With television slowly transitioning to 10bit video and new codecs,

because muh mobile first

Because people don't even know what color accuracy even means. He'll Android still has no mechanism for color correction via the use of ICCv4 files.

Because it makes fuck all difference.
High bit depth for imagery is like high bit depth for audio. Only really important when processing, and avoiding banding when using high levels of compression.

I don't think 4k will really take off.

HDTV in my country boomed because it was marketed just before my country went into recession. People were spending money they didn't have on shit they didn't need.
Now things are a lot tighter, and the improvements are less pronounced anyway.

Most people can't see the difference anyway. When I was in America a couple of years back, I saw multiple televisions with subscription tv boxes connected to their televisions with composite cables. They had them set to centre crop to 4:3, and the televisions set to stretch that picture to 16:9.
If you can't see that shit is fucking stretched, you can't see that it's 4k.

>>58675535
readability is more important than color accuracy on a smartphone.

hence qualcomm developed assertive display technology

Are those 10bit TN panels any good?

>>58675818
Most TNs can't even do native 8bit without resorting to tricks

Is OP an ignorant faggot?

DSLRs have been capturing in 12/14bit per colour for a long time now.

>>58675499
>why has digital photography and imagery on the internet similarly not improved?
What are you talking about? Almost all good digital cameras can shoot in Adobe RGB colour gamuts for many years already.
The rest does need to follow.

>>58675552
Our eyes are better than our ears. 8 bit colour depth is far from enough and colour banding is still a pretty big issue.

>>58675862
those are old TNs

>>58675875
This.
Any good SLR will shoot higher than 8bit if you shoot RAW. Even my entry level Nikon can manage it.

The main problems with photography are in distribution and processing.
Only a minority of photographers process their images at 16+ bit with a wide colour space. An even smaller minority distribute said images in such a format.

>>58675876
8 bit is sufficient for uncompressed images. Where it becomes trouble is with compression.
Compression ruins the dithering. The exact same is said to happen with 16 bit audio, although audio compression is comparatively more 'progressed' than image compression. The image standard that is still most widespread was standardised in 1992, and there have been few attempts to replace it with anything better.

As for Adobe RGB, the colour space used in 4K video (Rec. 2020) covers a greater amount of colours, but has issues with banding if you use it at a lower bit depth like 8 bit.


As much as I'd protest how bad it is that nobody uses any of this stuff, I doubt it'll change any time soon. People are more than happy with "good enough". We still have images on websites that are rarely taller than 400px.
Most of the differences aren't noticeable to the average layman.

File: The Last Guardian™_20170116181843.png (2MB, 3840x2160px) Image search: [Google]

2MB, 3840x2160px

>>58676126
>8 bit is sufficient for uncompressed images.
How incredibly wrong. This becomes painfully obvious when you try any game that supports HDR on PS4/XB1, and save a screenshot in 8bpp from it, since that doesn't include the dithering and different colour space they have for the SDR mode.

>Compression ruins the dithering.
...and this contradicts what you just said about 8 bit being sufficient. Dithering itself is a hack to make up for 8 bit being insufficient.

HDR is going to be adopted much like 4K, as these are just incremental upgrades with no downsides, but it'll take a few years.

>>58676178
For my next upgrade I'll just wait until OLED monitors with HDR, 4K, and high refresh rate pop up in the next decade. For now 4K IPS is sufficient.

>>58676476
If you have the money to burn, a 4K HDR OLED TV is already a pretty great option. Nothing over 60Hz just yet, but I give that a year or two.

>>58675499
because there is "no need" for it

you see, you can send your mom all the funny pictures and powerpoints
"mom you need a 10bit screen to see it in all its glory"
"looks good to me"

I'd say it's an equally slow transition in the computing world, maybe some day, when 10bit monitors become affordable or it somehow becomes the new standard, I might switch

>>58676706
Well sadly no I don't have money to burn quite contempt with my current setup for a long time until a worthy upgrade option presents itself.

144Hz would be nice for watching 24Hz video and animation most of which is set at 720 or 1080 16:9, 4K scales well with both however if 1440 content becomes a thing it won't play nice scaling wise.

Was tempted to get a 1080 144Hz TN monitor but 4K 60Hz IPS sounded like a better upgrade path overall despite a few nuances with UI scaling on certain programs here and there, it seems to have better support than 144Hz currently. Too many issues with high frame rate on older content . Some older games tie physics to frame rate or cap at 60Hz.

IPS shits all over TN viewing angle wise which is a big deal to me since I often sit at odd angles when watching stuff you even notice shift in colour on large enough TN panels when looking dead on, but there is backlight bleed on IPS so you can't win em all yet.

Hopefully burn in on OLED would be resolved otherwise it would be perfect.

>>58676178
Dithering doesn't happen if you take an image natively at 8bit. It's essentially an artificial quantisation process, which is a process which naturally occurs if you take an image at any bit depth, because a digital representation of an image will never entirely cover the complete continuous range of values which exist in reality.

If you are seeing banding in an uncompressed image which has had it's bit depth reduced (like the sample that you posted), that very likely means that it wasn't dithered at all. Dithering isn't a bad thing. It has to happen whenever you try to reduce a range of values into a smaller range.
You still need to dither going from 32bit -> 24bit as much as going 8bit -> 256 colours, and a 10bit image provides little benefit to an 8bit one when the values are properly quantised.

I'm not even trying to say that we shouldn't progress to 10bit for images, but it shouldn't be a priority, particularly when there are other issues when it comes to image handling and distribution.

>>58677233
thats literally the argument for 16 bit audio vs 24 bit audio

>>58676972
>144Hz would be nice for watching 24Hz video
It's a shame computer monitors aren't like european tvs.

They switch between 60/50hz (and these days 24hz as well) depending on the source content.

>>58675499
because webdevs are retarded

File: 1481793423196.png (42KB, 500x322px) Image search: [Google]

42KB, 500x322px

>>58675629
>when a family member using set top box scaling

File: 500full.jpg (9KB, 250x305px) Image search: [Google]

9KB, 250x305px

>>58675499
>With television slowly transitioning to 10bit video and new codecs


What is the point of 10 bit when all the content (including UHD Blu-ray) store the video signal in 4:2:0?

8bit 4:2:2 or 4:4:4 > 10bit 4:2:0

Thanks, chroma subsampling

File: 1321062489442.jpg (47KB, 362x565px) Image search: [Google]

47KB, 362x565px

>>58679941
Also, most UHD content is mastered and released in REC.709

>>58675499
Digital photography was the first to implement higher bit rates per channel. My camera shoots 14-bits per channel.

>>58679986
>My camera shoots 14-bits per channel.

42 bit?

>>58679995
In RAW format. 12-bit is common too. Kinda surprised /g/ isn't aware of this.

>>58679995
RAW photos usually only have one channel per pixel, the other channels are interpolated (when rendered at native resolution).

>>58680049
I think it's actually the way the sensors work. The RAW image has three different channels which are interpolated and rendered when you import them into lightroom or whatever program you use. Bayer filter is what they call it.

>>58680180
The data in a raw file is not that different from RGB data. A Bayer filter is just like a grid of RGB pixels.

>>58675499
speaking of color accuracy and correctness, you're using a CIE1931 diagram that shows incorrect hues. 520nm should be cyan, not green.

>>58675499
Because most people are still stuck with shitty 1080p or 1366x768.