[M4IF Technotes] contradictory conformance requirements?

[Deepali Arya]

> Hi all,
> 
> The visual conformance FDIS document has couple of contradicting
> statements. Can someone clear these up for me please?
> 
> Section 4.4 : Definition of visual decoder compliance
> The decoder shall reconstruct I-, P- and B-VOPs and sprites within +/-1
> pixel difference compared with that generated by the reference software.
> 
> 4.5.1 Static tests
> - The static tests that involve the use of IDCT, IDWT or sprite warping,
> in which case the test will check that the peak absolute error between the
> values of the samples reconstructed by the decoder under test and the
> values of the samples reconstructed by the reference decoder shall not be
> larger than 2 when the reference frames used by both decoders are
> numerically identical.
> 
> 
> 
> The first statement indicates that a decoder implementation can be off
> from the reference decoder by maximum difference of 1. The second
> statement indicates that a difference of 2 is acceptable, if reference
> VOPs are numerically equal. Which is the right requirement? If the IDCT is
> 1180 compliant and follows Annex A of the standard, how much difference
> between the final reconstructed pixels of the decoder is allowed?
> 
> Thanks.
> 
> Deepali Arya
> Improv Systems Inc, Beverly.
> Email: deepalia   improvsys.com
> Tel:  (978) 927-0555 x18
>