We Measured Seven IPL Long-Pass Filters (470–640nm): T50% Cutoff, Blocking, and Passband Data

When OEM clients ask about our optical filters, the first question is almost always the same: what is the actual cutoff wavelength? It is the right question. The cutoff determines where a device starts letting light through, and whether the short wavelengths are actually blocked. The number printed on most spec sheets is typed, not measured. And if that number is off, the product you customize for a target market may simply be cutting the wrong spectral window. So we took the seven long-pass filters we sell — one for each nominal cutoff from 470nm to 640nm — and measured every one on a real spectrophotometer. The data is published here in full. Numbers beat marketing claims.

How We Tested

Instrument: a Shimadzu UV-1800 UV-Vis spectrophotometer, wavelength scan mode, slit 2, scanned from 325nm to 1100nm at 1nm intervals. Samples: seven long-pass cutoff filters, nominally 470, 510, 530, 550, 580, 600, and 640nm. The source reports are lab printouts; we digitized and cross-checked them, and the T50% values are taken directly from the instrument’s printed output. Spectrum testing is part of our incoming-filter inspection, and finished devices go through a full test flow before shipping.

The incoming-inspection standard behind this test was written by Eric, who leads iShine’s product development and quality system. And this report is not a one-off showcase — it is one of the standard steps for every batch of optical filters we receive. When we promise OEM clients spectral consistency, that promise rests on this inspection data, not on a verbal guarantee.

The Results in One Table

Nominal (nm)	Measured T50% (nm)	Deviation	Max blocking T	Avg passband T	10–90% width
470	477.7	+7.7	2.20%	95.06%	17.3 nm
510	505.4	−4.6	4.18%	95.03%	16.7 nm
530	526.5	−3.5	2.06%	96.16%	29.2 nm
550	543.6	−6.4	1.70%	92.38%	2.9 nm
580	581.5	+1.5	4.41%	95.18%	18.5 nm
600	600.2	+0.2	2.34%	95.99%	22.1 nm
640	641.6	+1.6	4.81%	93.51%	25.4 nm

Blocking region = more than 60nm below the T50% cutoff; passband = 80nm above T50% to 1100nm; width = the wavelength span from 10% to 90% transmittance.

Here is what each metric actually decides. The T50% cutoff sets where the spectral window starts cutting; how far it is from nominal decides whether the optical design your client built around the datasheet really holds. Blocking-region transmittance decides how cleanly the unwanted wavelengths are cut, which feeds directly into safety and energy distribution. Passband transmittance decides how much of the wanted light gets through — less loss means more energy reaches the skin at the same charge voltage. And the 10–90% transition width is the speed at which the filter switches from blocking to passing: the narrower it is, the clearer the boundary between the wavelength you want (say 550nm) and the one you do not (say 530nm), and the less energy is wasted.

Merged Transmission Spectra

All seven measured transmission curves from 325nm to 1100nm are above, with the T50% point marked on each. The blocking regions sit at the bottom, the curves pull up steeply past the cutoff, and the passbands land on a 92%–96% plateau.

Can You Trust the Nominal Cutoff?

The measured T50% values are 477.7, 505.4, 526.5, 543.6, 581.5, 600.2, and 641.6nm. Against the nominal values, the smallest deviation is 0.2nm and the largest is 7.7nm — the worst case is still under 1.7% of the nominal cutoff.

The takeaway: nominal values are fine for a spec sheet, but where each filter actually cuts is decided by the measured curve, not the printed number. For the bigger question of how to choose a wavelength at all, we covered that separately (Choosing the Right Wavelength for IPL Devices).

Blocking: Are the Short Wavelengths and UV Actually Cut?

In the blocking region, the maximum transmittance of all seven filters stays under 5%, and for most of them the average is only 1.5%–2.3%. Short-wavelength photons carry more energy, and UV needs to be cut cleanly — this is the layer that decides whether a device is safe (UV Filtering Wavelengths: Why IPL Is Safe). Cutting the short wavelengths so the light targets the melanin in the hair follicle instead of dumping energy on the epidermis is also the basic premise of the American Academy of Dermatology’s laser hair removal guidance.

Passband: Does the Energy Get Through?

Average passband transmittance runs from 92% to 96%. The 550nm filter is the lowest at 92.38%; the rest are above 94%. Higher passband transmittance means less of the lamp’s energy is lost in the glass, so more reaches the skin at the same charge voltage. Energy that gets through is energy that works — a randomized controlled trial comparing IPL with diode laser hair removal confirmed IPL as safe and effective for hair reduction. The wavelength range printed on a device spec sheet — say 510–1200nm — is a printed window; we wrote separately about how that maps to the measured curve (IPL Wavelength Range).

Transition: How Clean Is the Cut?

From 10% to 90% transmittance, the seven filters span from 3nm to 29nm. The 550nm filter is the steepest at roughly 3nm — a very clean cut. Reading a curve that steep carries more error, though: for this filter the curve-derived T50% sits about 9nm away from the instrument’s printed value, so we take the printed 543.6nm as authoritative. A narrower transition means a sharper boundary between blocking and passband, which matters for devices that need a fixed spectral window.

What This Means for OEM Buyers

When you read a filter datasheet, do not stop at the printed wavelength range. Look at three things: how far the T50% is from nominal, how low the blocking region goes, and how high the passband transmittance is. These three decide whether the output window is clean and how much energy is wasted. If you want the cutoff moved for a target market, or a different substrate such as sapphire or coated glass, our optical filters page lists the configurable options, and we can design the optical stack together with the whole device. At the device level, home-use IPL safety is governed by international standards such as IEC 60335-2-113; cutting the unwanted wavelengths with a filter is part of that optical-safety design, and we keep the measured spectrum of every filter batch on file for review.

Data Notes

These results are single-sample measurements, not batch statistics. The spectra were digitized from lab printouts; wavelength accuracy is about ±5nm and transmittance accuracy about ±2%, with T50% values taken from the instrument’s printed output. Want the raw data, or the full transmission curve of one filter at a specific wavelength? Ask us directly.

Bottom Line

Seven long-pass filters, accurate cutoffs, clean blocking, low passband loss — that is our actual answer to OEM clients. Want to talk customization, or get the full test data? Contact us.

Related Reading

• Choosing the Right Wavelength for IPL Devices
• UV Filtering Wavelengths: Why IPL Is Safe
• IPL Wavelength Range: 510–1200nm vs 610–1200nm
• Quality Control in IPL Manufacturing: Best Practices
• IPL Optical Filters