By looking at the amplitude of the interference pattern we can deduce the quality of the measurement, and the phase give us the time difference. Here, we can only deduce the time difference in fractions of a wavelength (at this frequency, traveling at the speed of light). Understandably, the precision is very high, but also, the prediction of the time difference before the actual measurement needs to be precise to within one wavelength. The prediction is done by modeling the various variables (e.g. the rotation and movement of the earth, the atmospheric delay, etc), and the measurement serves to update and correct the model. This way the model and the measurements are interdependent, in a continuously refining and readjusting relationship.
The calculations of amp/phase relationships for each pair of antennas are used in the sound installation, where amplitude of measurement is mapped to amplitude of a sound, and the phase of measurement is mapped to the positioning of this sound. Each pair of antennas in an experiment is mapped to a specific pitch, where the actual choice of pitch has been done “according to taste” by me. I use several different pitch sets over the course of the installation running time, to create larger variation in the sound. Here is an example of the amp/phase data set being mapped to the pitches of a major scale: