- Timestamp:
-
11 Aug 2015, 14:05:59 (9 years ago)
- Author:
-
ksuehring
- Comment:
-
Split into two reports.
Gerhard told me, that he can confirm the issue remaining in this report.
Legend:
- Unmodified
- Added
- Removed
- Modified
-
-
Property
Summary
changed from
Two potential HTM14.0 reference model bugs
to
layer_set_idx_for_ols_minus1[i] code length calculation diff between spec description and HTM code implementation
-
initial
|
v1
|
|
1 | | '''Potential Bug1:''' |
2 | | layer_set_idx_for_ols_minus1[i] code length calculation diff between spec description and HTM code implementation |
3 | | |
4 | | Bug description: |
5 | | This syntax use u(v) to code. |
6 | | Spec description: code length should be Ceil(Log2(NumLayerSets - 1)) |
7 | | HTM implementation: Ceil(Log2(NumLayerSets )) |
| 1 | This syntax use u(v) to code. |
| 2 | - Spec description: code length should be Ceil(Log2(NumLayerSets - 1)) |
| 3 | - HTM implementation: Ceil(Log2(NumLayerSets )) |
| 4 | |
8 | 5 | Influence: |
9 | | VPS high level syntax |
10 | | We can always handle this potential problem due to SW implementation, no HW related. |
| 6 | - VPS high level syntax |
11 | 7 | |
12 | | |
13 | | |
14 | | |
15 | | |
16 | | '''Potential Bug2:''' |
17 | | • Inter-view bi-prediction mismatch with identicalMotion Mismatch case conditions: |
18 | | V2 are bi-predicted from V0 and V1. (V0->V2<-V1) |
19 | | No Weighted Bi prediction is used. |
20 | | Forward MV0 is equal to backward MV1. |
21 | | RefPOCL0 is equal to RefPOCL1 |
22 | | Influence: |
23 | | 2 views beyond (such as 3 views) |
24 | | If we only target 2 views, we can ignore this potential bug. |
| 8 | We can always handle this potential problem due to SW implementation, no HW related. |
| This list contains all users that will be notified about changes made to this ticket. These roles will be notified: Reporter, Owner, Subscriber, Participant - Derek Chang(Reporter)
- Gerhard Tech(Owner, Subscriber, Participant, Always)
- jct-3v@…(Subscriber)
- Karsten Suehring(Participant, Always)
|