as some of you may know i have been on a marathon test encoding session for a while, creating batch encodes of dozens of jobs at a time. in testing various parameters within x264, something occurred to me, is the motion estimation setting at odds with the sub motion estimation setting?
consider the entries from the x264 wiki:
"Set the full-pixel motion estimation method."
"Set the subpixel estimation complexity."
0. fullpel only 1. QPel SAD 1 iteration 2. QPel SATD 2 iterations 3. HPel on MB then QPel 4. Always QPel 5. Multi QPel + bi-directional motion estimation 6. RD on I/P frames 7. RD on all frames 8. RD refinement on I/P frames 9. RD refinement on all frames, so if my reading is correct, subme=0 only uses full pixels in motion estimation, thus "me" comes into play and higher values result in more indepth motion estimation analysis.
by the same token, i would take the above entries to mean that with anything other than subme=0, quarter pixel motion estimation is used and thus "me" settings should be ignored, even if they do get written to the stream as encoding parameters.
does the above reasoning sound right to you guys?
+ Reply to Thread
Results 1 to 2 of 2
Last edited by deadrats; 13th Dec 2012 at 16:39.
no, they are complementary
--me specifices the search pattern. --subme refines the interpolation (it goes to subpixel, or "inbetween" values) unless --subme 0 is used