New H.266 VVC codec shows promise for 4K and 8K streaming

[nanoguy]

Something to look forward to: Video data is expected to account for over 82 percent of the global Internet traffic by 2022, and with the proliferation of 4K and 360-degree video, better video codecs are needed to reduce the bandwidth and space requirements of your favorite content. That's where Versatile Video Coding (VVC) comes in, which is able to achieve the same subjective quality as HEVC at only half the bitrate.

The Moving Pictures Expert Group has three different video codecs lined up for release this year: Low Complexity Enhancement Video Coding (LCEVC), Essential Video Coding (EVC), and Versatile Video Coding (VVC).

The first is part of the MPEG-5 standard and is designed for live streaming on services like YouTube and Twitch. To that end, LCEVC acts as an add-on codec that takes an existing one like AVC or HEVC and improves the encoding speed. It does so by taking a base image encoded at a lower resolution and adding detail and sharpness by combining that with two layers of encoded residuals. Preliminary tests also showed improvements in decoding efficiency, which could lead to better battery life on mobile devices.

Essential Video Coding is presented as an alternative to AVC (H.264) and HEVC (H.265), but it really is more of a workaround to the overly complicated and broken licensing structure of the latter codec. EVC simplifies things with a baseline profile that's based on royalty-free technology and a mainline profile that will incur licensing costs. The first should offer bitrate savings of around 31 percent when compared to AVC and the second should achieve similar image quality to HEVC Main 10 with a 26 percent lower bitrate.

The most interesting of the three new codecs by far is VVC, which is also known as MPEG-I Part 3 or H.266. According to the Fraunhofer Heinrich Hertz Institute, who penned the official announcement today, it's the result of ten years of research and development backed by companies like Apple, Intel, Microsoft, Qualcomm, Huawei, Ericsson and Sony.

At least on paper, VVC should offer the same visual fidelity as HEVC (H.265) with around 50 percent of the bitrate, which will lead to significant storage and bandwidth savings. Keep in mind that this new codec was made with ultra high definition (4K and 8K) and 360-degree video content in mind, so the efficiency gains are smaller for lower resolution video.

However when compared to the royalty-free AV1, VVC sounds a little less impressive. For one, AV1 is already around 30 percent better than HEVC, and doesn't come with royalty expenses, while VVC is encumbered by patents -- this is why companies like Netflix have already started using the former to serve content to Android devices. More importantly, web browsers like Firefox and Chrome already support AV1.

Hardware decode support is well underway for AV1, while the first software decoder (and associated encoder) for VVC is expected this fall at the earliest. One thing we still don't know is how much compute power is required to encode videos using VVC, but neither HEVC not AV1 are particularly easy to use without powerful hardware. Preliminary tests show that VVC is anywhere between four to ten times more complex to encode when compared to HEVC.

Ultimately, the success of the new VVC codec will largely depend on how many organizations decide that its efficiency improvements are better than AV1's simple, royalty-free licensing. Leonardo Chiariglione, who is the co-founder and former chairman of the Moving Picture Experts Group, doesn't think VVC (or EVC and LCEVC, for that matter) will see a better adoption rate than HEVC unless patent holders can "get their acts together."

Permalink to story.

https://www.techspot.com/news/85889-new-h266-vvc-codec-shows-promise-4k-8k.html

 

[VitalyT]

One thing we still don't know is how much compute power is required to encode videos using VVC, but neither HEVC not AV1 are particularly easy to use without powerful hardware. Preliminary tests show that VVC is anywhere between four to ten times more complex to encode when compared to HEVC

Nobody cares how much power it needs to encode, since it is done by content providers. It is the power to decode that's important, which is what every client device will have to implement.

 

[Plutoisaplanet]

Nobody cares how much power it needs to encode, since it is done by content providers. It is the power to decode that's important, which is what every client device needs to implement.

Agreed, or to put it another way you upload once and encode once, then download several times and decode several times. I’m with you.

 

[brucek]

Video encoding seems pretty relevant to me - FaceTime, Zoom, recording the grand kids or the cat, etc etc. But maybe that's not what these particular codecs are about.

I want to hear more about what the royalty arrangements are. I'm not sure why it's in anyone's interest to potentially clog the internet more than it needs to be by making it a conflict for device makers to use them.

 

[umbala]

Nobody cares how much power it needs to encode, since it is done by content providers. It is the power to decode that's important, which is what every client device will have to implement.

People like you who love giving their opinions on the internet actually know next to nothing about what goes on behind the scenes. Content providers most definitely care about how much power any given service requires.

 

[VitalyT]

People like you who love giving their opinions on the internet actually know next to nothing about what goes on behind the scenes. Content providers most definitely care about how much power any given service requires.

And people like you come to criticize others, without offering any valid view of their own, because they don't really have any.

Providers encode their content only once, and that's it, while consumers decode the content unlimited number of times, while provider only serves the data from there on, and cares only about the size of the data, which is what costs them money when streaming, and not how much it took them to encode once in the beginning.

 

[Evernessince]

People like you who love giving their opinions on the internet actually know next to nothing about what goes on behind the scenes. Content providers most definitely care about how much power any given service requires.

If a codec requires more resources but produces a higher quality or smaller output file, they will certainly pursue that regardless of the increased computational requirements.

They only have to encode the video once for each format they offer, the files are simply propagated after that. They are going to reap the benefits of the reduced bandwidth usage on their servers as will the customers. Netflix has how many customers? 182.8 million. It's not hard to realize that even a small size difference across that many customers is going to vastly outweigh any single time encoding cost. Not even mentioning the quality improvement that comes with more resource intensive encoding as well. It is truly amazing to me how much smaller you can make videos without visual impact. Well, at least in the hand of someone who knows what they are doing.

 

[VitalyT]

If a codec requires more resources but produces a higher quality or smaller output file, they will certainly pursue that regardless of the increased computational requirements.

They only have to encode the video once for each format they offer, the files are simply propagated after that. They are going to reap the benefits of the reduced bandwidth usage on their servers as will the customers. Netflix has how many customers? 182.8 million. It's not hard to realize that even a small size difference across that many customers is going to vastly outweigh any single time encoding cost. Not even mentioning the quality improvement that comes with more resource intensive encoding as well. It is truly amazing to me how much smaller you can make videos without visual impact. Well, at least in the hand of someone who knows what they are doing.

Correctamundo!

 

[Polaris1983]

Well there it is. 8k and 4k native cable TV packages. With and without colorization and 70mm effect added at correct aspect ratios.

 

[Mister_K]

People like you who love giving their opinions on the internet actually know next to nothing about what goes on behind the scenes. Content providers most definitely care about how much power any given service requires.

Decoding these delivery codecs is a *****. They're great and efficient but a ***** to work with. However indeed, content providers want the least amount of bandwidth with maximum amount of quality used, it's why these codecs do so well. Then there is VP9, fairly under utilised imho...

 

[Polaris1983]

The ai software and hardware will take care of that issue ok? Give it some time.

 

[DjoCoeur]

Well I wish better succes to H266 than the failure that is H265. The vast majority of people still prefer H264 because they know any device will be able to play the video and it can be encoded much faster. Unless you care a lot about file size and 4k or 8k display, the disadvantage of not using H264 are still too important until all devices are able to hardware encode and decode.

 

[hk2000]

Nobody cares how much power it needs to encode, since it is done by content providers. It is the power to decode that's important, which is what every client device will have to implement.

I'm sure the author knows that. It's probably a typo!

 

[hk2000]

Everybody here seems to be forgetting about the Licensing fees. Makes a big difference to content providers! For an incremental savings in storage space compared to royalty free codecs, it won't make a dent. And trust me, 4K and 8K will never be as much in demand as people who frequent Tech sites like this think it it will be. Most consumers don't care and providers either already know that fact, or will very quickly find out.

 

[mrvco]

Video encoding seems pretty relevant to me - FaceTime, Zoom, recording the grand kids or the cat, etc etc. But maybe that's not what these particular codecs are about.

I want to hear more about what the royalty arrangements are. I'm not sure why it's in anyone's interest to potentially clog the internet more than it needs to be by making it a conflict for device makers to use them.

Exactly regarding royalties, the next royalty quagmire that the 'Motion Picture Experts Group' standards body and their licensing group, MPEG-LA, avoid will be the first one.

 

[killmess]

Video encoding seems pretty relevant to me - FaceTime, Zoom, recording the grand kids or the cat, etc etc. But maybe that's not what these particular codecs are about.

This is a codec for 4k, 8k or higher resolution videos.
Unless you have a 4k or 8k webcam, you won't see any benefit in Facetime and Zoom with this new codec.
;P

 

[brucek]

This is a codec for 4k, 8k or higher resolution videos.
Unless you have a 4k or 8k webcam, you won't see any benefit in Facetime and Zoom with this new codec.
;P

No. From the linked press release:
"Overall, H.266/VVC provides efficient transmission and storage of all video resolutions from SD to HD up to 4K and 8K, while supporting high dynamic range video and omnidirectional 360° video."

These codecs would have very limited current appeal if they could not provide benefits for resolutions below 4K.

 

[ampvick]

Everybody here seems to be forgetting about the Licensing fees. Makes a big difference to content providers! For an incremental savings in storage space compared to royalty free codecs, it won't make a dent. And trust me, 4K and 8K will never be as much in demand as people who frequent Tech sites like this think it it will be. Most consumers don't care and providers either already know that fact, or will very quickly find out.

It's all marketing. If service A is "only" 1080p and service B is 4K, people will choose service B. They will never know that 99% of the content is upscaled and even if they did they won't be able to tell the difference!

 

[Aryassen]

Well I wish better succes to H266 than the failure that is H265. The vast majority of people still prefer H264 because they know any device will be able to play the video and it can be encoded much faster. Unless you care a lot about file size and 4k or 8k display, the disadvantage of not using H264 are still too important until all devices are able to hardware encode and decode.

I didn't know h265 was a failure. I loved h264, but then realising how much space I can save, I have encoded my entire library (hundreds of titles) to h265. OK, I admit, it is also a kind of a hobby for me The point is, I never had issues with h265. It plays well on my 3 year old Android phone (model is from 4 years ago, actually), also on the missus' iPhone, on my Android tablet(s), on my laptop(s), and of course, on the desktop (even before my upgrade, on a 9 year old 2600K).

I admit I did encoding only on the desktop, but speeds were comparable. I think the reason is that for h265 I could settle for "medium", as it provided the nearly same visual qulities and compression ratio as the heavier settings (difference was around the 2-3% mark). Medium preset on h265 was actually a fair bit faster than "slow" for h264 (and for that the difference between presets was far more noticeable, both for size and quality, so I had to stick with "slow", the least), quality was similar or slightly better, and output size was significantly smaller. What's there not to like? For me h265 is great!

 

[DjoCoeur]

For me h265 is great!

If you have a fast computer that can hardware encode and decode, h265 is better. But as soon as you want to share videos, you'll get lots of people say they can't play the video, or it's like a 2 FPS photo display.
Not that I download lots videos from illegal places, but when look there, you find that h264 is used for like 95% of the videos. That's why I said h265 is a failure, since it's not used half as much as it could.