JVET-AM0017 JVET AHG report: Testing of video coding technology beyond CTC (AHG17) [J.-R. Ohm, M. Wien (co-chairs), M. Abdoli, E. Alshina, V. Baroncini, J. Chen, R. Chernyak, Z. Deng, P. de Lagrange, L. Li, D. Rusanovskyy (vice chairs)]
Activities
The draft CfE document JVET-AL2026 was finalized and uploaded to the JVET document site. The test sequences for all categories defined in the CfE document were collected in a dedicated folder on the JVET content server.
Four online AHG calls were held on 2025-04-15, 205-05-06, 2025-06-04 and 2025-06-17, respectively. The reports of these meetings are available in JVET-AM0041.
Preparation of bitstreams according to the draft CfE conditions
The categories of the draft CfE include:
- SDR RA UHD/4K: Representing the use case of distribution of standard dynamic range UHD/4K video content e.g. in a streaming scenario, using a random-access configuration.
- SDR RA HD: Representing the use case of distribution of standard dynamic range HD video content e.g. in a streaming scenario, using a random-access configuration.
- SDR LB HD: Representing the use case of conversational and other low delay applications at HD resolution, correspondingly using a low-delay configuration.
- HDR RA 4K: Representing the use case of distribution of high dynamic range UHD/4K video content e.g. in a streaming scenario, using a random-access configuration.
- HDR RA Cropped 8K: Representing the use case of distribution of high dynamic range 8K video content e.g. in a streaming scenario, using a random-access configuration. In order to reduce the encoding workload for assessment of this category and allow investigation on 4K displays, cropped regions of 3840×2160 resolution are used.
- Gaming LB HD: Representing the use case of online gaming with a low-delay configuration.
- UGC RA: Representing the use case of user generated content at 1080×1920 or 1920x1080 resolution using a random-access configuration.
For all categories, test sequences and corresponding target rates are specified in JVET-AL2026. VTM anchor bitstreams as well as ECM bitstreams matching the target bitrates for study of improved compression performance were generated. Furthermore, for the study of reduced complexity encoding, bitstreams for the following encoder configurations were generated:
- VTM RT1: VTM with reduced runtime setting at about 65% of the runtime of the default VTM configuration;
- VTM RT2: VTM with reduced runtime setting at about 50% of the runtime of the default VTM configuration;
- VTM RT3: VTM with reduced runtime setting at about 25% of the runtime of the default VTM configuration;
- ECM 5x: ECM with a reduced runtime setting at about a factor of 5 of the runtime of the default VTM configuration.
The bitrates and indicative PSNR figures are provided in the accompanying Excel sheet.
Related contributions
JVET AHG report: Testing of video coding technology beyond CTC (AHG17) | J.-R. Ohm, M. Wien (co-chairs), M. Abdoli, E. Alshina, V. Baroncini, J. Chen, R. Chernyak, Z. Deng, P. de Lagrange, L. Li, D. Rusanovskyy (vice chairs) | |
AHG17: AhG meeting notes | ||
AHG17: Generated VTM anchor bitstreams using LambdaScaleTowardsNextQP | ||
[AHG17] Suggested results reporting template for constrained encoder complexity category | ||
AHG17: Suggestion to enable DMVR encoder control for VTM anchor | ||
AHG17: Analysis on encoding time and mode cost test for draft CfE sequences | ||
On Intra Period of Random Access Test Case for HD and below Content | X. Li (Google), L. Zhang (ByteDance), X. Wang (Kwai), S. Liu (Tencent), A. Duenas (Warner Bros. Discovery), A. Norkin (Netflix), W. Zhang (Disney), Y. Ye (Alibaba), A. Segall (Amazon) | |
On Analysing and Reporting Encoding and Decoding Complexity | ||
AHG17: Getting VTM to run five times faster | ||
[AHG17] Encoder runtime for the constrained complexity configurations under the CfE draft test conditions | ||
[AHG17] Overview of the 3GPP codec testing in TR 26.955 | ||
On constrained encoding and decoding experiments | L. Li, M. W. Park, M. Park, Y. Kim, M. Budagavi, K. P. Choi (Samsung) | |
Suggestions on GOP Size Setting of Random Access Configuration for Live-streaming Applications | Y. Wu, Y. He, K. Zhang, L. Zhang (Bytedance), X. Li (Google), S. Liu (Tencent), W. Zhang (Disney) | |
AHG17 ECM/VTM performance under CfE requirement | K. Naser, E. François, F. LeLéannec, F. Galpin, T. Poirier (InterDigital) | |
[AHG17] On HDR coding and metrics | D. Rusanovskyy (Nokia), E. François (InterDigital) |
Recommendations
The AHG recommended:
- To progress the work in AHG4 towards finalizing the draft test set for a Joint Call for Evidence on video compression with capability beyond VVC.
- To conduct subjective tests assessing the proposed rate points and coding conditions in the draft CfE.
- To consider inclusion of bitstreams of JVET-AM0045 and potentially also JVET-AM0078 in the subjective evaluation.
- To review the remaining input contributions related to AHG17 during the meeting.
- To finalize and publish the CfE document.
Potentially more ACR test cases are possible from JVET-AM0045, JVET-AM0078 and JVET-AM0200. Mathias was asked to clarify offline with proponents, and present plans on Friday at 0900 (first item of 2nd day).