How to handle energy dependent deadtime/livetime?

[jknodlseder]

The deadtime (or livetime) for CTA is in principle energy dependent (as different telescopes and cameras will have different deadtimes). The question is how to handle that. Fold the deadtime into the IRF? Of having explicitly an energy dependent deadtime? This is even worse if there are different camera types on the same site, and events seen with one camera type will have a deadtime different to events seen by other camera types.