Back to Search Document details
40th Meeting: Geneva, CH, October 2025 2025-08-21 22:11
AHG18: On test conditions for ultra-low latency and packet loss resilience with full QP range
Abstract
In 39th JVET meeting several comments and suggestions concerning ULL CTC were discussed, mainly based on contributions JVET-AM0109, JVET-AM0202, JVET-AM0204, JVET-AM0222, JVET-AM0325. This contribution follows up this discussion and addresses the comments in aspects of LDP/LDB configuration, broader QP range, spatial scalability, new packet trace files and automatic bandwidth rate points selection for matching channel capacity to video bitrate.
JVET-AN0046 AHG18: On test conditions for ultra-low latency and packet loss resilience with full QP range [S. Ikonin, I. Gribushin, V. Khamidullin, B. Shevchenko, E. Alshina (Huawei)]

In 39th JVET meeting several comments and suggestions concerning ULL CTC were discussed, mainly based on contributions JVET-AM0109, JVET-AM0202, JVET-AM0204, JVET-AM0222, JVET-AM0325. This contribution follows up this discussion and addresses the comments in aspects of LDP/LDB configuration, broader QP range, spatial scalability, new packet trace files and automatic bandwidth rate points selection for matching channel capacity to video bitrate.

Was already presented in AHG meeting, but discussed again.

The basic problem that this contribution tries to address is the fact that it is tried to simulate cases where errors occur, and if the same network model would be used for QP22 and QP40, the latter would be less affected by errors due to lower rate. It is noted that a similar effect is used when using spatial scalability where the base layer has lower rate and is prioritized

One comment was made that it is difficult to judge the appropriateness of the error patterns suggested for the different QP values. It was also pointed out that it might be useful to look into similar investigations made in other bodies such as 3GPP.

Further study in AHG.

The basic problem that this contribution tries to address is the fact that it is tried to simulate cases where errors occur, and if the same network model would be used for QP22 and QP40, the latter would be less affected by errors due to lower rate. It is noted that a similar effect is used when using spatial scalability where the base layer has lower rate and is prioritized

One comment was made that it is difficult to judge the appropriateness of the error patterns suggested for the different QP values. It was also pointed out that it might be useful to look into similar investigations made in other bodies such as 3GPP.

The contribution also includes the software patch for spatial scalability which requires further confirmation by an independent party.

Decisions
The contribution also includes the software patch for spatial scalability which requires further confirmation by an independent party.
Citation