Transmitted light, dark field or reflected light: why lighting is decisive for opaque films and nonwovens

In optical web and film inspection, line scan camera resolution alone does not determine whether a defect is reliably detected. The decisive factor is the optical contrast between the defect and the intact base material. Day-to-day laboratory and production line practice reveals three typical pitfalls – and how to avoid them.

1. White and opaque films in transmitted light
Transmitted light (bright field) is the standard choice for transparent films: the camera and LED light face each other, and dark inclusions such as black specks or insects stand out sharply. With heavily pigmented, white, opaque, metallised or coated films, however, pure transmitted light reaches its physical limits. The material scatters or blocks the light so strongly that the resulting contrast on the surface is often insufficient for reliable evaluation.

The practical solution is to switch to reflected light. Here, the LED unit illuminates the web from the camera side at a defined angle. Surface defects, coating flaws, contamination and gloss variations become clearly visible without the light having to pass through the web.

2. Fine, transparent gels without colour contrast
The opposite situation occurs with highly transparent films: unmelted or cross-linked gels and fisheyes often have almost the same light transmittance as the surrounding film. In standard bright-field lighting they barely cast a shadow. Dark-field lighting (or a dark-field aperture) makes these defects visible: direct light is blocked so that only light refracted and scattered at the edges of the gel enters the lens – causing the gel to light up against a dark background.

3. Strongly fluctuating camera signals on nonwovens
Nonwovens pose a special challenge due to their open, irregular fibre structure. Unlike a homogeneously extruded film, their local density (“cloudiness”) naturally fluctuates considerably. If operators attempt to run automatic camera calibration on a heavily textured nonwoven using recipe parameters designed for smooth film, the system will often detect excessive signal fluctuation and abort the automatic routine.

The remedy is a targeted recipe adjustment in the CAMEN inspection software: exposure time, camera calibration and the upper and lower thresholds are matched to the natural structural variance of the nonwoven. This allows the system to learn the baseline texture of the fibres as good material while reliably reporting genuine production defects such as holes, thin spots, oil stains or foreign fibres.

Testing in advance instead of guessing on the line
Whether transmitted light, dark field or reflected light delivers the best results for a specific material – and how an inline system such as the FIS 1000 or ITG 850 – LC should be configured – can be verified in advance in the laboratory. On the ITG 500 – Bench Analyser laboratory bench, we test submitted samples up to A3 format using a 4K line scan camera (approx. 60 µm resolution at a 245 mm field of view) across all three lighting modes.

To ensure that a sample analysis at our laboratory in Flörsheim am Main delivers reliable data straight away, we recommend four simple preparation steps: include a defect-free reference sample for each material, keep the inspection area free from adhesive tape or permanent marker ink, mark known defects outside the test area and state the material type, colour and layer structure. You will receive a PDF report detailing the findings per sample, defect sizes in mm and mm², defect images and a clear system recommendation.