Machine vision component maker REVOX will use VISION 2026, the industry's trade fair in Stuttgart, to argue that lighting — not the camera sensor — now decides whether an imaging system meets its inspection specification. The company has framed its presence around what it calls machine vision's "battle for light."

The premise is not new to anyone who has commissioned an inspection cell, but it is one that purchasing decisions routinely underweight. A camera with a given quantum efficiency and resolution will only deliver its datasheet performance if the photons arriving at the sensor are controlled in geometry, spectrum, intensity and timing. REVOX's core business sits precisely at that interface: engineered LED illumination for industrial image capture.

Why does lighting dominate the imaging budget?

The physics is straightforward. Every camera specification — minimum detectable contrast, signal-to-noise ratio, effective frame rate at usable exposure — is a conditional statement. It holds only for a defined photon flux at the sensor plane. Change the illumination angle, and a specular surface either blinds the sensor or disappears into shadow. Change the spectrum, and the reflectance contrast between a feature and its background can invert.

Lighting engineers therefore treat the light source as the first component selected in a vision system, not the last. The geometry question comes first: bright field, dark field or diffuse illumination, chosen against the surface finish of the part. Then the spectral question, driven by the materials being separated — a task where the application, not the camera, dictates the wavelength. Finally the temporal question, whether the source runs continuous or strobed, with strobe pulses synchronized to the exposure window so that a moving part is frozen without pushing the sensor to short exposures it cannot support at the available intensity.

Each of those choices is an engineering trade measured in concrete units — irradiance at the working distance, spectral half-bandwidth, pulse width, trigger jitter. That is the ground on which REVOX's "battle for light" is fought.

What conditions does a claim like this rest on?

It is worth separating what a lighting supplier can demonstrate from what a vision integrator measures in the field. A vendor can characterize a source's output stability, lifetime derating and spatial distribution on an optical bench. Whether that translates into a stable inspection result depends on the target's reflectance, ambient light rejection, part presentation and the mounting constraints of the cell.

VISION, held at Messe Stuttgart, is where these questions typically get settled in person — on demonstration stands, against actual samples brought by prospective customers rather than against datasheets. Exhibitors in the lighting segment historically compete on the breadth of geometries and spectra offered, controller integration, and the mechanical packaging that lets a source survive washdown environments or fit inside a constrained optical path.

For REVOX specifically, the significance of a dedicated platform at VISION 2026 is that lighting is being presented as a system-level decision rather than a catalogue lookup. That framing shifts the buying conversation from "which ring light fits this lens thread" to "which illumination geometry resolves the contrast this inspection requires."

Where does this leave the integrator?

The development raises a practical adoption question for anyone specifying inspection hardware in 2026: at what stage of the design cycle does lighting enter the specification? Teams that select cameras first and retrofit illumination late routinely pay for it in rework, rejected frames and overstated system performance. Teams that characterise the optical signature of the part first — its reflectance, geometry and motion — tend to find that the sensor choice becomes the easier decision.

REVOX's message at VISION 2026, reduced to its core, is that the industry should invert its habitual order of specification. Whether visitors to the Stuttgart fair accept that argument, and whether competing lighting vendors converge on the same framing, will say something about how machine vision bill of materials are built over the next procurement cycle.