New H.265 twice as efficient as H.264!

[lobosrul]

Originally Posted by Penton-Man

Lobos Rule!

Wolf Howling - YouTube

C’mon, lobos have got better hearing than that!

Hehe, my handle actually refers to these Lobos: http://www.golobos.com/ We did not rule at all last Saturday

In short, a subjective increase in video quality is.... subjective.

[Penton-Man]

Originally Posted by lobosrul

...In short, a subjective increase in video quality is.... subjective.

Yes, but it is an entirely accurate parameter to evaluate the HVS (human visual system) as long as the test methodology is not flawed…and editorial review by peers, of either journal article, has found none.

Unlike flawed ‘objective’ measurement which is touted as accurate, like, for instance, that used as a basis for this tech journalist’s *insight*….http://reviews.cnet.com/8301-33199_7...vs-are-stupid/

namely… “but the average person with 20/20 vision can resolve 1 arcminute”,

ergo..“One arcminute of resolution is a best-case scenario."

lah-dee-dah-dee- dah with the math exercise from there.

Which, b.t.w., forms the basis of this chart which is touted on internet forums as a Commandment, rather than just a rough guideline for 4K viewing… http://carltonbale.com/1080p-does-matter/

When the fact of the matter is that this basic ‘objective’ measurement (one arcminute of resolution) is incorrect/limiting in its applicability to real world viewing because other than “resolution”, professional imaging scientists know and hold valid that there are other types of visual acuity which also play an important role in human vision, namely ‘recognition’, ‘detection’ and ‘hyperacuity’.

I don’t have time right now to elaborate further but, perhaps I’ll save that for another post on a rainy day.

[Anthony P]

Originally Posted by 5th Of November

If this only needs half the bit rate, then wouldn't it allow 4K enough space to fit on a 50 GB Blu-ray?

don't know what "half the bit rate means" but 4k is 4x the resolution.

[Anthony P]

Originally Posted by vargo

I don't know if adding to Bluray specs is the correct term. 4K using H.265 is a as different from Bluray as Bluray was from DVD.

that would be like saying 3D BD is as different from Bluray as Bluray was from DVD

obviously the larrter is not true, it was added to the BD format and now we have 3D BD.

[Anthony P]

Originally Posted by Penton-Man

Yes, but it is an entirely accurate parameter to evaluate the HVS (human visual system) as long as the test methodology is not flawed…and editorial review by peers, of either journal article, has found none.

I don't fully agree, there is subjective and objective but there is also ignorance and knowledge. I never trust early subjective results not because there is anything wrong with it but usually when you change something (resolution, compression schemes, encoders....) things will be different and so people are ignorant. What I mean is that my guess with H.265 they found ways to improve issues that existed with H.264 (that is usually the definition of "efficient"). Anyone used to H264 (and good for such a test) will be knowledgeable with PQ issues of h.264 over compression, now if h265 fixes those issues but causes issues else where (can't pay Peter without robbing Paul) the person at this time might not have the experience to judge them as critically.

[vargo]

Originally Posted by Anthony P

that would be like saying 3D BD is as different from Bluray as Bluray was from DVD

No, not all all.

My point was that 4K H.265 on a BDXL disc is technically feasible. 128GB is more than enough.

But even if you are still reading a Bluray disc with a blue laser, it is very unlikely such a format would be called 'Bluray something' (ie Bluray Ultra etc) or added as an extension of the Bluray standards. Backwards compatibility is not possible. It is too technically disparate.

You would start fresh with new branding and new specifications.

[space2001]

When sony releases they 4k streaming service, I was told anyone who buy's the blu-ray will be able to unlock the 4k movie from the streaming device. They don't' know how they will implement it yet but that is how it is for now. Also it will only be sony movies for now, unless other studios sign on.

[Penton-Man]

Originally Posted by Anthony P

I don't fully agree, there is subjective and objective but there is also ignorance and knowledge. I never trust early subjective results not because there is anything wrong with it but usually when you change something (resolution, compression schemes, encoders....) things will be different and so people are ignorant. What I mean is that my guess with H.265 they found ways to improve issues that existed with H.264 (that is usually the definition of "efficient"). Anyone used to H264 (and good for such a test) will be knowledgeable with PQ issues of h.264 over compression, now if h265 fixes those issues but causes issues else where (can't pay Peter without robbing Paul) the person at this time might not have the experience to judge them as critically.

Anthony, now that’s funny.

What I mean is that according to all the marketing wanabees on the internet who fancy themselves as prognosticators of the ultimate success, or failure, of 4K adoption by the consumer-at-large… the feeling is that in order for 4K to truly succeed amongst the masses, the increased clarity or sharpness, if you will, must be apparent to the most inexperienced observers…not only videophiles or experienced imaging professionals who can identify things like aliasing, for instance. That’s been one of the premier arguments, i.e…Joe6pack has to appreciate it.

Anyway, if you are skeptical of the laboratory findings of some of the best imaging minds in the business as regards to their methodology related to subjective response, then I can tell you that all ‘experienced’ viewers (searching for compression related artifacts), whom I’ve spoken with, were convinced of the ~ 50% efficiency claim comparing HEVC with H.264/AVC, also by Broadcom, once they moved from the floor….

Getting 4K with Broadcom's HEVC Encoding - CES 2013 - YouTube

to seeing the Broadcom demo in their private booth.

P.S.

Also, for some perspective, if you have a partly philosophic problem with subjectivity testing and methodology in general, then you must be whole-heartedly disenchanted with the .RED codec roll-out, because to the best of my knowledge, ALL their demos have been based solely on public subjectivity and they’ve never posted any objective PSNR comparison metrics, like for instance Qualcomm did nearly a year ago -

http://reviews.cnet.com/8301-13970_7...n-h.265-video/

[Anthony P]

Originally Posted by Penton-Man

Anthony, now that’s funny.

What I mean is that according to all the marketing wanabees on the internet who fancy themselves as prognosticators of the ultimate success, or failure, of 4K adoption by the consumer-at-large… the feeling is that in order for 4K to truly succeed amongst the masses, the increased clarity or sharpness, if you will, must be apparent to the most inexperienced observers…not only videophiles or experienced imaging professionals who can identify things like aliasing, for instance. That’s been one of the premier arguments, i.e…Joe6pack has to appreciate it.

agree, never meant the opposite.

Anyway, if you are skeptical of the laboratory findings of some of the best imaging minds in the business as regards to their methodology related to subjective response, then I can tell you that all ‘experienced’ viewers (searching for compression related artifacts), whom I’ve spoken with, were convinced of the ~ 50% efficiency claim comparing HEVC with H.264/AVC, also by Broadcom, once they moved from the floor….

I did not mean to question integrity or methodology or even be skeptical ( in the honesty of the findings), it is a simple statement of human nature. A long time ago when I was still active on AVS I decided not to "teach" people how to spot the different artifacts because once you learn how to spot them it is impossible not to do so, in essence it becomes second nature to focus on them. This is one case where I decided "ignorance is bliss" actually holds some merit. In my experience when the game changes a lot (such as much higher resolution such as going from DVD to BD) a better encoder and even a better CODEC it kind of negates the learnt behavior that leads to seeing the artifacts. But as a library of content gets added to the human brains database, it becomes more and more intrusive because you are re-teaching yourself to find the new issues. I am not saying that it is not ~50% but in my opinion only time will tell if it is ~50% (or less or more) and if it is all the better and I hope that 50% is used to bring us better stuff and not only to reduce BW since I am not really interested in BW but in movies.

PS let me put it this way, have you ever met a girl (or even a friend) with some quirks that at first are acceptable or even cute but that after time those same quirks start becoming real annoying? That is what I am talking about, short term and subjective might make something look more acceptable and only time will tell how acceptable the quirks(in human nature) or artifacts (in PQ) will really be.

[Penton-Man]

Originally Posted by Anthony P

PS let me put it this way, have you ever met a girl (or even a friend)...

Shucks, I don’t know Anthony, as there have been so many girl-friends in the past and I was having so much fun ...it’s all a blur now…hard to recall the negative details.

Before I respond to your query, let me just say something for some readers’ perspective….subjective analysis has long been considered of value by imaging people as a supplement to objective measurements….
http://www.itu.int/rec/R-REC-BT.500/en

with the most recent tweaking (approved 1/2012) of the methodology here…
http://www.itu.int/rec/R-REC-BT.500-13-201201-I/en

And you’ll note if you download the pdf http://www.itu.int/dms_pubrec/itu-r/...1-I!!PDF-E.pdf
that it is acceptable if observers are expert….or non-expert (naïve), or a combination of both.

Bottom line for how all this translates in the end…I would say that if investigators have the PSNR objective measurement determined, then, for humans, one can comfortably bump-up that PSNR-rated efficiency, to some degree, because of the inherent nature of the human visual system, regardless of whether ‘expert’ or ‘non-expert’ observers. Humans don’t see like a computer, i.e. in PSNR, because for one thing, of the saturation effect of the human visual system.

You’re right though, how much that objective figure is bumped-up by subjective analysis, may, in some cases, be dependent upon the expertise of the observers. In fact, I’ll make your point for you with a better analogy than the girl or friend thing. Back in ’03, D-Cinema testing was done at DCL in Hollywood, (which was eventually published in SMPTE journal sometime in the autumn of ’04) to determine how many bits (8, 10, 11, or 12) should be used for Digital Cinema (I’m talking theatrical here).

In a nutshell, the test patterns created steps in luminance corresponding 8, 10, 11 and 12 bits. The subjects (75, if I recall) consisted of both ‘expert’ and ‘non-expert’ observers.

Anyway, to make a long story short, to make your point, the subjects, through repeated viewings, became more sensitive and essentially learned to identify more luminance steps. So, in this case, yes, ‘expertise’ over time, made a difference compared to the initial viewing. B.T.W. for those who happen to be wondering, as the result of this testing, D.C.I. adopted 12 bit (@ gamma 2.6).

But, back to the real world. How many ‘experts’ with learned experience for a particular imaging characteristic actually watch movies? For instance, how many theater-goers identified, as a deficiency, the blown highlights ( see – https://forum.blu-ray.com/showthread...ts#post6879108) in the 3D HFR screenings of the Hobbit? A handful….if any?

Heck, even the ‘professional’ reviewers/critics were too consumed with whether or not the 48fps looked too real, or satisfyingly real, for a narrative motion picture. I mean, even if they noticed the clipped highlights esp. at the beginning of the film, but failed to object, it is naïve for them to think that harsh look was an intentional aesthetic choice by the color correcting crew.

[Anthony P]

Originally Posted by vargo

No, not all all.

My point was that 4K H.265 on a BDXL disc is technically feasible. 128GB is more than enough.

But even if you are still reading a Bluray disc with a blue laser, it is very unlikely such a format would be called 'Bluray something' (ie Bluray Ultra etc) or added as an extension of the Bluray standards. Backwards compatibility is not possible. It is too technically disparate.

You would start fresh with new branding and new specifications.

why, didn't they have BD live, 3D BD..... maybe they will or maybe they won't

[Anthony P]

Originally Posted by Penton-Man

Before I respond to your query, let me just say something for some readers’ perspective….subjective analysis has long been considered of value by imaging people as a supplement to objective measurements….

never said objective is the only choice or knocked subjective testing. In the end there is no other real choice. if objective data is too lenient (i.e. since we are talking images a sensor that measures brightness will be useless in this discussion) or too strict (i.e. we tested two audio files and the dog whistle is missing from one of them) it does not say much and in that way our senses are the best test. On the other hand (to stick with an easy audio example again) there are sounds that younger people can hear very well but older people cannot pick up.

But, back to the real world. How many ‘experts’ with learned experience for a particular imaging characteristic actually watch movies?

Don't know. But does it matter? it is a simple discussion, I did not mean to cause issues. I just think the 50% needs to be taken with a grain of salt since time might show it is less or even more. That was all I meant to say. Now if someone out there does not care, or if the people that really care (and know) are few. I agree it is the case but still if the question of "is it the same", to me it boils down to at least one person saying no for it to be no. It is like when people say "who needs ____ audio, see those X people could not correctly confirm the difference", but there are those Y people that did and so they should be the ones to judge.

[Penton-Man]

Originally Posted by Anthony P

never said objective is the only choice or knocked subjective testing....

I realize that , I was not referring to you. I was offering an informational comment to the general reading audience.

[Penton-Man]

Originally Posted by Anthony P

...Don't know. But does it matter?

Well, as a matter of fact, yes it does. Again I’ll refer you to the pdf above…. http://www.itu.int/dms_pubrec/itu-r/...1-I!!PDF-E.pdf

Namely, in the Intro…”Subjective assessment methods are used to establish the performance of television systems using measurements that more directly anticipate the reactions of those who might view the systems tested”.

Let me offer a practical real world example as to why this is important. Highly experienced 3D experts…I’m talking about people who work in/view stereoscopic imaging for hours every day, such as at the studio or post house level, are, by-and-large, in terms of distractive eye discomfort, relatively immune to the appearance of windows violations. Whereas, the vast majority of the general 3D viewing public, whether D-Cinema or home enthusiast, exhibit significant eye strain to the appearance of windows violations, if they are not corrected.

[Penton-Man]

Originally Posted by Anthony P

it is a simple discussion...

Tell that to the ITU membership who, in the past, has had to read the findings/recommendations from study groups like VQEG - http://vqegstl.ugent.be/

[Penton-Man]

Originally Posted by Anthony P

...I did not mean to cause issues.

Not taken that way . Not in the least . And I hope you don’t perceive my responses as causing issues either. That was not the intent at all.

On the contrary, I’ve enjoyed engaging in the conversation because there is very little posted about objective/subjective analysis and the back-and-forth discussion between you and I, has, I feel, contributed to that knowledge base.

But anyway, moving on, especially for those, unlike you , who are afraid or dismissive of 4K…..RELAX, for something which also goes unrecognized and thusly unmentioned by bloggers, is that HEVC (given its inherent bandwidth savings) has the potential to enable 1080p60 broadcasting someday, which would be a boon to HD fast-action sports.

What I mean is that, with HEVC, 1080p60 is comparable to today’s 1080i data rates. And for those content providers who have expressed an initial reservation to 4K, that (doable-to-the-consumer 1080p60) would seem to be of some interest, especially given the fact that “the network is ready for whatever comes next”…
http://www.engadget.com/topics/hd/20...ew-l-a-studio/ or, should be, as it’s been almost three years now.

So, HEVC 4K benefits aside, HEVC can also be thought of as the potential key to unlock the bandwidth bottleneck to the nirvana of HD fast-action sports, namely 1080p60 acquisition and delivery.

[Anthony P]

Originally Posted by Penton-Man

What I mean is that, with HEVC, 1080p60 is comparable to today’s 1080i data rates. And for those content providers who have expressed an initial reservation to 4K, that (doable-to-the-consumer 1080p60) would seem to be of some interest, especially given the fact that “the network is ready for whatever comes next”…

agree (especially if the 50% is right) but the issue with that is HW. Someone that watches ATSC or QAM (cable) directly on the TV would need a new TV or some new converter box with ATSC/QAM tuner. If they have a cable or satellite box it will be useless since it can't do HVEC and so they would need a new one. That is unless you assume that the broadcast distributors use extra channels and then you killed the argument since the extra channels will need more BW (i.e. I don't know what is offered, but to make it clear let's assume the BDU offers ESPN SD and ESPN 1080i60, they will need to continue them for the people that watch SD and HD and add 1080p60 for the people that buy the new box and can handle HVEC, since in the end you can't expect every customer to change their devices all in one day)

[Penton-Man]

1080p/60 (and for that matter, 1080p/50 for the Euros) is of course an evolutionary format which will require an investment cycle in production as well as distribution. But the first ‘issue’, is, and always has been, resolving the bandwidth bottleneck, which HEVC has accomplished.

[img eL]

A h.264 claiming 75% file size reduction http://www.prnewswire.com/news-releases/190832551.html

http://beamrvideo.com/

[vargo]

Originally Posted by img eL

A h.264 claiming 75% file size reduction http://www.prnewswire.com/news-releases/190832551.html

http://beamrvideo.com/

I downloaded one of their samples and it was clearly encoded using x264.