Irradiance vs dose: the math nobody does for you
Verdict: The most useful thing you can do after unboxing a panel is calculate your actual dose — and almost nobody does. The formula is one line. The inputs are where the fight is, because manufacturer irradiance numbers are marketing until someone independently measures them. Here's the math, the derating logic, and a worked example. Bookmark this; every dosing recommendation on this site links back to it.
Dosing confusion — how long, how far, how often — is the dominant post-purchase question in the red light therapy community (ev028). This page is the answer to all three, in the order the math demands: distance first, then time, then frequency.
The formula (this is the whole page)
Dose (fluence) = irradiance × time:
J/cm² = mW/cm² × seconds / 1000 [estimated] (ev018)
That's it. One multiplication, one division. If your panel delivers 50 mW/cm² [measured] at your treatment distance and you sit there for 10 minutes (600 seconds), your dose is:
50 mW/cm² [measured] × 600 / 1000 = 30 J/cm² [estimated]
Now the hard part: where did that 50 come from, and can you trust it?
Why the input number is the fight
Manufacturer irradiance specs are [claimed] — they are marketing until independently measured (ev010). Three problems stack:
- Brands publish peak-spot or close-distance numbers. The figure on the spec sheet is usually the brightest point on the panel at the closest distance, not what you'll actually receive across your treatment area (ev010).
- Distance quoting is inconsistent. Most brands quote irradiance at 6 inches; Joovv quotes at 12 inches, making cross-brand comparison misleading without normalizing to the same distance (ev021).
- Independent testers measure 30-60% below claimed irradiance on many panels (ev010). Solar power meters — the tool most people use — over-read LED panel output, so even community measurements can be generous (ev010).
Meanwhile, independent irradiance and spectral measurement exists for only a few dozen devices out of hundreds on the market (ev017). Most panels have no verified numbers at all.
Distance changes everything
Panel irradiance falls off steeply with distance (ev020). This is not a clean inverse-square law — an LED panel is an extended source, not a point — but the practical rule holds: double your distance and you get roughly half or less irradiance in the near field (ev020).
This means a panel that delivers 70 mW/cm² [claimed] at 6 inches might deliver ~35 mW/cm² [estimated] at 12 inches. Same panel, same LEDs — half the dose per minute just by sitting farther back.
Pick: Measure your distance before you calculate dose. "About arm's length" is not a number. Get a tape measure.
What dose are you aiming for?
The PBM literature most often reports effective doses in the ~3-60 J/cm² [estimated] range at the tissue surface, depending on the target (ev019). No single agreed number exists (ev019). We are not going to pretend otherwise.
What the literature does show is a biphasic dose response — beyond a threshold, more dose can mean less effect, not more (ev013). The direction is well-documented; the exact thresholds per condition are not established (ev013). Anyone publishing a single "optimal dose" number for all uses is overclaiming.
Here's how to use the range without overclaiming:
| Target | Ballpark dose range | Source |
|---|---|---|
| Skin surface (red, 630-670nm) | ~3-15 J/cm² [estimated] | Literature range, not a prescription (ev019) |
| Deeper tissue (NIR, 810-850nm) | ~10-60 J/cm² [estimated] | Literature range, not a prescription (ev019) |
These are the ranges the academic literature works in, not targets we're declaring. If your calculated dose lands somewhere in the 3-60 J/cm² [estimated] band, you're in the ballpark. If you're at 120 J/cm² [estimated] because you bought the biggest panel and sit 2 inches away for 20 minutes, you may be past the useful range — and you won't know it because the threshold isn't published (ev013).
Worked example: a typical panel session
Let's say you have a mid-size panel. The spec sheet says 100 mW/cm² [claimed] at 6 inches. No independent measurement exists for your specific model.
Step 1 — Derate the claimed number. Independent testers find 30-60% below claimed across the market (ev010). Split the difference: derate by ~45%.
100 mW/cm² [claimed] × 0.55 = ~55 mW/cm² [estimated] at 6 inches
Step 2 — Adjust for your actual distance. You sit at 10 inches, not 6. Irradiance falls off steeply with distance (ev020). From 6 to 10 inches is roughly 1.7× the distance; call it ~60% of the 6-inch value.
55 mW/cm² [estimated] × 0.60 = ~33 mW/cm² [estimated] at 10 inches
Step 3 — Pick your time. You want to land in the ~10-30 J/cm² [estimated] range for a general skin/wellness session (ev019).
33 mW/cm² [estimated] × seconds / 1000 = 20 J/cm² [estimated]
seconds = 20 × 1000 / 33 = ~600 seconds = ~10 minutes
Result: 10 minutes at 10 inches gives you ~20 J/cm² [estimated] — comfortably in the literature's effective range for surface targets (ev019), and below the point where biphasic dropoff becomes a concern for most endpoints (ev013).
If you moved to 6 inches instead: 55 mW/cm² [estimated] × 600 / 1000 = 33 J/cm² [estimated]. Still in range, but higher. Whether that's better or worse depends on the condition — and the thresholds aren't established (ev013).
The derating decision tree
This is the pick per branch:
- Your panel has independent measurement (from a reputable tester with a proper meter): Use the [measured] number at your distance. Calculate dose directly. This is the gold path; few panels have this data (ev017).
- Your panel has no independent measurement: Derate the [claimed] irradiance by 30-60% (ev010), then adjust for distance (ev020), then calculate time to hit your target J/cm² [estimated]. You're working with an estimate, not a measurement — label it that way in your own notes.
- You have a solar power meter and want to DIY: Your reading will over-read LED output (ev010). Treat your number as an upper bound, not ground truth. Derate it by at least 20% before calculating dose.
Frequency: how often should you dose?
The biphasic dose response (ev013) means "more sessions" is not automatically better. The literature does not establish a universal optimal frequency any more than it establishes a universal optimal dose (ev019). What we can say:
- The realistic results horizon for skin endpoints in the literature and honest reviews is ~8-12 weeks (ev033). If you're dosing correctly, that's the timeline to expect visible change — not three days.
- The weeks 3-8 doubt window is where most buyers churn because they expected faster results (ev033). Knowing the math doesn't make the wait shorter, but it does tell you whether you're in the game or wasting your time.
- Biphasic response suggests a rest day between sessions is defensible — you're not "leaving gains on the table" by skipping a day (ev013). The exact rest-to-treatment ratio is not established.
Pick: Start with 3-5 sessions per week at a calculated dose in the 3-60 J/cm² [estimated] range. Reassess at week 8 based on whether anything changed (ev033). If nothing changed, the problem is probably not frequency — it's either dose (too low or too high per biphasic response) or the endpoint itself isn't well-evidenced for RLT (ev016).
The honest caveats
We are doing arithmetic on estimates. Here's what we know and don't know:
- The formula is physics, not biology: J/cm² = mW/cm² × seconds / 1000 [estimated] is correct photometry (ev018). The math is not in question.
- The dose ranges are literature-derived, not consensus: ~3-60 J/cm² [estimated] is where studies report effects, but no single agreed number exists (ev019).
- The derating factors are empirical averages from independent testing, not laws: 30-60% below claimed (ev010) describes the market pattern, not any specific panel. Your panel could be better or worse.
- The biphasic response is real but its thresholds are not published per condition (ev013). We know too much can be worse; we don't know where "too much" starts for your specific use case.
- Wavelength selection (660 vs 630, 850 vs 810/830) is a permanent unresolved community debate; head-to-head consumer evidence does not exist (ev011). This page covers dose math, not wavelength optimization.
What we'd do
Calculate your dose once. Write it down. Use the derating tree that matches your evidence level. If your panel has no independent measurement, derate aggressively — the 30-60% gap is real (ev010), and pretending your [claimed] number is your actual number is the most common dosing error in the community.
The single highest-value action: stop trusting the spec sheet and start doing the arithmetic. It takes 90 seconds and it's more useful than another hour of reading reviews.