It should be obvious that, in its wavelength reflection ability, the grating approximates a simple (slightly rough) mirror when its front surface is vertical. When its front surface is not vertical, then the difference in optical paths between adjacent facets is d[sin(thetar)-sin(thetai)] where d is the facet spacing, thetar is the angle of reflection with respect to the normal of the grating's front surface median plane and thetai is the incident ray angle with respect to that same normal. Since both of these angles are much less than pi/2, we can simplify the path difference to dL=d(thetar-thetai) and that too can be simplified to read dL=2d*thetaG where thetaG is the counter-clockwise measured tilt angle of the grating front surface. The first diffraction order wavelength that is "seen" by the detector is then just this path difference. Second and higher order diffractions will be much weaker than the first because the blaze angle is usually chosen to be close to maximum efficiency for the first order diffraction. Instead of drawing waves as sinusoids, I have drawn just the line through the phase of maximum intensity which I call the wavefront. I should point out that I have chosen to show just the colors of the waves that would sensed through a very small aperture placed in front of the detector.