Let's call the frame of reference located at the surface of the earth O, and the one at a height above the earth's surface O'. Both observers agree to record the number of observed sunsets over a particular interval of time. Over some interval of time, O records N sunsets for a total elapsed proper time of τ. O' measures N' sunsets over a proper time τ'. We'll say that the day has a length d in the coordinates of O, and d' in the coordinates of O'. Then, the number of sunsets recorded in O is N = τ/d, and in O', N' = τ'/d'. For simplicity, we'll represent the effect of gravitational time dilation by ξ, so that τ' = ξτ, and d' = ξd. Then, the total number of recorded sunsets in O' is N' = τ'/d' = ξτ/ξd = τ/d = N. So, they record the same number of sunsets. Why? Because, ultimately, they make their observations over different intervals of time (because of gravitational time dilation), and the length of the day also varies (because of gravitational time dilation). Janus accurately stated this.