For the evaluation of virtual acoustics for mixed realities, we distinguish between the paradigms 'authenticity', 'plausibility' and 'transfer-plausibility'. In the case of authenticity, discrimination tasks between real sound sources and virtual renderings presented over headphones are performed, whereas in case of a plausibility experiment, listeners need to rely only on their expectation of a sound when listening to the rendering, without the presence of an explicit reference. In the case of transfer-plausibility, however, different real sources are active alongside virtual sources, potentially in different spatial locations, leading to a certain degree of comparability. This resembles the case of forthcoming augmented reality systems. Here, we show an experiment, which assesses the transfer-plausibility of rendered speech sources in a variable acoustic environment. We demonstrate the influence of the similarity between real and virtual source material and their spatial location on the transfer-plausibility of measurement-based headphone rendering.