in most cases, the baseband provides the best voice quality results, but it requires some efforts on the frequency planning.
if your cells are all the same height, and evenly distributed, the synth. hopping can give identical results as well, but with much less effort on the frequency planning.
if the network is multilayer, or irregular, it is almost impossible to achieve a good performances with synth. hopping. However, it is very simple to plan, and easy to maintain…
There is nothing much to elaborate, the best is to see it with your own eyes : perform field trials ! For instance, choose 2 cities of same cell density. On the first city, apply BBH, on the second city, apply SFH.
You will have clear results, perfectly adapted to your case and that will serve as a reference in the future for your network. Plus, it gives you more experience as a frequency planner and a qos monitor engineer 😉 it’s a win-win situation, isn’t it ?