vector<Double32_t> specifying precision

ROOT
1

Does the syntax

have a parallel for vectors of Double32_t? Can I specify that the on-disk values should be (eg.) 12-bit integers, just like for a Double32_t member?