Opened 12 years ago

Closed 12 years ago

#606 closed defect (fixed)

Weighted prediction process needs major revision

Reported by: bbross Owned by: bbross
Priority: major Milestone: D7
Component: Text Version: D7 (I1003) d5
Keywords: Cc: bbross, wjhan, jct-vc@…


The following issues are present in the WP process

  1. The input "bit-depth of the chroma component, bitDepth" is wrong for luma invocation.
  1. Variables shift1, shift2, offset1 and offset2 are derived in but used in Default weighted sample prediction process
  1. The whole process is inconsistent w.r.t. luma/chroma in Weighted sample prediction process:
  • there is a "with H being replaced by Y for luma samples and by C for chroma samples". How do we know that is is luma or chroma?
  • For the derivation of logWDC, o0C, o1C, and w0C, w1C, "If C is equal to L"... Where comes C from?
  • Why is BitDepthC and BitDepthY used in this derivation although corresponding bitDepth is input parameter and set to the global variables when invoked?
  • Wouldn't it be better to pass a cIdx and derive variables depending on that variable?
  1. logWDC should be log2WDC and derivation of these variable should be moved at the beginning (see #433)
  1. Invocation of the WP process for chroma should be split into two paragraphs as it is done for other processes.
  1. Minor improvement: Since it is not used somewhere else, "the same as specified in sub clause" should be replaced by the actual inputs and outputs. (see #433)

Who is going to fix this?

Change History (2)

comment:1 Changed 12 years ago by DefaultCC Plugin

  • Cc bbross wjhan jct-vc@… added

comment:2 Changed 12 years ago by bbross

  • Resolution set to fixed
  • Status changed from new to closed

Fixed in JCTVC-I1003_d8.

Note: See TracTickets for help on using tickets.

This list contains all users that will be notified about changes made to this ticket.

These roles will be notified: Reporter, Owner, Subscriber, Participant

  • Benjamin Bross(Reporter, Owner, Subscriber, Participant)
  • jct-vc@…(Subscriber)
  • karl.sharman@…(Always)
  • Karsten Suehring(Always)
  • Woo-Jin Han(Subscriber)