Starting in 2012 a new inharmonicity model is used in TuneLab software. This model is required if you want to employ split-scale tuning. It takes better advantage of a larger number of inharmonicity readings. Unless you specifically set the "Use old IH" switch in Settings, TuneLab will use the new model.

The old inharmonicity model in use since 2010 is a four-parameter formula whose graph looks something like this:

The red dots represent actual inharmonicity measurements that you make and the curve represents the assumed inharmonicity that the software derives from those measurements. The curve is a best-fit four-parameter curve that is as close as possible to the red dots, but it is constrained by the fact that it only has four adjustable parameters, so it cannot always coincide with the measurement points. Using the same inharmonicity readings the new inharmonicity model would look something like this:

The graph now hits every red dot exactly. Therefore the new inharmonicity model more exactly follows the actual measurements, which are connected by smooth exponential functions. There is no best-fit approximation involved, so the new model can benefit from a larger number of measurements. Unless you have a strong reason for using the old inharmonicity model, we recommend that you use the new inharmonicity model.

Because the new inharmonicity model follows actual measurements more precisely, the deviation curve on the tuning curve adjuster page may appear more irregular with the new model. However the irregular appearance more accurately represents the actual characteristics of the piano and therefore is more desirable.