The battle of the codecs: H.265 vs VP9 vs AV1
In this post, we are going to try to shed some clarity on a fundamental question: which codec is better?
There are numerous studies that focus on analyzing in detail the performance of video coding systems and, in Digital-Content.es, we are going to try to summarize some of their results to clarify the battle between the H.265, VP9 and AV1.
At the beginning of this decade, work on high-efficiency video coding with standards such as H.265 (MPEG-H Part 2 or HEVC, High Efficiency Video Coding) or VP9 saw the light. In recent years, the Alliance for Open Media has focused on the development of AV1. As we already commented from this blog, “AV1 aspires to become the industry standard, displacing VP9 and competing directly with HEVC / H.265, being more efficient than this and, also, open source” (Why should you start to use the AV1 codec?).
The questions that arise are: is it true that AV1 is much better than the rest of the standards? … And, between H.265 and VP9, which one is better?
Regarding the bit rate, according to a study entitled “Comparison of Compression Efficiency between HEVC / H. 265, VP9 and AV1 based on Subjective Quality Assessments”, it is shown that the AV1 and H.265 codecs account for 28% and a 27% bitrate savings respectively compared to VP9. However, the same analysis specifies that, although it is true that H.265 assumes a bitrate gain in relation to VP9, the differences are not very significant when working with very high bit rates. For its part, the fight in this regard between AV1 and H.265 is served, with very similar performances, although slightly favorable to H.265, largely conditioned to the audiovisual content used for encoding.
In terms of encoding time, another study titled “Future Video Coding Technologies: A Performance Evaluation of AV1, JEM, VP9, and HM” reveals that AV1 requires more time to achieve greater efficiency in video compression. In this sense, the analysis specifies that AV1 requires approximately 58 times more execution time than VP9. For its part, the research entitled “Study on H.265 / HEVC against VP9 and H.264: On Space and Time Complexity for Codecs” concludes that VP9 consumes approximately 35-45 times more encoding time than its competitor H.265 .
Following the data of these two parameters in the studies presented, the H.265 codec clearly outperforms its competitors. If we analyze the visual distortion generated by these encoders, according to parameters analyzed in “Comparison of Compression Efficiency between HEVC / H. 265, VP9 and AV1 based on Subjective Quality Assessments”, we can determine, based on objective and subjective metrics, that AV1 contains lower distortions than H.265 and VP9. This same analysis clarifies that at very high bit rates, the differences between all of them are not significant. AV1 and H.265 are subjectively perceived as very similar, so they compete strongly.
Therefore, taking into account the three factors that we have determined in this analysis and following the studies presented, we can consider H.265 as the winning codec in this battle. We trust in the evolutionary power of AV1, especially considering that it is an open and royalty-free solution.
The H.265 War vs. AV1 is still open!
