In 1900 Hartmann proposed a test for optical controlthat has since been used for a large range of applications.1 As an example, 100 years after Hartmann’sstudy, the primary mirror of the Zelentchouk gianttelescope has been tested during its manufacturewith a setup close to that described by Hartmann.2The success of this technique is probably due to itssimplicity. It uses a mask of holes, usually arrangedin a regular square grid, placed in the plane of analysis.These holes break the incoming light intobeams, which are deflected according to the local distortionsof the sensed wave front. If observed in aplane at a certain distance from the mask, the grid istherefore warped, with the displacement of each spotdirectly proportional to the local tilt of the wave frontand the distance between the mask and the observationplane ~see Fig. 1!. The sensitivity and the dynamicsof this test can thus be adjusted continuouslyby simple longitudinal translation of the observationplane. Because this operating mode is based on raytracing, the test is achromatic.A number of innovative variations to the classicaltest have been suggested, to adapt it to specific wavefrontsensing experiments.3 For example, Shackproposed replacing the mask of holes with a grid ofmicrolenses.4 With this modification, all the lightcoming from the analyzed source is focused at thecommon focal plane of the microlenses, allowing forbetter light efficiency. This property is of particularimportance for astronomical applications.5,6 However,the sensitivity is no longer adjustable, since themeasurement plane is restricted to the focal plane ofthe microlenses.In the classical Hartmann test, although the sensitivityis continuously adjustable, limitations appearwhen the observation plane is placed far from themask, owing to diffraction effects. The ray-tracingapproach is then no longer valid, and the hole grid isblurred. Our purpose in this paper is to present anew arrangement of the Hartmann test that allowssensitivity for a increase by minimization of blurringeffects that are due to diffraction. In Section 2 a descriptionof the classic Hartmann test is presented,issued from a study of the Talbot phenomenon. Thenwe present the theory of this extended Hartmann testin Section 3. Finally, we conclude this paper in Section4 by showing experimental results that give whatto our knowledge is the first evidence of the ability ofthis new Hartmann test.