<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.0 TRANSITIONAL//EN"> <HTML> <HEAD> <META HTTP-EQUIV="Content-Type" CONTENT="text/html; CHARSET=UTF-8"> <META NAME="GENERATOR" CONTENT="GtkHTML/4.6.6"> </HEAD> <BODY> Le vendredi 21 mars 2014 à 03:38 -0700, Jaya a écrit : <BLOCKQUOTE TYPE=CITE> Should i use only vaapisink after the vaapidecode? is there a way to access the decoded data ? </BLOCKQUOTE> libVA does not decode into system memory, hence will not work with the pipeline you proposed. There used to be an vaapidownload element for that, but I'm not sure if this is still the case. Note that by writing raw images to file you'll loose all the image information like width, height, stride, offsets, etc. Nicolas </BODY> </HTML>