In “Post process flow with Pre processed Y as input”
mud and rain algorithms needs to be modified , RGBtoGRAY conversion has to be removed and resizing has to be done on Y cannel alone,
effect of these steps on quality is not known,
please let us know your opinion on this ?