Use Q0 quantizers to derive lambda for Q0 pixel domain

Open Thomas Daede requested to merge ycho:lambda_fix1_for_q0 into master

Created by: yushincho

The quantizer tables that rav1e currently use are those from libaom, which are all in Q3 precision (except, qindex=0, lossless mode, which is Q2). Thus, if we derive lambda which applies to pixel domain in Q0, we should use quantizers in Q0.

Merge request reports