That’s true for dt. In RT it is always applied on full raw data and more of 95% of the cases using it is a no brainer