ITU-T J.249 PDF

This RRNR-TV (reduced reference and no reference objective models for standard definition television) test addresses two video formats (525-line and 625-line). Reduced reference (RR) models have limited bandwidth access to the source video. For an RR model to operate correctly, the unimpaired source video should be available for the models to extract parameters.

The hypothetical reference circuits (HRCs) in each experiment spanned both coding-only artefacts and coding with transmission errors. The coding schemes examined were MPEG-2 and ITU-T H.264 (MPEG-4 part 10). The MPEG-2 encoders were run at a variety of bit rates from 1.0 to 5.5 Mbit/s. The ITU-T H.264 encoders were run at a variety of bit rates ranging from 1.0 to 3.98 Mbit/s. Each experiment included 12 source sequences, of which two were secret source. Each experiment included 34 HRCs, and 156 processed video sequences (PVSs). Of these PVSs, 40 contained transmission errors and 116 contained coding only.

Although the models defined in this Recommendation may be used to compare video system performance (e.g., comparing two codecs), readers should be aware that the models contained in this Recommendation were not validated for such applications.

None of the evaluated models reached the accuracy of the normative subjective testing.