Performance comparison on pipelineEncode sample with InputFile.nv12 v/s InputFile.bgra
I have done time comparison on pipelineEncode sample of AMD media sdk 1.1 (amf sample) with two different Input files(1- .nv12 and 2- .bgra). I have written one time calcullation code GetTickCount() to calcullate time taken by SubmitInput() in Pipeline.cpp file InputSlot class Run function at SubmitInput() call. The cases are:
CASE 1 : Pass command line parameter as -i Input.nv12 -o Output.h264 -c exampleConfig.cfg
RESULT : Time taken by SubmitInput function is 0~15 millisecond.
CASE 2 : Pass command line parameter as -i Input.bgra -o Output.h264 -c exampleConfig.cfg
RESULT : Time taken by SubmitInput funcion is 110~125 millisecond.
Is there any way to minimise the time taken in CASE 2.