A test signal retains 75% of its previous level each cycle. You only log full cycles, so the useful answer is the first recorded value below half its starting level.

Enter 0.75 in B2 and 0.5 in C2. Use =LN(C2)/LN(B2) to locate the model crossing. It comes at about 2.41 cycles. This calculation assumes a fixed retention fraction between zero and one.

A sample model starts at full signal and retains 75 percent of the prior level per cycle.

Example model. Express the retention and target as fractions.

Check the whole cycles around the crossing

At cycle 2, the modeled level is 0.75^2, or 56.25%. At cycle 3, it is 0.75^3, about 42.19%. The first whole cycle strictly below half is 3. If another example lands exactly on half at a whole cycle, that cycle reaches the threshold and the next one is below it.

A chart shows 56.25% at cycle 2 and 42.19% at cycle 3 around the half threshold.

Example whole-cycle check. Confirm the first observed point below the target.

Compare the model with the actual signal if the retention rate can vary from cycle to cycle.

Microsoft’s LN function reference.