[HN Gopher] Inside Netflix's bet on advanced video encoding
___________________________________________________________________
Inside Netflix's bet on advanced video encoding
Author : marban
Score : 59 points
Date : 2024-06-23 16:00 UTC (7 hours ago)
(HTM) web link (www.theverge.com)
(TXT) w3m dump (www.theverge.com)
| paulryanrogers wrote:
| Wait, doesn't Netflix have a reputation for terrible encoding
| quality at 4K? (I don't recall being impressed when I last tried
| it)
| Havoc wrote:
| I think flying close to the sun is precisely aim of the game
| here. So anyone sensitive to it is likely to conclude it is
| terrible.
| Omniusaspirer wrote:
| Not just 4K- all resolutions and it's getting worse with plenty
| of comparisons available online for those who care to look.
| Streaming video quality has generally declined across all
| platforms in the last few years with the exception of Apple TV+
| and contrary to this puff piece it's to the point it's very
| noticeable to the discerning end user.
|
| Enshittification continues relentlessly.
| SSLy wrote:
| Movies Anywhere is consistently great quality.
| akira2501 wrote:
| > Enshittification continues relentlessly.
|
| It's just rampant monopolization.
|
| Enshittification makes it sound like it's a natural expected
| process, for companies to buy their way into dominance the
| market, destroy competition, then deceive and underserve
| customers while raking in undeserved profits.
|
| It also fails to lay the blame on Congress and the
| Administration's doorsteps, where it belongs.
| jfengel wrote:
| You mean, dinging them for false advertising? Or creating a
| standard?
| lukifer wrote:
| I support your general thrust in spirit, but I think
| "monopoly" isn't a helpful frame here. Consumers have a
| glut of options in streaming services: Prime, Max, AppleTV,
| Hulu, even YouTube (contrast with one of the classic
| monopolies, AT&T, which for many was the only game in town,
| unless you consider "write a letter to your friend instead"
| to be a valid substitution good).
|
| I feel the correct frame is the general category of "market
| power" (including, but not limited to, the "anti-
| competitive practices" that forms the basis of current
| regulatory policy).
|
| In the case of streaming, the closest analogue to
| monopolization is vertical integration: the custom-produced
| content which is only accessible on a particular streaming
| platform (sometimes including no releases on physical
| media).
|
| On the one hand, Netflix could be considered to have a
| monopoly over the distribution of "Squid Game" (and Prime
| over "The Boys", Apple over "Silo", etc), meaning they can
| raise prices, not for the product itself, but for the "club
| goods" distribution platform, which the consumer may not be
| interested in otherwise.
|
| On the other hand, entertainment/culture are not the same
| kind of "inelastic demand" necessity as telephone service
| in the past, or internet service now. And when we look at
| actual outcomes, I'd say quality for vertically-integrated
| productions are better on average than older business
| models; and while prices are going up, I wouldn't call them
| ridiculous price-gouging either (in addition to competing
| with each other, the streaming services are also competing
| with piracy and password-sharing).
|
| Anyway, I think it's absolutely the case that
| enshittification isn't inevitable, and we should be wary of
| succumbing to learned helplessness in the face of capital
| consolidation. I _would_ claim it 's a "natural expected
| process" of the shareholder corporation, and the perverse
| incentives of short-term profits and stock prices, even
| when it comes at the expense of long-term brand equity. But
| solving that general problem is probably out of scope for
| "the bitrate is too damn low and the subscription fees are
| too damn high".
|
| I'd love to see more regulation in the space, but
| realistically it'd be tricky (I don't expect any Doctorow-
| approved abolition of DRM, or mandatory interoperability,
| anytime soon). One could perhaps establish streaming-
| quality standards, where a company couldn't advertise "4K"
| unless it hit some threshold. But when it comes to the
| issue of streaming quality in particular, I suspect the
| tragic reality is that the majority of users simply won't
| notice or care.
| jsheard wrote:
| I recall several streaming services reducing their bitrates
| during the pandemic, ostensibly to offset the increased
| network load from everyone watching more while stuck at home,
| did those reductions ever actually get walked back or did
| they just become the new default?
| chgs wrote:
| are torrents no longer available in a high bitrate capacity?
| tambourine_man wrote:
| I haven't watched them in many years but my experience was that
| compression sucked on dark scenes even in 1080p. On the other
| hand, you were able to watch on a 3G connection without any
| buffering bumps
| erikpukinskis wrote:
| ... and one of these things hurts engagement.
| SSLy wrote:
| New 18Mb/s transfers are better than their back stock.
| malux85 wrote:
| Is it just me or has anyone else noticed a big increase in the
| number of encoding artifacts recently?
|
| Rapidly oscillating patches, blurry smears, it's only Netflix
| (other video streaming services are fine) its getting quite
| annoying, a few times it's been so severe I thought it was a
| problem with my projector
| MaxikCZ wrote:
| Rarrr, no issues here
| noahtallen wrote:
| Definitely noticeable on Netflix, but I also have some specific
| issues with Amazon Prime, where certain scenes have these
| weird, thin horizontal artifacts.
| blackeyeblitzar wrote:
| I wonder if the issue is the fact that these artifacts are
| layered on top of processing artifacts from TVs, receivers,
| streaming devices, and all of that. Does Netflix test their
| encoding in isolation or in these real life setups?
| Havoc wrote:
| I'm a little surprised that we haven't seen something more AI
| driven yet. (yes yes buzzword I know)
|
| i.e. slice it not just into scenes but also into objects and do
| bitrate on that level. i.e. Face and objects in foreground get
| more. It seems we now have pretty small models that can do that
| sort of stuff (see Apple & MS ones recently) so should be
| feasible at scale.
|
| I'd imagine you can also train an LLM on patterns that encoders
| choke on...chequered patterns etc.
| cornstalks wrote:
| > _i.e. Face and objects in foreground get more._
|
| Encoders have been doing that for decades at this point.
|
| AI-related models are actively being researched for multiple
| applications in video encoding. From generating predicted
| frames, to generating fine details, choosing tools and
| parameters for each frame, narrowing the search space, etc.
| 42lux wrote:
| You mean transformer/diffuser based when you are talking about
| AI in this context? Because there are a lot of ml papers out
| there for encoding video. Stadia (RIP), Parsecs, Epic (Unreal
| Pixel streaming), META, Apple and Nvidia doing amazing stuff
| for their streaming services/VR/AR plays.
| kmeisthax wrote:
| After he got kicked out of MPEG, Leonardo Chiariglione opened
| his own standards organization[0] called MPAI, which is focused
| primarily on... standardizing machine learning techniques for
| video coding tools.
|
| I'm not sure if they've actually shipped anything yet, though.
|
| [0] With blackjack, and hookers:
| https://blog.chiariglione.org/the-mpai-framework-licence-app...
| coldsmoke wrote:
| We've experimented with that at SVT (Sweden's public service
| broadcaster). Or rather, the video team has, I'm just a web
| dev, so I don't know the details other than what's in this blog
| post:
|
| From the eyes of the viewer: Attention-aware video compression
| for improved low bitrate video https://medium.com/the-svt-tech-
| blog/from-the-eyes-of-the-vi...
| karmakaze wrote:
| Then we can have many 6-fingered characters and other
| hallucinations. It's all fiction anyways so what's one for a
| other? I hope it either gets great fast or phases out as a fad.
| coretx wrote:
| Decreasing the gamut before encoding is not my definition of
| "advanced video encoding". I"d like to call it "ghetto video
| encoding" instead. What bothers me most about it, is that many
| people are slowly getting used to it. It also makes me feel sorry
| for the people who worked hard on the production. Very few people
| will ever see the true quality of their work.
| kmeisthax wrote:
| Per-shot encoding sounds like something that should be handled by
| multipass encoding. Presumably it isn't - so I'm wondering what
| is failing in those encoders to make it necessary to tweak
| settings that much.
|
| Or are they just aggressively searching for corners to cut to
| save bits?
| refulgentis wrote:
| You're 100% right, the first instinct a programmer would have
| as a codec dev is "gee we def don't need that many bits to
| encode pure black frames"
|
| I can't wrap my mind around this article and exactly what's
| going on because AFAIK VBR has been default ~forever. You don't
| even need multipass for that because you just have to look
| ahead N frames, it was received wisdom, statistically backed,
| as far back as 2007 that multipass was a waste of time because
| it was extremely marginally, like .0001%, better than just
| doing VBR, and using CBR was really dumb and wrong unless you
| had very specific edge case technical needs. I can't even
| remember what, maybe poorly designed initial H.264 decoders or
| people feeding into legacy broadcast system
|
| Little more context here:
| https://news.ycombinator.com/item?id=40770734
| gyudin wrote:
| Wonder how much has Netflix payed for this article to the Verge.
| ilrwbwrkhv wrote:
| Netflix video quality is decreasing. I don't know what they are
| doing, but it is not working.
| aeturnum wrote:
| I don't really know what people "generally" think of Netflix
| encoding, but my experience is that it's Very Bad. It's a
| combination of factors:
|
| - Netflix will only send you "high quality" (over 720p) streams
| on certain browsers (IE on windows, Safari on OSX, ???? on Linux)
| that support Encrypted Media Extensions.
|
| - Netflix also auto-scales the quality they are sending based on
| their understanding of your connection.
|
| - Also apparently they do per-title re-encoding passes.
|
| This all combines, in my experience, to on average the worst
| streaming quality overall and also the most opaque. There are
| debug modes you can enable to see some of this, but generally
| it's very hard to tell what quality you are looking at and what
| is preventing you from getting a nicer quality. I also find that
| Netflix's "low profile" content (i.e. not max quality) looks bad
| - i.e. the 720p stream looks quite bad in addition to being low
| resolution.
| vosper wrote:
| > Netflix also auto-scales the quality they are sending based
| on their understanding of your connection.
|
| Netflix dropping to potato quality after a few minutes is
| infuriating. Refresh the browser and quality goes back up for a
| few minutes before it drops again. My wife often doesn't notice
| it; it drives me nuts. My theory is that they try to save
| bandwidth by dropping quality after some time and hoping you
| don't notice.
| LegitShady wrote:
| 100% agree with all of your points
| Retr0id wrote:
| > Netflix will only send you "high quality" streams on certain
| browsers [...] that support Encrypted Media Extensions.
|
| They won't send you anything at all unless you have EME,
| coupled to a supported DRM engine. Getting >1080p requires a
| "more secure" DRM engine (e.g. Widevine L1, PlayReady SL3000)
| MaxBarraclough wrote:
| It's similar with YouTube's premium content, i.e. DRM-protected
| films or series available for rent or purchase on YouTube. It
| plays at a maximum of 480P on most desktop browsers, including
| Google's own Chrome for Windows. (Disclaimer: I haven't checked
| this recently, I'm assuming the situation hasn't improved.)
|
| * https://support.google.com/youtube/answer/3306741
|
| * https://support.google.com/youtube/thread/89050028/my-
| movie-...
| jamesponddotco wrote:
| Because of this quality issue, I always wondered how people who
| rip Netflix content handle this, since they often offer 2160p
| Dolby Atmos versions of movies or even TV show episodes for
| download.
|
| How do they maintain the quality long enough to rip the
| content? When I used Netflix years ago, I could barely keep a
| constant 1080p stream at times.
| gruez wrote:
| They use HDCP strippers, devices which can strip protections
| from the hdmi signal, at which point they can feed it into an
| capture card to reencode.
| AltF4me wrote:
| I think the OP's question was maintaining the higher
| quality stream, not the act of ripping.
| mock-possum wrote:
| For whatever it's worth, I've never had any problem with it.
|
| I've got a bookshelf computer hooked up to a 52" 1080p screen
| in my living room, nothing fancy, and I've never noticed any
| issues at all with quality, or buffering times. I just open the
| video through the website in chrome, and play it, and it's
| fine.
| ChrisArchitect wrote:
| Related?
|
| _German Court Fines Netflix EUR7.05 Million for Continued
| Infringement of Broadcom HEVC Patent (2023)_
|
| https://www.broadcom.com/company/news/product-releases/61711
| metadat wrote:
| I wonder how Anne Aaron came to be the leader for this aspect of
| the business. Netflix is really only doing re-encoding, which is
| essentially what all first-tier media pirate groups specialize
| in. It's interesting the pirate groups deliver much higher
| quality releases in very bandwidth efficient formats (x265 is
| incredible), yet Netflix still consistently looks blocky on my 4k
| LG with 1Gbit Internet. Also, every dark scene looks doubly
| atrocious.
|
| You'd think netflix could do better than the ad-hoc groups of
| individuals who do it for free in their spare time.
| refulgentis wrote:
| This is some really glorified PR that's either dumbed down or
| extremely excessive in attributing basic insights and algorithms
| to Netflix itself
|
| I got into programming/software by encoding my _cough_ well-
| sourced _cough_ movies /TV shows to MP4 for my iPod video.
|
| Far be it from me, maybe it's insufferably geeky detail, but the
| slow decade-long march described as "gee each movie is different"
| and "gee each scene is different" followed by Herculean work of
| FAANGers insufficiently appreciated by creative types was solved
| by VBR years upon years earlier. (VBR = variable bit rate)
|
| Once you're getting to "we'll use ML as a VBR algorithm!", that's
| original, but the problems described and solution was
| understandable and solvable by a 18 year old non-programmer in
| 2007 with free software.
|
| VBR wasn't some niche thing either, it's a very very obvious
| optimization I've never seen a codec miss, from MP3 audio to MP4
| video. There's no caveats here or haughtiness or flippant
| "Dropbox is rsync + my weekend" dismissiveness on my part. It
| wasn't news to _anyone_, it's a very obvious optimization that
| was applied by everyone
|
| I'd be veeeeery curious if there was much contribution here
| beyond using x264, occasionally with patches, and then
| engineering a pipeline around it
___________________________________________________________________
(page generated 2024-06-23 23:02 UTC)