“For months, powerful companies including Google, Amazon, Microsoft and Facebook have been working to shrink online video sizes,” Stephen Shankland reports for CNET. “But their work just got more important, because Apple has now joined the partnership too.”
“The Alliance for Open Media is working on technology called AV1 that compresses video before it’s stored or sent over the network. That technology is crucial to keeping your phone from running out of storage space or your data plan from pushing past monthly limits,” Shankland reports. “But compression technology is useful only when it’s widely supported, and Apple was a major holdout.”
“Apple quietly joined the alliance as a founding member, according to the group’s website, which was updated with the change Wednesday,” Shankland reports. “Apple’s plans for AV1 aren’t yet clear — the company didn’t respond to a request for comment — but joining as a founding member sends a strong signal of support.”
“The dominant standard [is] called H.264 or AVC. Apple in 2017 championed its successor, called HEVC or H.265. But HEVC has been mired in patent problems as companies that contributed technology wrangle for lucrative royalty fees,” Shankland reports. “AV1 is still a work in progress, though the first version of the technology should be finalized in coming weeks. Mozilla, which supports an early version, said in November that AV1 cuts file sizes 25 percent to 35 percent compared to HEVC and VP9. Better compression can come with a problem, though: longer times to compress video and a greater burden on scarce computing resources like memory and battery life.”
Read more in the full article here.
MacDailyNews Take: AV1 just became real, thanks to Apple.
Now, let’s see if it can avoid being the subject of myriad video compression patent claims.
Lossy or not- that is the question.
If it is lossy you are ruining the picture in order to fit it in a smaller space and that data cannot be recovered.
Any compression more than about 2.1:1 is lossy. A few years back in the lossless imagery, motion imagery, and video worlds a “breakthrough” happened that got truly lossless compression for generic, random sets of those types of data got past the 2.08:1 that was the limit for several years. (Note: There are a couple groups out there with supposed mechanisms that *claim* they can do lossy compression where the induced compression noise is no greater than the noise of the video/imagery capture sensor itself. I intend to run a study to test this claim in the coming months.)
Current compression methodologies, e.g., H.264, H.265, and VC9, are all lossy compression.
Think of it this way: UHDTV (most often incorrectly referred to as “4K”, which is really a Digital Cinema standard with a different aspect ratio than current TVs) with 12 bit Dolby Vision HDR imagery (forget audio for a moment) is approximately 9 Gbps to 15 Gbps depending on the original, raw coding and metadata included. The theoretical maximum for UHD Blu-ray is just 128 Mbps. That’s at least a 80:1 compression implementation for some of the best consumer video out there, which is *way* into the lossy range. Netflix compresses things significantly further in trying to get things down to around a 10 Mbps data rate.
However, these are not “archive” copies that are being sent to consumers. Those archives are sometimes stored with lossless compression, but, unfortunately, most are not.
The whole point of the various lossy schemes is to make video as close to “visually lossless” as possible, meaning “the average observer will not notice any lossy compression artifacts”. The goal is not to have pristine video with zero artifacts.
The reality is that unless you’re doing your own video and creating your own movies, everything you see today is compressed using lossy compression.
“Lossy or not”
It’s LOSSY! The question is HOW LOSSY?!
All video we use today on computers is already lossy compressed by default. Compressing lossy video AGAIN via another lossy algorithm is a very bad idea IMHO. But I’ll await final judgement until after I get to see the final results. Low expectations.
Like I said, the signal passed by most Cable TV systems compared to the OTA is quite stunning.
Kind of sad to spend a truckload on great equipment and then have the signal compressed all to hell.
Absolutely. I Cut-The-Cable as I am sick of the price gouging scam that cable TV has become. Let it die.
Contrary to MDN’s take, it is more likely Apple joined so they wouldn’t be left behind.
Since AV1 is based largely on Google’s VP9 they indicated last year that YouTube would be able to move to support AV1 within 6 months of it being finalized.
Netflix also had announced their intention to be an early adopter of AV1.
https://en.wikipedia.org/wiki/AOMedia_Video_1
I agree, for the most part.
Video compression standards require two things:
1) they must be better at producing visually lossless imagery at a consistently higher compression ratio than the generation before (virtually all are variable bit rate [VBR] so there is no constant compression ratio), and
2) they must be widely adopted, e.g., AAC is better than MP3, but AAC had a small fraction of the adoption rate of MP3, similarly ALAC is marginally better than FLAC for audio, but FLAC has a much larger following.
If this new VC1 compression scheme is even 25% better, on average, than H.265 and it gets widely adopted by both content creators and content distributors, then we will have a new “standard” within a couple years.
Based on the article regarding Demuxed held 2 months ago AV1 was stated to be rated at 25% efficiency over HEVC/H.265 with a goal of 35% efficiency. 🙂
https://www.giantbomb.com/profile/cliffordbanes/blog/talks-about-4k-codecs-hevc-h265-av1-and-hdr-master/116176/