Custom Query (1442 matches)

Filters
 
Or
 
  
 
Columns

Show under each result:


Results (25 - 27 of 1442)

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19
Ticket Resolution Summary Owner Reporter
#594 fixed 14-bit issues in Weighted Prediction bbross pandrivon
Description

In 8.5.2.2.3 Weighted sample prediction process, variable offset1 is set equal to 1<<(shift1-1) but shift1 is set equal to 14 - bitDepth. If bitDepth equals to 14 then we have a negative shift. Basic non-invasive lightweight solution would be to set offset1 to 0 when bitDepth equals to 14. Another hardcore one would be to shift all MC (WP, IF...) precision on 15 bits instead of 14 bits that would be very invasive in the code. Btw, the same issue happens in HM7.0. (WP, IF). Any thoughts?

#595 fixed 14-bit offset issues pandrivon
Description

Offset issues w.r.t. 14-bit:

in HM7.0, TComWeightPrediction::addWeightBi()and

TComWeightPrediction::addWeightUni()appearing three times (luma and chroma) for each. { …

Int shiftNum = IF_INTERNAL_PREC - ( g_uiBitDepth + g_uiBitIncrement ); Int shift = wp0[0].shift + shiftNum; may be null Int round = (1<<(shift-1)) * bRound; beget negative shift here!'

… } and in HM7.0 TComInterpolationFilter::filterCopy() { … Int shift = IF_INTERNAL_PREC - ( g_uiBitDepth + g_uiBitIncrement ); may be null Short offset = IF_INTERNAL_OFFS + (1 << (shift - 1)); beget negative shift here!' … }

Lightweight solution is to set round and offset to 0 when bit-depth = 14... Same issue in Draft 7 d4 (Cf. Ticket #594) Any thoughts?

#478 fixed 2nd & 4th "If" statement in combine list initialization is vague and not precise. bbross tktan
Description

The language of the 2nd and 4th "if" statement in step 3 of subclause 8.3.4.3 is vague and not precise, as it is really difficult to know what "first occurance of the reference picture" actually means.

I propose editorial to change Step 3 of Subclause 8.3.4.3 as follows:

Replace – If the entry RefPicListL0[ refIdxL0 ] is the first occurance of the reference picture,

PredLCToPredLx[ refIdxLC ] = Pred_L0, (8 12)

.....

with – If the element RefPicList0[ refIdxL0 ] is not in the reference picture list RefPicList2,

RefPicList2[ refIdxL2 ] = RefPicList0[ refIdxL0 ], (8 12) PredL2ToPredLx[ refIdxL2 ] = Pred_L0,

......

And similarly, replace

– If the entry RefPicListL1[ refIdxL1 ] is the first occurance of the reference picture,

PredLCToPredLx[ refIdxLC ] = Pred_L1, (8 13)

.....

with

– If the element RefPicList1[ refIdxL1 ] is not in the reference picture list RefPicList2,

RefPicList2[ refIdxL2 ] = RefPicList1[ refIdxL1 ], (8 13) PredL2ToPredLx[ refIdxL2 ] = Pred_L1,

......

Also note that there is a defect in the List name RefPicList0 and RefPicList1, they sometimes appear as RefPicListL0 and RefPicListL1. I suggest that all occurances of RefPicListL0 and RefPicListL1, be replaced with RefPicList0 and RefPicList1, to be consistent with the AVC text.

While we are at it, we may as well change the following variable names....... as "LC" and "lc" are not good variables as they are used elsewhere for other tools.

PredLCToPredLx to PredL2ToPredLx RefIdxLCToRefIdxLx to RefIdxL2ToRefIdxLx num_ref_idx_lc_active_minus1 to num_ref_idx_l2_active_minus1 ref_pic_list_modification_flag_lc to ref_pic_list_modification_flag_l2 refIdxLC to refIdxL2 (I already used this above) mvd_lc to mvd_l2 Pred_LC to Pred_L2 mvp_lc_flag to mvp_l2_flag luma_weight_lc_flag to luma_weight_l2_flag delta_luma_weight_lc to delta_luma_weight_l2 luma_offset_lc to luma_offset_l2 chroma_weight_lc_flag to chroma_weight_l2_flag delta_chroma_weight_lc to delta_chroma_weight_l2 delta_chroma_offset_lc to delta_chroma_offset_l2

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19
Note: See TracQuery for help on using queries.