Video Area Home >> Video-Codec Comparisons >> HEVC/AV1 Video Codecs Comparison 2019

HEVC/AV1 Video Codecs Comparison 2019

Fourteen Annual Video-Codecs Comparison by MSU

Video group head: Dr. Dmitriy Vatolin
Project head: Dr. Dmitriy Kulikov
Measurements, analysis: Dr. Mikhail Erofeev,
Anastasia Antsiferova,
Sergey Zvezdakov,
Denis Kondranin,
Egor Sklyarov,
Stanislav Grokholskiy

100 videos were used for comparison! (report part I, FullHD)
If you want to receive notifications about our reports, please

subscribe


Navigation


Participated codecs

Codec name Report parts Standard
Part I. FullHD Part II. Subjective Part III. 4K Part IV. High Quality
1 aom
AOMedia
None
None
None
YES AV1
2 arowana xvc
Divideon
YES
Ripping use case
YES None
None
xvc
3 Bytedance V265 Encoder
ByteDance Inc.
YES
Fast, Universal use cases
YES YES No HEVC
4 HW265
Huawei Technologies Co., Ltd.
YES No YES No HEVC
5 MainConcept HEVC
MainConcept GmbH
No No YES No HEVC
6 rav1e
rav1e
No No No YES AV1
7 SIF Encoder
SIF Encoder Team
YES
Ripping use case
YES YES
Universal use case
No SIF
8 SVT-AV1
Open Visual Cloud
No No YES
Universal use case
YES AV1
9 SVT-HEVC
Open Visual Cloud
No No YES YES HEVC
10 SVT-VP9
Open Visual Cloud
No No YES YES VP9
11 sz265
Nanjing Yunyan
YES YES YES No HEVC
12 Tencent V265 Encoder
Tencent
YES YES No No HEVC
13 UC265
Ucodec Inc.
YES
Ripping use case
YES No No HEVC
14 VP9
The WebM Project (Google)
YES
Ripping use case
YES YES No VP9
15 WZAurora AV1 Encoder
Visionular
YES
Ripping use case
YES YES
Universal use case
No AV1
16 x264
x264 Developer Team
YES YES YES YES AVC
17 x265
MulticoreWare, Inc.
YES YES YES YES HEVC
18 xin265
YES YES No No HEVC

Reports

Main report (Objective comparison, FullHD videos)
Three encoding use cases measured on 100 FullHD videos
Released on October, 21
Free version Enterprise version
Use cases Universal (partially) Fast, Universal, Ripping
Per-sequence-results 2 of 100 sequences (only Universal use case) All 100 sequences for all use cases (in interactive charts)
Metric: YUV-SSIM YES YES
Description of video sequences YES YES
Codec info (developer, version number, website link) YES YES
Other objective metrics (in addition to YUV-SSIM) None YES
Y-VMAF(0.6.1), Y-VMAF(0.6.2), Y-VMAF(0.6.3), Y-VMAF(0.6.1, Phone), Y-VMAF(0.6.2, Phone), Y-VMAF(0.6.3, Phone), Y-SSIM, U-SSIM, V-SSIM, YUV-PSNR, Y-PSNR, U-PSNR, V-PSNR
Per-frame metrics results (in HTML version of the report) None YES
All metrics for all sequences and use cases (10000+ charts)
Relative quality analysis None YES
Download links for video sequences None YES
Encoders presets description None YES
PDF report 58 pages 83 pages
HTML report 28 interactive charts 14000+ interactive charts
Price Free $950
Download PDF & HTML

Descriptions of 100 used videos can be found on this page or in separate PDF (41 MB)
Buy

You will receive enterprise versions of all reports (FullHD, Subjective, 4K and High Quality)

Subjective Report
Subjective comparison conducted on Subjectify.us platform
Released on November, 1
11 codecs
Bytedance, sz265, Tencent V265 Encoder, UC265, x265, xin265, arowana xvc, SIF Encoder, VP9, WZAurora AV1 Encoder, x264
Free version






Enterprise version

Buy

You will receive enterprise versions of all reports (FullHD, Subjective, 4K and High Quality)
Contact us if you want to buy only Enterprise Subjective Report
732 unique observers
25784 valid answers
5 video sequences
Short fragments from Crowd Run, Kayak Trip, Making Alcohol, Tractor, Wedding Party
Special Subjective Encoding Use Case
At least 1 FPS
6 metrics
Subjective score and 5 objective: YUV-SSIM, Y-SSIM, YUV-PSNR, Y-PSNR, Y-VMAF(v.0.6.1)
HTML and PDF documents
118 interactive charts and 33 pages

4K Report
Comparison conducted on 4K (UHD) videos
Released on March, 6
12 codecs
Bytedance V265 Encoder, HW265, MainConcept HEVC, SVT-HEVC, sz265, x265, SIF Encoder, SVT-AV1, SVT-VP9, VP9, WZAurora AV1 Encoder, x264
Free version

PDF report


HTML report


Download all in one archive (zip)


Enterprise version

Buy

You will receive enterprise versions of all reports (FullHD, Subjective, 4K and High Quality)
Contact us if you want to buy only Enterprise 4K Report
11 4K video sequences
From 24 to 60 FPS
Two Encoding Use Cases
4K Universal (1fps) and 4K Fast (20fps)
9 metrics
YUV_SSIM, YUV_PSNR, Y_VMAF (v0.6.2), Y_SSIM, U_SSIM, V_SSIM, Y_PSNR, U_PSNR, V_PSNR
HTML and PDF documents
710 interactive charts and 71 pages

Report on high quality encoding
Comparison conducted under slow-speed requireents
Released on March, 30
7 codecs
aom, rav1e, SVT-AV1, SVT-HEVC, SVT-VP9, x264, x265

Enterprise version (for free)

PDF report


HTML report


Download all files in one archive (zip)
PDF, HTML report + download links for videos + HTML report with per-frame metrics results
6 FullHD video sequences
From 24 to 60 FPS
Special Encoding Use Case
Formal limitation 0.005 fps (not strict)
9 metrics
YUV_SSIM, YUV_PSNR, Y_VMAF (v0.6.2), Y_SSIM, U_SSIM, V_SSIM, Y_PSNR, U_PSNR, V_PSNR
HTML and PDF documents
594 interactive charts and 42 pages


Ovarall Conclusions


Main report (Objective comparison, FullHD videos) summary

According to just quality scores (YUV-SSIM), the best codecs among our competitors evaluated for all three use cases are the following:

  • First place: HW265
  • Second place: Tencent V265 Encoder
  • Third place: sz265

The biggest number of codecs took part in comparison with high-quality encoding presets (Ripping use case). The winners for only high-quality encoding are the following:

  • First place in high-quality (ripping) use case: HW265
  • Second place in high-quality (ripping) use case: Tencent V265 Encoder
  • Third place in high-quality (ripping) use case: VP9

We tested three encoded use cases (see the description in section Test Hardware Characteristics). Here is one of the rate-distortions charts for universal encoding (Cion video sequence):

The universal-encoding use case has two Pareto optimal encoders in terms of mean speed and quality: HW265 and Tencent V265 Encoder. Nevertheless, the differences emerge for particular sequences and use cases.

Free version contains the results for 2 of 100 video sequences, while full results are available in enterprise version.


Subjective report summary

According to subjective quality scores, the best codecs among our competitors evaluated for all three use cases are the following:

  • First place: WZAurora AV1 Encoder
  • Second place: Tencent V265 Encoder
  • Third place: arowana xvc

There is no absolute winner in the comparison, since different encoders take first place at different test video sequences: for example, on Crowd Run (short) three encoders show Pareto-optimal results: SIF Encoder, Bytedance, WZAurora AV1 Encoder, arowana xvc.
All graphs are available in enterprise version, which is free for all buyers of enterprise main report, and enterprise main report is also free for all buyers of enterprise subjective report.

4K report summary

According to SSIM quality scores, the best codecs among our competitors evaluated for two use cases are the following:

  • First place: HW265 by Huawei Technologies Co., Ltd.
  • Second place: Bytedance V265 Encoder
  • Third place: Mainconcept HEVC
The results slightly differ for other objective quality metrics. For example, Mainconcept HEVC shows the best result according to VMAF.

Mean overall quality (SSIM):


Overall quality for universal use case (1fps, SSIM):


In this comparison, the results for universal 4K encoding (1fps) differ from overall scores. The following encoders showed the best performance results:

  • First place: WZAurora AV1 Encoder
  • Second place: HW265 by Huawei Technologies Co., Ltd.
  • Third place: Mainconcept HEVC

The above plots show only quality gain, hovewer, all encoders slightly differ in encoding speed. The following speed-quality chart shows the picture for Duck Take Off video sequence.


All graphs, including speed-quality trade-off for all videos, are available in enterprise version (which is free for all buyers of enterprise main and subjective reports).


High quality encoding report summary

According to YUV-SSIM, YUV-PSNR and Y-VMAF quality scores, the best codecs among our competitors evaluated for two use cases are the following:

  • First place: aom
  • Second place: SVT-AV1
  • Third place: x265
rav1e shows the best scores according to V-SSIM and V-PSNR metrics.

Mean overall quality (SSIM):


As we analysed the case of offline encoding, the limit on encoding speed was not strict in this part of the comparison. The following speed-quality chart shows the landscape of competitors relative speed and quality for all video sequences.


The following plot shows per-frame VMAF scores for axebat video sequence at 2 Mbps. SVT-AV1, SVT-HEVC and SVT-VP9 show quality decrease in a part of the video with high motion and complex scenes (running baseball players, splashing water). Enterprise report includes an HTML report with per-frame metrics results for all videos and bit rates.


Comparison Rules


HEVC codec testing objectives

The main goal of this report is the presentation of a comparative evaluation of the quality of new HEVC codecs and codecs of other standards using objective measures of assessment. The comparison was done using settings provided by the developers of each codec. Nevertheless, we required all presets to satisfy minimum speed requirement on the particular use case. The main task of the comparison is to analyze different encoders for the task of transcoding video – e.g., compressing video for personal use.

Test Hardware Characteristics

  • CPU: Intel Socket 1151 Core i7 8700K (Coffee Lake) (3.7Ghz, 6C12T, TDP 95W)
  • Mainboard: ASRock Z370M Pro4
  • RAM: Crucial CT16G4DFD824A 2x16GB (totally 32 GB) DIMM DDR4 2400MHz CL15
  • OS: Windows 10 x64

For this platform we considered three key use cases with different speed requirements:

  • Fast/High Density – 1080@60fps
  • Universal/Broadcast VQ – 1080p@25fps
  • Ripping/Pristine VQ – 1080p@1fps and SSIM-RD curve better than x264-veryslow

See more on Call-for-codecs 2019 page

Videos

Videos for testing set were chosen from MSU video collection via a voting among comparison participants, organizers and an independend expert. In this year, the collection was enlarged up to 18418 sequences which were chosen from 700000+ user-generated videos posted on Vimeo website. All videos have 4K or FullHD resolution and high bitrate (50 Mbps was selected as a lower bitrate boundary).

Number of videos in MSU video collection
Year # FullHD videos # FullHD samples # 4K videos # 4K samples Total # of videos Total # of samples
2016 3 7 882 2902 885 2909
2017 1996 4638 1544 4561 3540 9299
2018 4342 10330 1946 5503 6288 15833
2019 4945 12402 2091 6016 7036 18418

Bitrate distribution of videos in MSU video collection Videos bitrate distribution

Final video set consists of 100 sequences: 8 from the old data set and 92 new ones from Vimeo and media.xiph.org derf's collection. The average bitrate for all sequences in the final set is 218.9 Mbps, median - 143.2 Mbps. City walk (61.5 Mbps), Nancy (67.6 Mbps) and Oman museum (69.7 Mbps) sequences have minimal bitrates.

Descriptions of all test videos are presented on this page and in a separate PDF provided with our reports.

Video sequences selection

Codec Analysis and Tuning for Codec Developers and Codec Users


Computer Graphics and Multimedia Laboratory of Moscow State University:

  • 15+ years working in the area of video codec analysis and tuning using objective quality metrics and subjective comparisons.
  • 27+ reports of video codec comparisons and analysis (H.265, H.264, AV1, VP9, MPEG-4, MPEG-2, decoders' error recovery).
  • Methods and algorithms for codec comparison and analysis development, separate codec's features and codec's options analysis.

We could perform next tasks for codec developers and codec users.

Strong and Weak Points of Your Codec

  • Deep encoder parts analysis (ME, RC on GOP, mode decision, etc).
  • Weak and strong points for your encoder and complete information about encoding quality on different content types.
  • Encoding Quality improvement by the pre and post filtering (including technologies licensing).

Independent Codec Estimation Comparing to Other Codecs for Different Use-cases

  • Comparative analysis of your encoder and other encoders.
  • We have direct contact with many codec developers.
  • You will know place of your encoder between other newest well-known encoders (compare encoding quality, speed, bitrate handling, etc.).

Encoder Features Implementation Optimality Analysis

We perform encoder features effectiveness (speed/quality trade-off) analysis that could lead up to 30% increase in the speed/quality characteristics of your codec. We can help you to tune your codec and find best encoding parameters.

Thanks


Special thanks to the following contributors of our previous comparisons

Google Intel AMD NVidia
ATI Adobe ISPhone dicas
KDDI R&D labs Dolby Tata Elxsi Octasic
Qualcomm Voceweb Elgato Telecast
Huawei MainConcept Vitec Tencent

Share with us your opinion about comparison



Contact Information



compression.ru
in cooperation
with
Lomonosov MSU
Graphics & Media Lab
(Video Group)
Dubna State University
Institute for Information
Transmission Problems RAS

Subscribe to report updates



Materials about MSU Codec Comparison


Call for cloud encoding comparison participation 2020
See all MSU Video Codecs Comparisons

MSU video codecs comparisons resources:


Other Materials


Video resources:

Last updated: 02-July-2020


Server size: 8069 files, 1215Mb (Server statistics)

Project updated by
Server Team and MSU Video Group

Project sponsored by YUVsoft Corp.

Project supported by MSU Graphics & Media Lab