Practical comparison
Compare backlight, ring, dome and directional light on the same samples. This procedure separates light geometry from pulse timing and records image quality, inspection decisions and installation conditions.
A lighting trial should answer whether the required defect stays distinguishable from acceptable variation. A bright, attractive image is insufficient. This article provides a compact comparison procedure; the lighting guide explains the optical and electrical choices in detail.
Label the samples and record their expected outcomes before adjusting the inspection. Include good parts, known defects and borderline examples. Keep some independent parts for validation. Use the actual surface finish, including permitted oil, film wrinkles or printing variation.
| Candidate | Compare it for | Check its boundary |
|---|---|---|
| Backlight | Outer contour, open holes or missing material. | It does not reveal front-surface print on an opaque part. Refraction can move apparent edges on transparent parts. |
| Ring light | Surface presence and print contrast within a compact field. | Direct reflections can obscure the feature. Test product tilt and field corners. |
| Dome light | Print or surface detail on curved, reflective products. | Diffuse light may suppress scratch contrast. Check space and reflections from the camera opening. |
| Directional bar or low-angle light | Edges, scratches or embossed features. | Harmless texture may also become visible. Rotate the part within its allowed range. |
Pulsed light is an operating mode, rather than another geometry in this table. Several geometries can use either continuous or pulsed operation if the equipment supports it.
First keep the camera position, lens, focus and image processing fixed. Capture each identified part under every candidate light. Record light position, intensity, exposure, aperture and any filters. Keep automatic exposure or gain disabled where the equipment allows, so the camera does not conceal the effect being compared. Avoid saturation.
Then optimise each promising candidate within the permitted exposure and equipment limits. A common exposure makes the initial comparison easier, but final selection must account for each option's achievable image quality and installation requirements. Record every changed setting.
For each image, inspect the feature region and its background. Look for overlap between good and defective appearances, glare, shadows and loss of detail. Run the intended inspection tool and retain its output alongside the image. Compare independent validation samples after setting the decision limit.
Test minimum, nominal and maximum speed and the shortest expected interval between parts. Repeat at the position and height extremes with production shielding and protective windows installed. Check whether cleaning access or a format change can disturb the light position.
Calculate the exposure limit from allowed displacement divided by product speed. For example, 0.1 mm at 500 mm/s gives 0.0002 seconds, or 200 microseconds. This illustrative limit does not prove that enough light reaches the sensor. Test the signal at that exposure.
If pulsing is required, verify the sensor's exposure window, trigger delay, pulse width and repetition rate. Use the permitted light/controller current at that combination. A short pulse only limits effective motion blur when it dominates other illumination during the image.
Use a test record with sample ID, light configuration, speed, position, expected outcome, actual outcome and result time. Count missed defects and false rejects separately, with the number of relevant samples. Set acceptable limits before final tuning. Identify repeated passes of the same part rather than treating them as independent product variation.
Include light, controller, filters, mounts, cables and protective enclosure in the investment. Add trial and validation hours multiplied by the relevant rates, then cleaning and replacement effort. Estimate reduced interventions or rejects only from validated comparisons with the existing process.
Select the candidate that meets the agreed inspection limits and fits the installation and maintenance conditions. Retain reference images and a restorable settings record. For the next discussion, bring the samples, comparison images, available space and speed range so the remaining uncertainty can be tested.

Send photographs or samples of the product and defect, together with working distance and line speed. This gives us enough information to advise on lighting geometry and control.