be met by colour temperature alone on this platform. It is met by dropping illuminance — 10 melanopic EDI at the eye is reached by dimming, not by pushing the spectrum warmer than 2700 K. Any tender clause that demands both M/P < 0.40 and a 2700 K minimum is internally inconsistent and should be rewritten as an m-EDI-at-the-eye cap.

Why 500 lux is not 500 lux

This table is the article in one screen. Both columns are the same measured illuminance at the same working plane. The only difference is the spectrum.

Source: rows 2–4 from the BY001 design-target curve; rows 1 and 5 are worked examples at M/P 0.30 and 0.90 (SUMLINK internal estimate, chosen to bracket the range a specifier is likely to meet in the market). The point of the table is arithmetic, not measurement.

Read it as a warning against a common procurement error: a brief that specifies "500 lux, CRI 90" and stops there can be satisfied by a source that delivers only 150 m-EDI — technically compliant, circadianly inert. The engineering response is not necessarily a higher illuminance. It is a spectrum with more melanopic content at the same lux, which is exactly what M/P quantifies.

One office day, in M/P and m-EDI

Here is a six-node schedule from the office program, translated from colour temperature into melanopic terms. The m-EDI column is obtained by scaling the 500 lx design curve linearly with illuminance — it is an internal calculation, not a laboratory measurement, and it is the method a specifier should use to model a schedule before commissioning.

Source: schedule nodes from the live office program (`/lighting-programs`); M/P values interpolated from the BY001 design-target curve (4500 K = linear interpolation between 4000 K and 5000 K = 0.74); m-EDI = illuminance × M/P, internal calculation.

The schedule shows where the useful work happens. Between 08:00 and 14:00 the system holds melanopic EDI above the WELL tier-1 threshold and above the 250 m-EDI research minimum for the first 90 minutes. By 18:30 it is at 112 m-EDI — which is low, but still an order of magnitude above the 10 m-EDI research recommendation for the pre-sleep window. The evening band cannot be reached by this schedule while the office is still lit to 250 lx. Getting to research-level evening exposure requires dropping illuminance to roughly 20–25 lx at the eye, which is a different design decision (task lighting, or simply turning the main run off) rather than a colour-temperature setting.

That is the practical lesson: M/P tells you about the spectrum you specify, and m-EDI tells you what the occupant actually receives. You need both columns in a specification. If a supplier gives you only one, you cannot model the schedule.

What a single ratio cannot tell you

M/P is a scalar summary of an entire spectral power distribution. That compression is useful and also lossy. Four limitations matter in practice:

• It does not describe the shape of the spectrum. Two sources can share an M/P ratio and differ in where their short-wavelength energy sits relative to the 480 nm melanopic peak. The ratio measures the weighted integral, not the fit.
• It says nothing about timing or duration. A 0.92 M/P source at 09:00 and the same source at 22:00 are, photometrically, identical. The circadian system treats them very differently. Exposure research is explicit that timing dominates (Brown et al. 2022).
• It is blind to light history. Prior hours of bright light change the magnitude of the response to the light in front of you. No single-source metric captures that.
• It does not capture uniformity at the eye. Vertical illuminance at eye level — not horizontal illuminance at the desk — is the quantity the research recommendations and WELL clauses are written against. A well-lit desk under a dark ceiling can leave the occupant's eye-level exposure far below the design value.

Source: limitations 1–2 and 4 are properties of the metrics as defined in CIE S 026:2018; limitation 3 and the timing emphasis follow Brown et al. 2022.

What the peer-reviewed literature actually recommends

The most-cited consensus recommendation for indoor exposure is Brown et al. (2022), written after the Second International Workshop on Circadian and Neurophysiological Photometry. It is a research recommendation, not a standard, and it is stated at the eye:

Source: Brown TM, Brainard GC, Cajochen C, Czeisler CA, Hanifin JP, Lockley SW, et al. (2022). Recommendations for daytime, evening, and nighttime indoor light exposure to best support physiology, sleep, and wakefulness in healthy adults. PLOS Biology 20(3): e3001571. doi:10.1371/journal.pbio.3001571. Source checked 2026-10-02 via the publisher's article page.

Read against the table above, this is why the morning band targets M/P ≥ 0.90 at working illuminance. It is also why the evening band is a dimming problem, not a colour problem: 10 melanopic EDI at the eye is roughly what a 2700 K, M/P 0.45 source delivers at about 22 lx. No tunable-white fixture reaches it at office illuminance.

How to read SUMLINK's M/P numbers

Every figure in this article taken from our own platform is labelled Design target. On our catalogue pages that label has a specific meaning: the value is an engineering design value derived from the LED and driver platform specification. It is not a laboratory report. Where a value is backed by a test report, the catalogue says so and the report is available on request.

We state this distinction because the alternative — a badge that implies third-party certification without a named laboratory and a report number — is not something a specifier can act on. If you are writing a tender, the clause you want is not "supplier-verified" but "a test report from an accredited laboratory, naming the instrument and the measurement geometry, delivered with the BOM." Ask for that, and the M/P number stops being a marketing claim.

Source: SUMLINK catalogue policy — `Design target` vs. test-report-backed values, as published on the BY001 page.

Related Reading

• The M/P ratio specification guide
• Our measurement and metric explainer
• Run this as a scheduled program
• The BY001 linear platform

Frequently Asked Questions

Is the M/P ratio the same as melanopic DER?

Yes. CIE S 026:2018 calls the quantity the melanopic daylight efficacy ratio (melanopic DER). M/P is the industry shorthand for the same division: melanopic EDI divided by photopic illuminance.

What M/P ratio should I specify?

It depends on the band. For morning activation, M/P 0.90 or higher at working illuminance. For core daytime, enough to hold WELL v2 L03 (at least 150 EML, or 136 melanopic EDI) for at least four hours. For the evening, stop thinking in M/P and write a melanopic EDI cap at the eye instead, because 10 melanopic EDI is reached by dimming rather than by spectrum.

Can a higher M/P ratio make up for too little light?

Partly. Melanopic EDI is the product of illuminance and M/P, so a spectrum with higher M/P delivers more circadian signal at the same lux. But the recommendations are written as an absolute melanopic EDI floor, so a dim room with a very cool source can still fall short.

Is a high M/P ratio bad in the evening?

Yes. The same spectral quality that supports alertness in the morning delays sleep onset in the evening. That is why a tunable system is worth specifying: it lets one installation serve both bands.

Do SUMLINK's M/P figures come from a test report?

Values published as Design target are engineering values from the LED and driver platform specification, not independent laboratory results. Where a value is test-report backed, the catalogue page says so and we supply the report with the BOM on request.

Send us a room schedule and we will return it in melanopic terms. We will convert your existing illuminance plan into M/P and m-EDI per time node, flag any band that misses its target, and issue the per-SKU specification sheets for the fixtures that close the gap. → Request a room-by-room M/P schedule (`/science`)