echo
26th December 2003, 17:30
Hi everyone and Merry Christmas to you all
Does anybody know what P block masking exactly is and how it works? For instance I know that darkness masking lowers the quantizers in dark areas and lumi masking does the same with bright areas. From experience I can tell that it only works in P frames (hence the name :D ) by increasing the quantizer in some MB's and only in second pass, meaning that it does not have any effect at all in the first pass. The results I got were suprisingly good even at extreme values (=1) in quite a few movie encodings I have made with it. Overall the quality in each frame that I tested was better with P block masking on, probably because it also lowered the quantizers where it was most needed within the frames. But still I would like to know more about how this option decides about where it lowers and where it increses the quantizers. I have searched through many libavcodec help files and pages but did not find anything that would clarify things a bit.
Regards
echo
Does anybody know what P block masking exactly is and how it works? For instance I know that darkness masking lowers the quantizers in dark areas and lumi masking does the same with bright areas. From experience I can tell that it only works in P frames (hence the name :D ) by increasing the quantizer in some MB's and only in second pass, meaning that it does not have any effect at all in the first pass. The results I got were suprisingly good even at extreme values (=1) in quite a few movie encodings I have made with it. Overall the quality in each frame that I tested was better with P block masking on, probably because it also lowered the quantizers where it was most needed within the frames. But still I would like to know more about how this option decides about where it lowers and where it increses the quantizers. I have searched through many libavcodec help files and pages but did not find anything that would clarify things a bit.
Regards
echo