The second line of evidence I have to present is only slightly more substantial that the quote above. While this time the U.S. patent office is involved, no results are available for inspection and one can assume a rosy bias on the effectiveness of anything submitted for patent. Nonetheless, in 1969 Charles Weller of Shell Oil Company filed for patent that describes a method of acquiring seismic data:
[s]eismic exploration is conducted with out using a seismic sound source by recording a plurality of relatively long stretches of ambient earth noise data at each of an array of seismic receiving stations...
The application covers the acquisition of data in any of several areal strategies on the surface of the earth. However, a linear array of six recording stations every 200 feet having returned[m]eaningful alignment of events ... corresponding to known geologic boundaries ... obtained to beyond 5 seconds delay time, involved a linear array of six recording stations every 200 feet. The application goes on to describe more specific issues relating to the successful acquisition and processing of the data set. The most interesting of these points is the first mention of an empirical understanding of just how long the records need to be. Weller makes the claim that approximately eight hours recording in the quiet Gulf coast region was 'satisfactory' for his images, and further that one to two hours was 'sufficient'.