Took at a look at the timer manager, looks like our timers are smarter than I thought. They use a running absolute expire time that they push forward the (number of calls this frame) * the tick rate. So you’re not going to get the kind of accumulated error I’m talking about. Floating point error isn’t really the issue. Not all timer systems do what ours does, some use a much more simplified approach. They may just have a time in the future, and when that time becomes the past, they execute the callback. Then they set a new time that’s Rate + current time in the future, for the next call, losing time because they don’t account for the overshoot. Point is, not all callback systems keep perfect time, so never hurts to plan accordingly.