Does the high fidelity sound reproduction system itself provides desired hearing conditions in a listening room? That is only a necessary condition, because sound fields in a recording room and/or a listening room affect on subjective preference. Results of subjective preference judgements of the synthesized sound fields reveal following optimal objectives, namely “Principle of Desire” of listeners [Y. Ando, J. Acoust. Soc. Am. 62, 1436–1441 (1977) and Y. Ando, 95th Meeting of Acoust. Soc. Am., Providence, paper JJI]: (1) Sound fields with a smaller value of the interaural crosscorrelation are always judged as preferred (Incoherence at the Two Ears). In order to obtain the small value of interaural cross correlation, the preferred direction of reflections is found in the range centered on ±55° in the azimuth to the listener. (2) The discrete reflection arriving after the direct sound is preferred, so that the initial time delay gap can be determined by the coherence of autocorrelation function of source signals and the level of reflections (Renewal of the Direct Sound). For example, a preferred scale factor for the dimensions of rooms exists for a given music source. Taking these points into consideration, it is possible to design a new sound system from recording to reproducing.