further clean up of errorperbit and sadperbit
authorYaowu Xu <yaowu@google.com>
Wed, 1 Jun 2011 18:41:05 +0000 (11:41 -0700)
committerYaowu Xu <yaowu@google.com>
Wed, 1 Jun 2011 21:44:06 +0000 (14:44 -0700)
commit5b2fb3296167da69ab4ad6df2e24b0b87fa1e515
treeeae38145c6c4ef46ada3cf6028722ac617ba69d9
parent50916c6a7da357eeee43c36b4d6ab6a723ddbe1e
further clean up of errorperbit and sadperbit

this commit makes the usage errorperbit and sadperbit consistent for
encoding modes and passes. Removed all different magic weight factors
associated with errorperbit. Now 1/2 is used for both sadperbit16 and
sadperbit4, the /2 operation is merged into initializations of the 2
variables.

Tests on cif set show .23%, 0.18% and 0.19% gain by avg psnr, overall
psnr and ssim respectively.

Change-Id: Ifa285c3e065ce0a5a77addfc9f95aabf54ee270d
vp8/encoder/firstpass.c
vp8/encoder/mcomp.c
vp8/encoder/pickinter.c
vp8/encoder/rdopt.c
vp8/encoder/temporal_filter.c