Slashdot Mirror
BBC Confirms 50% Bitrate Savings For H.265/HEVC Vs H.264/AVC (bbc.co.uk)
An anonymous reader writes: A research team from the BBC has done a series of tests to confirm earlier computations showing a ~50% savings in bit rate for H.265/HEVC compared to video using H.264/AVC at comparable quality. "The subjective tests used a carefully selected set of coded video sequences at four different picture sizes: UHD (3840x2160 and 4096x2048), 1080p (1920x1080), 720p (1280x720) and 480p (832x480), at frame rates of 30Hz, 50Hz, or 60Hz. The video content was chosen to represent diverse spatial and temporal characteristics, and then coded using HEVC and AVC standards at a wide span of bit rates producing a variety of quality levels." Here is the full published analysis. "The tests confirmed the significant compression efficiency improvements achieved in HEVC, verifying the results previously reported using objective quality metrics (PSNR based methods)." The team did not test against VP9, which is shaping up to be an impressive standard as well.
What is it's Weissman Score?
Another win for the mobile browsers, I guess.
I've played in this space in a former position. Interesting lessons learned:
- PSNR is nearly worthless: An image with almost the same score can look terrible. Not all the time, but enough of the time.
- The only quantitative test I found that worked reliably was an old analog Tektronix PQA500 (lots of work to use for digital CODEC.)
- Management didn't like the PQA data (it said our product was terrible), decided to use PSNR data (product is great!)
- Customers fixed this discrepancy and product line failed spectacularly (due to video quality, surprise!)
- I never could find any published information sufficient to recreate the Tek PQA algorithm.
Except it's not a for-profit company so who exactly is gaining from this "huge financial stake"? You could just as easily argue they could fluff the results in favour of H.264 in order to strengthen their claim for more funding.
Name another media company that went out of their way to develop a patent-free media codec that was independent and competitive with other codecs of the time? (Google Dirac)
The BBC are publicly-funded, and under immense pressure to justify their funding at the moment - there's talk of scrapping the TV licence, and with it the BBC. They receive no advertising revenue in the UK at all. They only get some foreign revenue from sale of media (not even their own codecs or patents), and that goes to their commercial arm which isn't funding stuff like this.
There's no profit in them evaluating codecs, only if they then go out and build their own hardware that uses it. They didn't manage to do that with Dirac either, so why they would with this I have no idea.
All they want to know is what's best to push through iPlayer and store in their archive.
Not-for-profit does not mean not-with-costs.
There are two types of people in the world: Those who crave closure
You need some serious hardware to play H.265 video without bringing your CPUs to their knees.
50% bitrate reduction vs H.264 sure, but not vs x264 which is the current gold standard for HQ video compression.
It's like comparing a new audio codec to the original fraunhofer MP3 encoder. LAME on the other hand is a significantly better MP3 encoder like x264 is a better H.264 encoder.
My own compression testing between HEVC and x264 show that at verty low bitrates, yes HEVC is better, but only at bitrates below what I would normally use and what I would consider "quality" encodes.
When you compare say a 10GB x264 encode of a full-length BluRay film, even 8GB for the HEVC does not provide an equal or superior copy.
If so then we still can't use it.
The only conclusion that can be drawn from this study is that for encoding at fixed quantization, the JM v18.5 AVC reference encoder is objectively and subjectively worse than the HM v12.1 HEVC reference encoder.
Raise your hand if you've ever used the reference encoder for either.
That means it'll take half the time to bittorrent Doctor Who episodes. :-)
The benefit for them is that they can reduce costs by reducing the bitrate, and claim not to have reduced picture quality. They have done it before - when HD broadcasts started they were around 18Mb/sec IIRC, but were later reduced to less than half that (average per channel, they actually balance about 18Mb/sec between two channels in a kind of VBR system). They claimed that the picture didn't suffer but it very clearly did, and it's now rather poor.
They are under immense financial pressure, and reducing bitrate (and picture quality) allows them to save something else, someone else's job.
const int one = 65536; (Silvermoon, Texture.cs)
SJW, n: "Someone I don't like, and by the way I'm a fuckwit" - AC
So what? Who cares? It doesn't matter.
H.265 is simply far too encumbered by "IP" costs and licensing restrictions to be of any use whatsoever. Sure, there are places (like the douchebags at BBC) who will decide to use it, but rational folks should run from it like the bubonic plague. Especially folks that create content. If you create something that uses H.265 as the master source, your copyright is worthless, because you content is now encumbered inside of format that requires you to pay fees if you want to extract it, convert it, distribute it, or even view it. "It can't really be that bad," you say. Sure, you can buy equipment and software that will do that for you, but part of your cost is the partial fee that the manufacturer is paying for licensing.
Someone have a chat with the BBC about putting resources into VP9 instead. It will save them and everyone else a lot of money, pain, lawyer fees and headaches into the future.
"Somebody has to do something. It's just incredibly pathetic it has to be us."
--- Jerry Garcia
Erm you do realise that this isn't actually a quality reduction?
If you read up on HEVC, you'll notice it's a completely separate codec to AVC that was designed specifically to hit 50% better (higher quality same bitrate or same quality at half bitrate). AVC was the benchmark codec that it was being compared to.
So the BBC is just confirming that they hit that mark.
https://en.wikipedia.org/wiki/...
Name another media company that went out of their way to develop a patent-free media codec that was independent and competitive with other codecs of the time? (Google Dirac)
There is nothing royalty-free about HEVC.
"Somebody has to do something. It's just incredibly pathetic it has to be us."
--- Jerry Garcia
Your paranoia is showing. The BBC reported the real difference between these codecs, and your first thought is that they're biased. How can you engage with the world at all while this far gone?
Umm, all the employees and in particular the leaders who draw multiple million dollar salaries?
Troll is not a replacement for I disagree.
I have done my own comparisons of AVC (using x264, single-thread, veryslow preset) and HEVC (using x265, disabling wavefront processing because it slightly reduces quality, veryslow preset). All 1080p video, significant because HEVC is supposed to scale to 4K better than AVC.
My conclusions:
1) x265 takes FAR longer to encode, but we knew that. Understandable.
2) When "low in bits", x265 blurs images rather than making them look blocky. This sometimes looks better but to me often looks worse.
3) x265 seems to force a denoise filter. Video is far easier to encode efficiently when denoised, so I figure this is part of the data savings. It's a bit of a cheat, however, because I can get far smaller file sizes by running a denoise filter myself for x264-encoded video.
I looked closely, for example, at Captain America the Blu-ray. Much of the detail of, e.g. car leather and grass and tree leaves is lost in an x265 encode, even at about the same overall data rate as x265/
x265 supports "--tune grain", roughly analogous to "--tune film" for x264, but it makes the video vastly larger -- often larger than x264's version, and it often looks worse. It does a better job of keeping grain, however.
My experience is very similar to many others' in forums. I had committed to switching my encoding to HEVC, but the results of my tests showed it is not ready for prime time. Some may not mind blurry ("soft" is probably a better word) video, or video that looks like it has been through a denoise filter, but I do.
This is not to say that x265 is junk. I am sure it will mature over time just like x264 had to over time. x264 started out as being not all that much better than divx, the previous generation.
Computer Science is no more about computers than astronomy is about telescopes. --E. W. Dijkstra
What would they know? If you spell BBC in Siryllic you get PRAVDA. Which means 'lying toady lapdogs'.
No commercial pressure, no magic of the market, they just mak things up.
P.S. Vote Trump
--
roman_mir
But if no quality reduction it is also a benefit to the user. Why stick to a patent encumbered standard when there's an alternative, just because someone is making money (which will happen with either standard)?
The problem is, HEVC is expensive. While the MPEG-LA made h.264 the way it is by making streaming free (if the viewer doesn't pay) and offering caps to the maximum license fees, thereby encouraging big users (Cisco, Apple, Microsoft, Netflix, etc) to switch to h.264 and merely pay the cap every year
Of course, MPEG-LA wanted to encourage the switch to HEVC by offering the same terms, and several patentholders balked which is why they pulled out of the MPEG-LA pool and created the HEVC Alliance which licenses without a cap, without free streaming (they want some money per HEVC stream), meaning the money you save in bandwidth might go straight to licensing fees.
And I'm sure the BBC streams under h.264 were basically cost-free since the streams were available at no charge (granted, you paid with your TV license, but the MPEG-LA doesn't count that), so switching will create costs.
Cannot be done in anything close to real-time, not with anything close to these results. The thing about this is, patent-holders are tired of having to "give h264 away for free" and are doing a GWX, and any cost. Suck. It. In. and Swallow.
Notice how NO PLACE gets slashdotted anymore?
Your paranoia is showing. The BBC reported the real difference between these codecs, and your first thought is that they're biased. How can you engage with the world at all while this far gone?
s/BBC/Fox/g
And your response would then be?
I've read many many posts indicating that for low bitrate stuff, 265 is doing very well but for the higher end stuff, it's only marginally better than 265 (more like 25%)
That's an approximation, I don't recall the exact figure, but I do recall it being significantly less than 50%
This is pretty disappointing, perhaps other tweaks and improvements will be eeked out over time, but as it stands, I can't see a 3TB movies folder being recompressed to 1.5TB with the same quality at this point in time.
(I know that would be lossy to lossy and stupid, that's not the intention of the post)
There seems to be this misconception that because the bbc is publicly funded it seeks to do everything "in the public interest " and generally doesn't behave like any other large corporation.
The truth as I see it at least is that there are a lot of highly paid media types with one eye on their careers who justify every decision using exactly the same metrics as the rest of the industry. The bbc news website is turning into a clickbait riddled buzz feed clone exactly because they judge success in terms of uniques and audience retention as if they were selling ads. TV programs are similarly judged on audience share rather than quality. This led to a major climb down a few months ago when someone sufficiently senior pointed out that spending 10's of millions trying to compete directly with ITVs Saturday evening prime time offering was probably not an effective use of what is essentially public money.
Similarly the website has come under fire for having a huge and poorly defined scope, along with costing a fortune to run. Being able to point at these figures and saying "but look what we saved on bandwidth!" Is either going to justify someone's continued employment or allow them to push through their 8k@120hz resume padding vanity project.
Right. This is why I think VP9 actually could win and become the new standard (replacing H.264).
H.265 and VP9 seem like they are definitely in the same ballpark on quality. And H.265 is heavily encumbered with patents; you have to pay royalties, and you never know what the royalties might cost in five years. VP9, on the other hand, is simply free: no royalties, no restrictions on what you may do with the video.
Even if VP9 takes a lot more CPU time to encode, and even if H.265 is slightly better than VP9, not having to pay royalties (not even having to keep track of what you do with the video!) is such a huge benefit. It seems like a no-brainer.
And Google will be making sure that all the Android phones at least will have good hardware support for VP9 decoding. VP8 never had a chance against H.264 because the hardware support wasn't there, and large companies were content to pay the capped fees as you noted.
All that's left is possible legal FUD around VP9, but even that seems pretty cut-and-dried to me. MPEG-LA tried for something like a year to find patents to put into their patent pool to extract royalties from VP8, and in the end Google gave them a one-time payment of (to Google) a relatively small amount of money. Thanks to that one-time payment we know MPEG-LA won't ever come after anyone for using a VP8-derived codec, and I have no reason to think anyone would be able to prevail in court if they try it.
Given all of the above, it seems to me that VP9 is the obvious choice for the new video standard, and I kind of wonder why anyone is still interested in using H.265 and paying the royalties.
lf(1): it's like ls(1) but sorts filenames by extension, tersely
Otherwise known as "complete bullshit."
Name another media company that went out of their way to develop a patent-free media codec that was independent and competitive with other codecs of the time?
All the companies involved with the Alliance for Open Media and NetVC.
Lawyer up.
"Win treats sysadmins better than users. Mac treats users better than sysadmins. Linux treats everyone like sysadmins."
You think they just get paid by what's left over from funding after expenses?
Do you think that is how normal for profit enterprises work????
Troll is not a replacement for I disagree.
Because the set of subsisting patents is so large that you can't prove a negative. It costs a substantial chunk of change to deploy set-top decoder hardware for a new data format. Once hardware implementing a decoder for an allegedly royalty-free format has been shipped out, patent holders can come out of the woodwork and claim that their patents apply to the new format. With a patented format, on the other hand, it's more efficient for a patent holder to just join the existing patent pool.
There's also Daala in the pipline, which uses entirely new compression techniques that are not patent encumbered.
"How can Fox engage the world when its that far gone?" :)
The BBC are publicly-funded, and under immense pressure to justify their funding at the moment - there's talk of scrapping the TV licence, and with it the BBC.
That's complete bollox, mate. It's been that way since the 1970s, and they only change is that they get more money to expand into areas they have no need to bother with. They have dozens of staff just responding to twitter for fuck's sake! Then there are the teams that reformat reddit fluff and red-top speculation and celebrity worship. This is the crap that needs cutting, and it needs cutting immediately.
The next step will be to close the iplayer loop-hole and ensure those not paying the TV tax have to start paying.
Like all large organisations, they do one thing, and that's expand. They are not a private business, they are not on the stock exchange and do not have a board and shareholder to respond to. It's about time they were pulled into line and all those filler jobs cut.
Furthermore, their IT people are morons. Their content is delivered through flash, not open formats, and they force OS upgrades to continue to access programming we've paid for, just because they're a bunch of Apple zealots and want to force people to buy new fucking devices.
I don't see any mention of how motion compensation and other "smart" features of the display devices influenced the perceived quality. Software in these devices may work in both positive and negative ways.
H.265 vs VP9 seems like BetaMax vs VHS. H.265 may be better but by being more expensive its adoption may well be lower which would give VP9 the clear shot at the title. Best doesn't always win.
Yeah, I heard this same bullshit about Theora/VP8.
Is there something intrinsic with all these codecs that Intel Quick Sync can do in hardware (CPU instruction set)? Or does the actual CODEC have to be baked in hardware?
Wikipedia did mention something that I wasn't aware of, but makes sense in hindsight: "Quick Sync, like other hardware accelerated video encoding technologies, gives lower quality results than with CPU only encoders. Speed is prioritized over quality.[6]".
I'm guessing people looking for the best quality won't be using Quick Sync after all, or rather haven't. This must be for the average person whom takes video with a cell phone, edits it on a computer, and posts to FB or YouTube.
Life is not for the lazy.
He's not talking about HVEC, he's talking about something completely different but similar. Essentially saying they've lied in the past about the same topic and they cannot be trusted when talking about visual quality.
You are aware the current DG is a Director of HSBC, a Tory party doner and a personal friend of David Cameron. I don't trust the BBC to tell me the time anymore
Build a Man a Fire, and He'll Be Warm for a Day. Set a Man on Fire, and He'll Be Warm for the Rest of His Life.
Switching from MPEG-2 to H.264 really is about a 50% savings. If your video looks bad at 18Mbps MPEG-2, you need better hardware. It won't look great, it's almost as much as the ATSC standard. It's not as bad as satellite TV that allocates less than 5Mbps H.264 to HD channels.
None of it's anything compared to the 20Mbps+ H.264 used on Blu-Ray.
I prefer to watch my videos on vinyl.
If I recall correctly Google also held some important media-related patents from its Motorola purchase that allowed Google to just basically cross-license the H264 patents, the small fee being for the much greater number of patents held by the MPEG-LA. These patents are besides the cellphone-related patents that Google wasn't able to enforce because they fell into the FRAND category of essential industry patents.
It went to Eton and then it was eaten?
Confucius say, "Find worm in apple - bad. Find half a worm - worse."
I haven't read the HDCP agreement myself, but I wouldn't be surprised if it forbade modification of the video signal.
Even if this is not forbidden, a DRM scheme operates on the principle that implementations are assumed noncompliant until proven compliant. The fixed costs of an equipment producer joining a DRM scheme include the cost of an audit (possibly probabilistic) of the compliance of the equipment producer's product on behalf of the scheme's maintainer. So even if it were compliant, whether a device is compliant or not does not matter if the prospective manufacturer cannot draw enough investors to make the device exist in the first place.
I haven't read the HDCP agreement myself, but I wouldn't be surprised if it forbade modification of the video signal.
You can definitely modify the video signal - http://support.xbox.com/en-US/...
Looks like there are HDCP certification services that will test your device for under $5000 (Simplay Labs). I assume they would be liable if the device was actually non-compliant.
Skewered
The truth as I see it at least is that there are a lot of highly paid media types with one eye on their careers who justify every decision using exactly the same metrics as the rest of the industry.
Two eyes.
When all you have is a hammer, every problem starts to look like a thumb.
the money you save in bandwidth might go straight to licensing fees
+1 Nailed it. And since bandwidth always increases while license fees usually do not decease, the deal gets worse the longer you play this game.
When all you have is a hammer, every problem starts to look like a thumb.
What does that have to do with Dirac?
I once wrote an H.261 video codec. Much more primitive than the current stuff but it gives me an idea.
There is a lot you can do to accelerate a video coder without baking the whole thing into hardware. For decode, a fast Inverse Discrete Cosine Transform (IDCT) done by a DSP chip or whatever would really take a lot of load off the CPU. H.261 is mostly IDCT plus video "mixing" to paint new, decoded image pieces over old ones in the reference buffer, and filtering. I just looked at the Wikipedia page for H.265 and it seems that it also has additional filtering modes and can use Discrete Sine Transform (and I couldn't tell you why you would use both DCT and DST, sorry).
You should be able to use the stream processors in a graphics accelerator to speed up decoding of any kind of video, even H.265. Per Wikipedia, H.265 was deliberately designed to be parallel decodable, so even just splitting the work to multiple processors of any kind should be a speedup.
For a better answer than this one, ask someone who is more up to date than me. Sorry I can't give a better answer, but I waited a few days and nobody else stepped up.
You could ask this question on StackOverflow and it seems like you could get a good answer there. It's not even opinion-based so it should still be permitted even though SO is really getting anal about which questions will be smacked down and closed.