Not just low-budget producers but also equipment manufacturers who could build it into their hardware without that hardware becoming more expensive.
Where I work we have bandwidth-intensive applications which have evolved from H.264 to H.265 and it makes a difference in terms of stability. We would love EVC baseline if all it meant was a free or near-free firmware update.
From the EVC paper:
“The results presented in this paper show that the MPEG-5 EVC standard can reach the same video quality as HEVC at 26% lower bitrate on average, over a set of UHD and HD sequences.”
That has improved a lot, but yeah still slow.
I suspect it will be in NVENC and Quick Sync Video as they both supports it now for decode.
YouTube has started delivering in it, and Shotcut users have already started reporting bugs playing WebM they downloaded.
Still, no support from Apple, no surprise.
Throw H266, aka VVC, into the mix.
Too many codecs and acronyms to remember.
Yeah, the AV1 speed improvements have been huge. For my purposes, libsvtav1 on a processor with AVX512 is in a usable range. I would choose it over libx265 every day of the week… but again, that’s for my purposes. Other people may have tighter deadlines to meet.
By “crumbling”, I was referring to the Sisvel patent pool situation that threatened the free nature of AV1. If AV1 ceased to be an open format, it would change the whole codec landscape. However, since I wrote that crumbling statement, the corporations behind AOMedia (AV1) have not-so-subtly threatened to crush opponents with a legal sledgehammer should they take AV1 patents to court. Given the deep pockets of these corporations and the density of research their legal teams have already done to ensure the free use of prior patents, AOMedia is expected to make good on those threats and protect the freedom of AV1, at which point maybe Sisvel will fade back into the dust where it belongs.
Basically, the future of AV1 has gotten a lot brighter since I made that statement.
As for VVC, that was already covered.
What would be your go-to encoder: AV1, EVC or VVC?
It’s really hard to say because each codec specializes at different things. I’m usually not one to speculate on the future because I’m not a prophet and I don’t want to sway somebody else the wrong direction. But since you’re asking about my individual purposes, my go-to codec would always depend on the nature of the project:
Family video: probably EVC for being fast and free. Maybe AV1 if space savings are significant since there is no rush in terms of encoding time. It’s hard to know until the EVC encoders mature for a fair comparison.
Encoding a video for YouTube on a rigid weekly schedule: probably EVC because it’s so fast to encode, which gives me much more time to film and edit
3D/360 video with spatial audio, or virtual/augmented reality integration: VVC because it has native support for these formats at a level the other formats do not even support
A self-produced instructional video series that I want to sell on Blu-ray data discs with the expectation of being played on a computer: probably EVC because distribution is royalty-free, and playback should be smooth on even low-end computers thanks to EVC’s lower decoding complexity (I would not want to handle support calls for choppy AV1 playback as a low-budget producer)
Designing a transcoding ladder for a streaming platform: probably AV1 or VVC because they support S-frames (switch frames) which is great for mobile devices needing to switch resolution on the fly. If the streaming company was large enough to hit the licensing cap of VVC, then I’d probably use VVC since it does have higher quality at lower bitrate (based on current encoders, subject to change). But a smaller streaming company might benefit from AV1 to financially save on the royalty situation.
Encoding a major film and targeting the widest device support in a couple of years: probably VVC. This is the weird part… AV1 and even FLAC and VP9 are free formats, but Apple devices in particular don’t play them. Apple is adding partial support for VP9 in iOS 14, but how many years after-the-fact? I call this weird because some companies (like Apple) that are members of MC-IF tend to support MC-IF formats almost exclusively, of which VVC is one (as was HEVC). So Apple would likely support VVC right out of the gate, as would Huawei and Samsung. But that’s speculation.
So yeah, my go-to isn’t clear-cut. It depends on playback device compatibility requirements and the financial royalty situation and bandwidth requirements and the type of project at a minimum, which includes the production schedule’s tolerance for encoding time. Curiously, AV1/EVC/VVC are all so good at compression that the war for the average person isn’t even about bitrate reduction anymore. We just need something with acceptable cost (ideally free) and widespread device support that doesn’t take all year to encode. AV1 is close to fitting that bill. But if it doesn’t get significantly faster or make significantly smaller files soon, then the space savings compared to EVC might be small enough that I begin to prefer EVC simply for the radically reduced encoding time… that is, when it comes to personal projects. I haven’t heard enough industry rumblings about EVC getting widespread device support yet to recommend using it for published projects. VVC seems to be the clear favorite for future device support. (AV1 does have significant support, but not having Apple and a few others is a pretty big deal. VVC would likely have them all.) Sigh… the future could go so many different directions. But on the upside, there is an ideal format for any project we can imagine now. 
This topic was automatically closed after 90 days. New replies are no longer allowed.