How could a dip be absorbed? From my experience it is not easy to absorb sound below, say 1..2kHz with just a carpet, even if it was a fat Nepalese or so. These dips are something we see with nearly every design. It is the back side of multi-way speakers, except they are coaxial and named KEF.
Hi GelbeMusik, specific directional dips are typically caused by cancellations in specific directions, due to destructive interference from differences in phase resulting from path length difference between direct and specific first reflections. You could reduce a dip by absorbing the relevant reflection, resulting in less destructive interference. Sometimes, people refer to "floor bounce" as a relative cancellation or dip resulting between 100-300 Hz from such a path length difference from the bass drivers to the floor, which typically occur in a relatively predictable fashion in most rooms (assuming that the speakers aren't elevated significantly and that the listeners are positioned within a relative distance from the speakers). Obviously, ceiling heights are much less predictable.
Floyd Toole published a relatively important book called Sound Reproduction: Loudspeakers and Rooms. If you have access to an early edition, perhaps you could look at Figure 21.3, which is captioned "At the top is shown the isolated infl uence of a good carpet underlay— 40 oz/sq yd (1.4 kg/m2) hair felt, which is typically about 0.43 in. (11 mm) thick. The same thickness of common rubber, plastic, or foam cannot deliver this level of acoustical performance. There may be other materials that are comparably good, but check for acoustical measurements. The shaded area in the bottom curve combines this underlay with different kinds of high-quality clipped-pile woven carpets (with porous backing because the sound must be allowed to penetrate into the felt underlay)." Here is a screenshot:
Since Amir was referring to the dip above 2 kHz, which is very clearly shown in his graph showing Early Reflections and which has the sentence that you reference before your comment "From my experience it is not easy to absorb sound below, say 1..2kHz with just a carpet, even if it was a fat Nepalese or so," I don't understand why you wanted to discuss your experience about absorption "below, say, 1.2 kHz with just a carpet".
Floyd Toole's book, which is worth reading if you haven't done so, also shows absorption data for a number of other materials and scenarios, which appears to be significant above 500 Hz, though decreasingly so below that, depending on the material and variations in mounting.