Object-sided telecentry

In most applications, the lenses are only telecentric towards the test object as the emphasis is on the correct capture of the part. An object-sided telecentric optical path is generated by means of a convergent lens in a diaphragm in the image-sided focal plane. A lens of this design is normally used in the applications described above: by means of object-sided telecentry, the perspective error can be avoided and parts can be measured to scale even if the working distance varies.

Image-sided telecentry

Quite a number of lenses is designed in such a way that the optical path is only parallelised towards the camera sensor, i.e. the focal point is in the infinite. An image-sided telecentric optical path is generated by means of a convergent lens in a diaphragm in the object-sided focal plane. The advantage of this design is the uniform vertical exposure of the camera sensor. Especially the interline transfer CCDs common today have micro-lenses attached to the individual sensor pixels in order to maximise the quantum efficiency. If the light hits these micro-lenses at an angle especially on the sensor edge, massive vignetting effects appear there. Without image-sided telecentry, the image would be strongly dimmed on the margin. The larger the sensor chip in proportion to the lens mount, the more often this effect can be observed. In this case, image-sided telecentric optics should be used.

Double-sided telecentry

If both methods are combined, lenses with double-sided telecentry can be designed. They have an excellent quality regarding optical errors, however, their production is expensive. These lenses typically have a particularly high object-sided depth of field. Furthermore these optics do theoretically not cause any geometric imaging errors, as the optical path is almost perfectly parallel both on the object and the image side.