Opened 13 years ago

Closed 9 years ago

#194 closed defect (fixed)

significance bit estimation in RDOQ (PCP_SIGMAP_SIMPLE_LAST defined)

Reported by: jzan Owned by:
Priority: minor Milestone:
Component: HM Version: HM-3.3
Keywords: Cc: fbossen, ksuehring, davidf, jct-vc@…


When PCP_SIGMAP_SIMPLE_LAST is defined (as default in HM-3.3 and HM-3.4), the last coefficient position is encoded before the significance map.

During significance map coding, the position of the last significant coefficient is indicated by the already encoded LAST position, and thus not encoded.

However, in the current RDOQ process, this is not taken into consideration: the significance map rate of the last significant coefficient is calculated and included into its RD cost, resulting in inaccurate estimation.

Attachments (1) (21.8 KB) - added by jzan 13 years ago.
Codes in FIX_RDOQ_SIGBIT might fix this

Download all attachments as: .zip

Change History (6)

Changed 13 years ago by jzan

Codes in FIX_RDOQ_SIGBIT might fix this

comment:1 Changed 12 years ago by davidf

  • Component set to HM

Updating component after adding WD (Text) tickets

comment:2 Changed 12 years ago by davidf

  • Cc fbossen ksuehring davidf added

comment:3 Changed 12 years ago by davidf

  • Cc jct-vc@… added

comment:4 Changed 10 years ago by ksuehring

  • Milestone HM-3.3 deleted

comment:5 Changed 9 years ago by karlsharman

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

This issue appears to have been fixed at some point in the
In the current software, the cost of coding coefficients
is calculated first, and then the xGetRateLast is called
on line 2496 of TComTrQuant.cpp, removing the cost of
coding the sigmap when working out totalCost on line 2497.

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

  • David Flynn(Subscriber, Participant)
  • Frank Bossen(Subscriber)
  • jct-vc@…(Subscriber)
  • Jinwen Zan(Reporter)
  • Karl Sharman(Participant)
  • karl.sharman@…(Always)
  • Karsten Suehring(Subscriber, Participant, Always)