To leave the path of classic holography and limit the space-bandwidth-product of the holographic reconstruction is one way to enable interactive real time holographic 3D displays. Thus, a couple of major problems - among several others - can be reduced to a practical level. This holds e.g. for the computation power, the data transfer rate and the pixel count of the spatial light modulator (SLM) used. Although this idea is almost twenty years old, the maximum time span of IP protection, displays based on space-bandwidth-limited CGH reconstruction, which also can be referred to as spacebandwidth- limited reconstruction of wave front segments, are still not on the market. There are several technological reasons for that. However, the technological barriers can be tackled gradually. One problem to be solved is the illumination of the entrance plane of the preferable complex valued spatial light modulator (CSLM). Here, CSLM means to modulate the phase and the amplitude of each pixel. The display diagonals of desktop and TV type CSLM might be e.g. 32 and 65 inch respectively. In other words, reasonable large collimated illumination wave fields are mandatory. In addition a small form factor is a must have in order to obtain commercial success. The solution is an optical system design, which is based on Bragg diffraction based volume gratings. Classic refractive optics fails here. In other words, Bragg diffraction based volume gratings are key components of illumination units of holographic 3D displays. They can be used within a parameter space, which cannot be addressed by surface relief type diffraction optics. But their layout depends on the parameters of the illumination wave field, which has to be tailored in regards to the optical system of the discrete, e.g. 1D or 2D encoded holographic 3D display. This will be described in more detail. The example used for the description is a double wedge type backlight unit. Furthermore, it will be explained why the use of complex valued secondary light sources is a must have in holographic 3D displays. For this, a short explanation of coherent retinal inter object point cross talk will be given too. Finally, the description of the wave field shaping (WFS), which is required in order to form the optimized complex valued light source planes, is provided. In other words, a description of a tailored coherence preparation is given, which is up to now not state of the art. The cause and effect relationship of the light propagating from the primary light sources, which are lasers, to the final receptor, which is the retina, will be pointed out. Although this tailored partial coherent illumination totally differs from the state of the art of information displays, it might help to understand a technology, which will come in the next decades.