--qpfile should specifying QP as float not integer
Issue #372
new
--qpfile input QP data as integer but --csv output QP as two digit float
If I want to use csv file as qpfile for reencode, qpfile should accept QP as float for maximum quality.
qpFile is used to directly force the qps to each slice. There is no Rate control step. So, it will be an integer.
The outputs shown in csv is only an approximation of qp used per slice - this is because in many cases, qps are computed and set on a CU basis. Hence we average the cu level qps to display for each frame in the csv.