In most tests on audio quality, a "low quality" anchor is provided concurrently with the sequences to be evaluated. A classic generic anchor consists of a low-pass filtered version of the audio reference; but evaluation of specific sound properties (for example spatial features) could require a specific anchor. In a first experiment, we evaluated five spatial anchors to find which was the most appropriate for a test including evaluation of Space quality. One of the tested anchors obtained low scores for all the excerpts used. In a second experiment, a copy of this anchor was integrated into an evaluation test of audio codecs, with assessment of the categories Timbre, Defects, and Space. The scores obtained in this test indicated that this anchor fulfilled the required criteria (a low score for the anchor over all excerpts, the correct range of codec scores for Space, and no serious degradation for Timbre and Defects).