Depth of Field and Aperture

Depth of field is the distance range in front of and behind the plane of focus over which the image is sharp enough to use. It is not a yes-or-no boundary — sharpness falls off gradually — so the range depends on what "sharp enough" means for your measurement. A station measuring the position of a whole part can tolerate more blur than one measuring the width of a 0.1 mm gap, and both are described by the same lens at the same aperture. The number only becomes meaningful once the tolerance is fixed.

What sets the range

Four things decide how much range you get. The aperture is the one you control directly: a larger f-number means a smaller opening, which makes the cone of light arriving at each pixel narrower, which means a wider range of object distances still land inside it. Magnification is the second: the more the lens magnifies, the narrower the cone and the shallower the range. Pixel size is the third, and it is easy to overlook — the same lens on a sensor with smaller pixels resolves more detail and shows blur sooner, so the usable range shrinks. Working distance is the fourth: the same field of view from further away gives more range.

Stopping down is not free

The natural response to a part that will not stay sharp is to stop the lens down. That does widen the range, but every stop removes light, and the light has to come back from somewhere: a longer exposure, a brighter light source, or more gain. Longer exposure reintroduces the motion blur the station was designed to avoid. More gain lifts the noise floor, which eats the contrast the measurement depends on. A brighter source costs money, heat, and space on the fixture. So stopping down converts a sharpness problem into a light problem, and the light problem has to be solved too.

There is also a ceiling that catches people out: past a certain point, stopping down makes the image worse rather than better. Diffraction at a very small opening softens the whole image, not just the out-of-focus regions, so beyond roughly f/11 to f/16 on typical industrial lenses the picture gets softer everywhere. The exact point depends on the pixel size and the lens, but the direction is always the same. If the range is still not enough at that point, the answer is to change something else — a different lens, a different working distance, or a different measurement strategy — not to stop down further.

Estimating the range you actually need

The range you need comes from the process, not from the lens catalogue. Start with the flatness of the part, add the variation in how it sits in the fixture, add the fixture's own repeatability, and add whatever the conveyor or robot contributes. On a flat part held against a locating face, that total may be a few tenths of a millimetre. On a stamped part dropped onto a nest, it can be several millimetres. The number you get is the range the station has to cover, and it is a requirement to write down rather than a detail to discover later.

Two refinements are worth the effort. First, only the measurement features need to be sharp. If the station measures a bore on one face, the rest of the part can fall out of focus without consequence, and the range only has to cover the variation of that face. Second, sharpness requirements differ across a feature: locating the centre of a circular hole tolerates more blur than measuring its diameter to a tight tolerance, because the centre moves much less than the edge under blur. Spending the depth of field only where it is needed often makes an impossible requirement possible.