Should I choose hard-coated or soft-coated bandpass filters?

Should I choose hard-coated or soft-coated bandpass filters? Let's break it down with a no-nonsense, practical comparison so you get the answer you need for your specific project.

Hard-coated vs. Soft-coated Bandpass Filters

Choose hard-coated (sputtered/IBS, dense films) for durability, spectral stability, steep edges, high blocking, higher laser power, and field use.
Choose soft-coated (evaporated, more porous) when you need very large sizes at lower cost or good enough performance for more gentle indoor environments.

What the terms actually mean
Hard-coated = ion-beam or magnetron sputtered (often with ion assistance). Layers are dense, hard, and non-porous.
Soft-coated = evaporated multilayers (typically without full densification). Layers are more porous/softer and can take in moisture.

Side-by-side comparison (what usually matters)

Quick chooser: When to pick hard-coated or soft-coated?

• Pick Hard-coated if you need:
  • Tight spectral control: steep edges (10–20 nm FWHM with OD 5–6 outside band), minimal ripple.
  • Specs to hold under high humidity, cleaning, and temperature swings.
  • Field or rental gear (photo/cine/stage), or outdoor sensors.
  • High irradiance or laser exposure.
  • Long service life and consistent re-orders.
• Pick Soft-coated if you need:
  • Large diameters (150–300 mm+), tighter budget.
  • Moderate performance is okay (OD 3–4 blocking, edges not ultra-steep).
  • Benign environments (clean lab, controlled humidity), limited cleaning.
  • Shorter lead time/cost matters more than ultimate robustness.

Common specs to check (for any filter):

• Center wavelength / FWHM at your AOI (e.g., 0° or 5°) and temperature (e.g., 23 °C).
• Blocking range and OD (OD ≥ 5 from 200–1100 nm outside the passband).
• Wavefront/flatness & scratch-dig (e.g., λ/4 P-V @ 632.8 nm, 40-20 for imaging; 60-40 can be fine for stage).
• Humidity/thermal drift (request ΔCWL and ΔT specs after high-RH soak and thermal cycle).
• Angle sensitivity (specify your AOI; even a few degrees may shift the passband).
• AR coatings on both sides to cut ghosting; ask for stray light data if important.

Pitfalls to avoid

• Treating soft coat like glass: aggressive wiping can micro-scratch or alter spectra over time.
• Ignoring humidity: soft coats can shift a few nm in the visible after humidity exposure.
• Under-spec'd blocking: OD 3 may look fine but can allow flare/ghosting in bright scenes.
• Assuming all bandpass filters are the same: edge steepness and out-of-band behavior vary a lot between methods.

Application-specific nudges

• Machine vision / factory lines: Hard-coated preferred (stable CWL, lower scatter, repeatable).
• Fluorescence/biomed: Hard-coated for high OD off-band and clean edges.
• LIDAR / NIR beacons: Hard-coated for stability and sun/temperature swings; better for high flux.
• Stage/film lighting: Hard-coated holds up better (if very large panels are needed on a budget, soft-coated can work with protective cover glass and looser specs).
• Large astronomy windows/shrouds: Soft-coated may be most economical for size, but mind moisture/temperature and verify blocking.

If you're torn, use this decision stub

• Environment harsh? (humidity, cleaning, temperature, high flux) → Hard
• Need steep edges/high OD? → Hard
• Diameter >150 mm and budget tight? → Soft (with cover/protection)
• Laser or bright source? → Hard
• Short, low-risk pilot build? → Soft can be a cost-saving bridge

Spec wording you can drop into a PO (edit values):

Bandpass filter, CWL ___ nm, FWHM ___ nm, AOI 0° (±___°) at 23 °C;
Blocking OD ≥ 5 from 200–1100 nm outside passband; peak T ≥ ___%.
Preference: hard-coated (sputtered); provide ΔCWL after 24 h at 95% RH and thermal cycle −20 to +70 °C.
Surface quality 40-20 (or 60-40 for non-imaging), AR both sides, substrate ___, diameter ___ mm, thickness ___ mm.

Bottom line

• If performance, stability, and robustness matter—even a little—go hard-coated.
• Choose soft-coated mainly for very large, budget-sensitive parts in gentle environments where moderate drift and blocking are okay.
• If you want, tell me your target wavelength, bandwidth, diameter, AOI, and environment. I can help you turn it into a precise, vendor-ready spec and suggest realistic tolerances.