When analyzing the emission of a large number of individual chromophores embedded in a matrix, the spread of the observed parameters is a characteristic property for the particular chromophore-matrix system. To quantitatively assess the influence of the matrix on the single molecule emission parameters, it is imperative to have a system with a well-defined chromophore nanoenvironment and the possibility to alter these surroundings in a precisely controlled way. Such a system is available in the form of the visible fluorescent proteins, where the chromophore nanoenvironment is defined by the specific protein sequence. We analyze the influence of the chromophore embedding within this defined protein environment on the distribution of the emission maximum wavelength for a number of variants of the fluorescent protein DsRed, and show that this parameter is characteristic of the chromophore-protein matrix combination and largely independent of experimental conditions. We observe that the chemical changes in the vicinity of the chromophore of different variants do not account for the different distributions of emission maximum positions but that the flexibility of the chromophore surrounding has a dominant role in determining the distribution. We find, surprisingly, that the more rigid the chromophore surrounding, the broader the distribution of observed maximum positions. We hypothesize that, after a thermally induced reorientation in the chromophore surrounding, a more flexible system can easily return to its energetic minimum position by fast reorientation, while in more rigid systems the return to the energetic minimum occurs in a stepwise fashion, leading to the broader distribution observed.