A dual-wavelength console presents you with two buttons and one fiber. The 980 nm button. The 1470 nm button. The brochure explains the difference in one line, and the line is about water absorption. Depending on which brochure you have in front of you, the 1470 nm wavelength is absorbed by water three to four times more strongly than 980 nm, or ten times, or forty, or sixty.
Those numbers cannot all be describing the same measurement, and they are not. The reason they disagree is more useful than the correct figure, because it is the same reason that “which one is better” is the wrong question to ask about these two wavelengths.
This piece is not a head-to-head with a winner at the end. It is an attempt to say what each wavelength actually does to a fat compartment, using optical properties that have been measured rather than asserted, and to be explicit about where the marketing story and the measured story separate. Both wavelengths are in our product line, which is a reason to be precise rather than a reason to be neutral.
The short version: the phrase two chromophores is correct, and the nouns are wrong. The two chromophores are water and haemoglobin. Neither of them is fat.
Two chromophores, and neither of them is fat
Start with what is actually inside the thing you are trying to destroy. Adult adipose tissue is roughly 60–87% lipid by volume depending on the site and the patient, with the remainder water, protein and extracellular matrix. The adipocyte itself is more extreme — a single fat cell is about 95% triglyceride droplet by volume, with the cytoplasm pushed into a thin rim. You are aiming at a bag of oil.
Lipid has its own near-infrared absorption bands, and they are not where the lipolysis market works. The C–H stretching overtones put lipid absorption peaks at approximately 930, 1040, 1200 and 1720–1760 nm, with an additional band near 1400 nm (Anderson et al., Lasers Surg Med 2006; Wilson et al., J Biomed Opt 2015). Anderson’s group went further and measured what those bands buy you: at 1210 and 1720 nm, photothermal excitation of fat was twice that of dermis, and at 1210 nm a subcutaneous fat layer several millimetres thick could be damaged without apparent injury to the overlying skin. That is genuine lipid selectivity, achieved with a free-electron laser, and it is the only clean demonstration of it in the literature.
Now place the two commercial wavelengths on that map.
- 980 nm sits in the trough between the 930 nm and 1040 nm lipid bands, a few tens of nanometres from the 970 nm water overtone. It is not on a lipid peak and not on a water peak. It is on the shoulder of the water band and, more importantly, on the shoulder of the haemoglobin band.
- 1470 nm sits on the flank of a band near 1430–1450 nm which is a combined feature — the first overtone of O–H stretching in water at ~1430 nm, overlapping a lipid band near 1400 nm. Of the two commercial wavelengths, this is the one closer to a fat/water band.
Neither is lipid-selective in the sense Anderson demonstrated. Both are, strictly speaking, water wavelengths, and the tissue you are treating is mostly not water. That single sentence explains most of the confusion downstream.
The human pilot of the genuinely lipid-selective 1210 nm wavelength is worth reading before anyone gets romantic about fat-specific lasers. Twenty-four subjects, abdomen, 70–90 J/cm²: the exposures were painful enough to require local anaesthesia in most subjects, the contact-cooling device malfunctioned, and two of fourteen subjects developed epidermal damage. At 4–7 weeks there were lipomembranous changes in 89% of treated sites — and in 33% of the untreated control sites, attributed to handpiece pressure causing tissue ischaemia rather than to the laser at all (Wanner et al., Lasers Surg Med 2009). A fat-specific wavelength that cannot be delivered comfortably, and whose endpoint shows up in skin that was never irradiated, is a useful corrective to the genre.
What 980 nm buys you: a wavelength that can tell blood from water
Here is the property that makes 980 nm interesting, and it is not the water number.
Water absorbs 980 nm weakly — its absorption coefficient is about 0.45–0.50 cm⁻¹, which corresponds to an optical depth of roughly two centimetres in pure water. Meanwhile oxyhaemoglobin at 975 nm still has an extinction coefficient of about 1.20 mM⁻¹ cm⁻¹, which at a whole-blood haemoglobin concentration of 150 g/L works out to roughly 6–7 cm⁻¹ (Wray et al., Biochim Biophys Acta 1988, with the standard scaling to whole blood at 45% haematocrit). Against the 0.45 cm⁻¹ of the interstitial water sitting beside it, a blood vessel at 980 nm is roughly fourteen times more absorbing than the tissue around it.
That ratio is the whole job. 980 nm is the only one of the two wavelengths that can distinguish a vessel from the fluid next to it. When the tip passes a bleeding point, the light is preferentially absorbed there, and the vessel seals. The early endovenous literature described the 810–980 nm diode generation exactly this way — as the wavelengths with “a high absorption coefficient of oxyhaemoglobin” — before the field moved to water-absorbing wavelengths above 1000 nm (J Vasc Bras 2017;16(1):23–30).
At 1470 nm that discrimination is gone. Oxyhaemoglobin’s contribution at 1470 nm is negligible, and blood is about 80% water, so blood and interstitial fluid absorb almost identically. A 1470 nm fiber does not find vessels. It finds water, which is everywhere, and does not care which compartment it is in.
So the first “job” is not “lipolysis versus haemostasis” in the marketing sense. It is selectivity versus uniformity: 980 nm can pick out a compartment, and 1470 nm cannot.
What 1470 nm buys you: confinement, not reach
The other job is easier to state and easier to mismeasure.
Water absorption at 1470 nm is about 24.8 cm⁻¹, against 0.45 cm⁻¹ at 975 nm — a factor of roughly 55 (J Vasc Bras 2017;16(1):23–30, Table 7). In pure water, 1470 nm light has an optical depth of about 0.4 mm; at 980 nm it is about 2 cm. A short-wavelength review states the same relationship more bluntly: the absorption coefficient of water is “over 60 times greater at 1440 nm than at the 970 nm NIR peak” (Wilson et al., J Biomed Opt 2015).
This is why the two best-selling claims about 1470 nm both contain the same word, and why that word is doing all the work: precise, controlled, shallow, superficial. All four describe confinement — the energy is deposited in a thin shell rather than a broad volume. None of them describes reach, and 1470 nm has less reach than 980 nm in any water-containing tissue, by construction. A brochure that sells 1470 nm for its “deep water absorption” is selling a contradiction.
Confinement has a real clinical consequence, and it is the reason 1470 nm became the wavelength of choice for the submental area and the jawline. If the energy lands in the first few tenths of a millimetre, the structures in that first few tenths of a millimetre get hot — and in the subdermal plane those structures are the reticular dermis, the fibrous septa, the vessel walls, and the interstitial water between adipocytes. Collagen denatures in the 48–50 °C range, so confining the deposit at the dermal boundary buys contraction with a smaller energy budget. It is also the fastest way to burn the skin if the fiber stops moving.
There is a second consequence that gets said less often. If 1470 nm is absorbed by water, and the adipocyte interior is 95% lipid, then 1470 nm does not absorb efficiently in the thing it is supposed to be destroying. Most of what it heats is the minority component around and between the fat cells. The adipocyte is killed downstream of that — by conduction from the hot interstitium, by the membranes and septa reaching coagulative temperatures, and by the mechanical disruption of vaporising water. That is a real mechanism, and it is why 1470 nm works. But it means “1470 nm melts fat” is shorthand, and the mechanism behind the shorthand is heat conduction from a thin hot boundary — a different dose strategy from heating the fat directly.
The comparison that matters is in fat, not in water
Which brings the argument to the only number that decides anything: the effective penetration depth in adipose tissue, not in a cuvette of water.
The measured optical properties of human subcutaneous fat are, roughly:
| Quantity | ~980 nm | ~1470 nm | Note |
| Water absorption, µa | 0.45–0.50 cm⁻¹ | 24.8 cm⁻¹ | pure water (J Vasc Bras 2017) |
| Ratio in pure water | 1× | ≈55× | derived |
| Fat absorption, µa | 1.0–1.6 cm⁻¹ | ≈4–5 cm⁻¹ (estimate) | Bashkatov 2005; Hamdy 2022 |
| Fat reduced scattering, µs′ | 9–24 cm⁻¹ | ≈7 cm⁻¹ | Bashkatov 2005; power law 1050.6·λ^−0.68 |
| Effective depth in fat, δ | ≈1.4 mm | ≈0.7 mm | calculated; cf. Wassmer 2010 |
The bottom row is arithmetic, not measurement, so the assumptions should be visible. Using the standard diffusion form δ = 1/√(3·µa·(µa + µs′)): at 980 nm with µa = 1.2 cm⁻¹ and µs′ = 13 cm⁻¹, δ ≈ 1.4 mm. At 1470 nm, if fat is roughly 20% water then water contributes about 5 cm⁻¹, and scattering has fallen to about 7 cm⁻¹, giving δ ≈ 0.7 mm. The result happens to land where the only direct measurement of this quantity lands — Wassmer and colleagues measured the penetration depth in fat across 900–1320 nm and found approximately 1.5 mm for all of them, with 1440 nm the sole exception because it is more absorbed by subcutaneous fat (Wassmer et al., Photomed Laser Surg 2010).
Read that again, because it is the pivot of the whole article. A fifty-five-fold difference in water absorption becomes roughly a two-fold difference in fat. Not fifty-five. Two.
Two things collapse it. First, the chromophore is a minority constituent: water is only about a fifth of what is in front of the tip, so its absorption coefficient is diluted by five before the light even starts. Second, and more importantly, fat is a highly scattering medium — the reduced scattering coefficient is roughly ten times the absorption coefficient at 980 nm — and once scattering dominates transport, further increases in absorption move the penetration depth only slowly. The medium does not care what the marketing ratio is; it cares what fraction of the light it can absorb per unit path length, and in fat that fraction is small at both wavelengths.
And a caveat that belongs in the article rather than in a footnote: the optical properties of fat are themselves disputed by an order of magnitude. Bashkatov’s group reported a mean absorption coefficient of 1.1 ± 0.03 cm⁻¹ for human subcutaneous fat across 620–1000 nm; Simpson and colleagues, with a different sample-handling protocol, reported 0.11 ± 0.03 cm⁻¹ for the same tissue and range. Bashkatov flags the possibility that his own values are overestimated because lateral light loss was not fully excluded (Bashkatov et al., J Phys D Appl Phys 2005). Any single number quoted for fat absorption — including the ones in the table above — is therefore an order-of-magnitude statement, not a figure. That applies to the marketing numbers and to ours.