4k, HDMI, and Deep Color
toonormal.com
toonormal.com
HDMI must die.
I suspect this has something to do with HDCP DRM, but I'm not sure if that's the case.
So attempting to send 1080p over a category 1 cable isn't uncommon, and it often "works" with occasional dropouts.
Just adding a ferrite tube at the end of the cable is not effective choke. and a cable without choke, is a interference antenna.
I don't know which of their cables the previous poster had trouble with, but I'm pretty sure he could have returned it for a refund.
(Disclosure: I have no relationship with BJC other than being quite happy with the speaker cables I've bought from them. Which I'm sure the Monoprice fans would tell me I overpaid for anyway, but, well.)
I read that people having an identical issue had better results with shorter cables. I don't know any of the details but the shorter cable had a huge effect.
The problem with Displayport is that the chips to render it are more expensive unless you built your monitor for direct drive. That would be great if every display was designed with the same subpixel specs, but no 2 manufacturers have the same specs for their displays so this doesn't work.
What happens is most DisplayPort displays convert the image and then render it which is also how you allow for brightness, contrast and sharpness adjustments on the display rather than on the device connected to the display.
If you used direct drive you'd control brightness and contrast on your Xbox, bluray, or computer rather than on the display.
(That said I use DP whenever possible)
I'd definitely appreciate that like 8 years ago, but if a device is old enough to worry about that, it still might not even support HDCP, in which case a large chunk of modern media wouldn't play on it anyway.
I don't have an everlasting flame of hatred for it in my heart, but yes, I do find it inane and a bit annoying to have marketing labels for the thing I already bought on the thing itself.
So yeah, in the "real world", DisplayPort doesn't exist outside of last-gen Macbooks.
Beyond your DisplayPort incident, Best Buy in general is a provincial consumer electronics store, with ignorant sales personel and quite narrow selection. It might have been good enough for 1995, but not in 2014.
It's like visiting a Walmart for delicatessen items.
There's a reason your local deli has one store with three clerks and is only open until 9pm, while Wal Mart is bigger than Jesus and open 24/7. We don't live in the real world. The real world doesn't demand 4K with cables that can carry the signal at full color because they have different priorities than we do.
I was using the term "delicatessen" as a quality indentifier, not merely as a product category. Any old cheese and lunch meat won't do.
>Likewise, Best Buy is indeed the real world. Mom is going to go to Best Buy to pick up a new laptop. Dad's getting his new TV there (with Monster cables so he doesn't get an HDMI virus) (...) Because, if Best Buy doesn't have it, there's no one to sell it to me.
Sure, but it's not like it's 1995 and the local Best Buy at Boise, ID is the only option. Tens of millions have an account and order from Amazon and lots of even more dedicated electronics sites.
What's difficult about "going on Amazon and shopping around"? We're not talking about our cousin or dad here, you were relating your story of visiting Best Buy for that cable for your Lenovo. Why even go there?
It makes people have to think, and in a lot of cases people don't want to have to think. I posted something longer on this subject recently:
https://news.ycombinator.com/item?id=7105674
>Why even go there?
Because I needed a cable today, not tomorrow or the day after. Amazon can't match that in my location. That's where I ran into the issues that DisplayPort is not mainstream, and Best Buy only carries things that Uncle Joe is going to need for the TV and laptop he bought at Best Buy.
I understand you used the word deli to mean quality, but there's a reason Apple still has less than 10% marketshare of desktops/laptops. People are more often than not satisfied with "good enough".
My Sager NP8255S has mini, fullsize, and HDMI ports.
It's unfortunate as well, because DP lets you daisy chain monitors together; it's the perfect form factor for notebooks all around. You only have to plug in a mouse, keyboard, and single DP dongle for full power to drive 4x monitors, and only using up 2 USB ports (assuming powered equipment). No messing around with USB hubs running multi monitor stuff with AC adapters crowding the power strip. This gets rid of the need for a dedicated docking unit as well.
Apparently Amazon can now claim cheaper, better selection, and more convenient.
The answer is more color range per pixel.
Chroma is subsampled compared to Luma because that's literally how vision works: the Human eye has far more Rods (brightness sensors) than Cones (color sensors) per unit area. Increasing the color resolution on a TV[1] to match the luma resolution is literally a waste of bits[2]. On the other hand, the color receptors are very sensitive to the exact shade of color[3], so increasing color bit depth is not a waste.
[1] Increased color resolution on a computer monitor is useful because you sometimes closely examine a small section of the monitor -- and possibly lean in to examine it better.
[2] The old NTSC analog system used the same principle: color was encoded using less bandwidth than the luma.
[3] As the author mentions, this is best observed in a gradient where the TV must present the gradations in sufficiently small steps to fool the eye into seeing a continuous gradation.
Doesn't decorrelating the quantization noise (dither) along with appropriate noise shaping make these two equivalent? It's been a long time since I've thought about this in context of video, but it should be possible to directly trade one for the other in the wire format, as far as I remember.
These are important facts for deciding, for instance, that alarm clocks, and shipboard navigation dashes should use only red light- since red light doesn't kill the sensitivity of your rods. It doesn't deactivate your night vision.
That's not to say you're not partly right, it's just the color/luminance processing doesn't happen in the retina. It happens in the visual cortex, or possibly with some help from processing that happens in or near the optic nerve.
As noted in the article, Deep Color is important so that sampling errors do not accumulate while images are filtered, processed & combined. But a television or monitor is the final step, it should be performing very little image processing. (Should is the operative word here, that's not necessarily true).
If you used HDMI to connect cameras to recorders or to connect effects processors, this would be important. Does anybody do this for 4K?
Dithering is certainly a solution, but higher bit depths also works, and in a way that's much cleaner.
EDIT: This implies of course that the final display is also capable of high bit-depth display. Obviously, if the display is limited to 8bit, then it's kinda useless. I think the author's point was that HDMI 2.0 -should- be kinda future proof, and -should- anticipate widespread adoption of high bit-depth displays.
If you have a RGB image were each color component uses a byte to represent it, you should have a effective Luma bit depth of 8 bit. Well, this is false for LCD screens because they have a effective Luma bit depth of 7 or 6 bits (or worst), so even using a VGA signal (analog signal) you get banding on these screens, were on a old good CRT screen you never appreciate it.
Also, are there even 12-16 bit color panels in production? I like to hope 4k / 8k mean the end of the pixel race, because we really need adaptive vblank (from Displayport) of at least 100 hz before we keep pushing the pixel density. Hell, most of my family don't recognize the difference between 480 and 1080p already because the colors are so bad on most consumer tv panels.
I agree that deep color is fascinating, especially in light of the fact deep color for end users has been right over the horizon for so long (I've ranted previously about my disappointment that 24-bit color has been the pinnacle of desktop computing for well over a decade). I personally find 24-bit banding quite annoying, especially with animation and fade effects, which visually exaggerate the limitation.
I would absolutely love a 50" concave OLED display with high-DPI, high-speed deep color. For the time being, in the real world of compromises, I'd be happy with either HDMI 2 or DisplayPort 1.2+, since I am presently dealing with 30Hz at 4K. Given that GPUs available today support 4K at higher refresh rates on DisplayPort, my current very slight preference is DisplayPort over HDMI 2 (which as far as I know is not supported by any GPU I can buy today).
First and easiest. Anything that you can display over DVI can be displayed on a display that has implemented DVI over HDMI, so DVI doesn't look better than HDMI for any technical reason. This is how all those non-standard resolutions over HDMI work, and allows for color combinations that are not part of the standard.
Now that is established. 4k doesn't actually require HDMI 2.0 to work because any combination of resolution and color that is supported by the "speed" of the wire and can be communicated in the "handshake" between devices can work now.
Unlike most standards HDMI has "requirements" for certification, but those are not the upper limits they are the lower limits.
So 4k 24 FPS at 4:2:0 could be negotiated on an HDMI cable that is 1.1 compliant.
Author should read less Wikipedia. When I was at Microsoft's Media Room I had to have IT block Wikipedia because engineers were getting so much wrong information from its articles. Read the standards, talk to device manufactures. Wikipedia is not a place for deep tech, it is a place for a quick answer. If it isn't a knol of data like a stat for transfer speed, assume someone like this author wrote it, and it will be inaccurate.
That sounds like the worst engineering environment I can imagine. What happened to managing one's staff properly?
Seems like the ideal display can:
1) match the dynamic range of the real world in luma and chroma 2) resolution that makes seeing pixels very hard 3) enough precision/gradation within the dynamic range to avoid banding 4) support high framerates
1. The CIE XYZ color space is different from the CIE RGB color space.
2. "Dynamic range" is not about gamut but the ratio between white and black. A monochrome display can have high dynamic range.
3. Neither the CIE XYZ nor the CIE RGB color space correspond to human vision. CIE RGB color space uses three monochromatic primaries (700 nm, 546.1 nm, and 435.8 nm) and so there are visible colors that it can't represent without using negative numbers (such as the blue color of an Argon laser). The CIE XYZ space is a linear transformation of the CIE RGB space designed to represent all visible colors without using negative numbers, the trade-off is that most of the colors in XYZ are imaginary -- they cannot be perceived by humans, recorded by cameras, or displayed by monitors. The color space closest to human perception is CIE LMS, which still contains many imaginary colors.
If you want better colors look no farther than Rec 2020, which is probably going to come around soon. It still won't represent Argon-laser-blue but the cost of representing all colors on a real, physical display is prohibitive (to the point that no prototypes even exist). Agreed that sRGB is a rather small color space but improvements on color representation have to take into account the engineering needed to make it happen. The question, "Which primaries should I use?" was thoroughly explored during the development of Rec. 2020 and I suggest you read their rationale.
But yes, you are right about the other stuff. But I'm not concerned about what can be achieved with real physical display panels today. I was talking about idealism, not incrementalism, I want to look at a display and almost not notice it's there, like looking through window on my wall. is Rec 2020 going to get us there?
1. Natural contrast ratios are so far beyond the limits of current display technology that it would take a complete revolution in material science just to figure out what to make the actual screen out of.
2. Natural colors are have a gamut that cannot be reproduced using the fixed primary model. You would have to do something like put a dynamically tunable laser inside each pixel.
It's like asking for a USB port on your computer in 1949.
However, if you're prepared to make some small compromises:
1. Okay, contrast ratios are limited by glare from ambient light.
2. Okay, we'll use three primaries: red, green, and blue.
THEN, Rec. 2020 is what you want.
> "dynamic range" (DNR) is not about white and black, it means the ratio of the largest/smallest values of a changeable quantity.
In the context of displays, the changeable quantity is the amount of light. White is the name of the largest value, black is the name of the smallest value.
But just as we cannot perceive individual frames at 24 Hz, but instead see motion, there's still an emergent effect that is perceived when we juice that frame rate up to 60, or 120, etc. There exists a frame rate where the perception changes from "really fast slideshow" to "motion" and that value is pretty low (something like 12 Hz?). Yet, higher frame rates still deliver a noticeable change in the quality of the video.
I think the same effect will apply to very high pixel densities. Moire effects that occur when the news anchor is wearing a striped jacket, for example. There's gotta be other emergent qualities that come out of having such a tight pixel density, even if no single pixel can be discerned at a 3 meter distance.
(Disclaimer: the above is pure conjecture, I have not compared a 4K display to an HD one from normal viewing distance)
I think there is (at least) one factual error in the blog post; 16-235 range was not chosen because of CRTs: http://en.wikipedia.org/wiki/Rec._709#Digital_representation
In that light I'd also claim that the expansion to full-range of values is probably one of the smallest advantages of xvYCC. The wider gamut is significantly more important.
Also if when watching video files you notice that blacks do not look very black, then you should change your video player. As the aforementioned wikipedia article says, value 16 is intended to represent pure black in Rec709 colorspace.
Chrome subsampling goes a long way to explain the issues I was seeing with the Seiki 4k TV at http://hardforum.com/showthread.php?p=1040546819#post1040546....
Can anyone explain why we have this disconnect between video/monitor standards and bitmap/png standards? What would be the problem with doing 8/10/12 bit RGB front to back, or YCbCr front to back?
The reproduction equipment has no need to support 240fps, 36bit color, 12k resolutions... that would be a total waste of money on hardware. But when editing the media, the extra quality is useful.
Audio does the same thing. Most music is 44.1kHz sample rate, sometimes 48kHz. Studios might record in 96kHz or more; the human ear can't appreciate 96kHz and most hardware can't play it, but when the mix it, the result has fewer distortion/artifacts.
I would tend to say higher resolution is more worthwhile. I obviously don't have a way to test this directly, but my guess is that 8k at 180 fps and 2-bit, using spatial and temporal dithering, will look better than 4k at 60 fps and 12-bit, even though those two signals would contain the same amount of information.
The reason is that the dithering allows the same color depth to be presented, but the discreteness is kept well beyond the boundaries of what your eye can pick up.
Obviously there are plenty of technical reasons that this isn't exactly practical.
4K at this point has little point, but the increased resolution will be nice once the quality catches up.
Added resolution has the nice property that it can be traded for color depth by the use of dithering.
Is this technique used anywhere?