Unisons - intentionally mistuning

It is commonly believed that the unisons for each note should be tuned to the same frequency, with zero beating. In fact this is standard procedure for many piano manufacturers. In contrast, the best tuning I have ever heard was done by a master tuner at a school of music who claimed that the unisons should never be tuned to exactly the same frequency, they have to be slightly off. He claimed that intentionally mistuning the three strings on a three string note leads to a richer sound with longer sustain.

This has been documented in the professional literature with careful quantitative adjustments and blind listening tests. Roger Kirk (Baldwin Piano Company) compared the preferences of a group of listeners (profession, trained, and untrained) for the sound of a piano with unisons containing differences of 0, 0.5, 1, 2, and 3 cents. There was a strong preference for a 1 to 2 cent deviation. More information can be found in the original paper:

Roger E. Kirk "Tuning Preferences for Piano Unison Groups", The Journal of the Acoustical Society of America, 31, 1644-8 (1959).

A detailed analysis of the physical basis for the phenomenon has been given in

Gabriel Weinreich "Coupled Piano Strings", The Journal of the Acoustical Society of America, 62, 1474-84 (1974).

It would be interesting to see more work in this area.