Slashdot Mirror
HP Introduces First-Ever 30-bit, 1 Billion Color Display
justechn writes "I recently had the opportunity to see, first hand, HP's new 30-bit, 1 billion color LCD display. I have to say I am impressed. Not only is the HP Dreamcolor LP2480zx capable of displaying so much more than standard LCDs, but it considered a Color Critical display. This means if you work with videos or photos you can be guaranteed that what you see is what it is supposed to look like. With 6 built-in color spaces (NTSC, SMPTE, sRGB, Rec. 709, Adobe RGB and DCI), you can easily switch to the one that best suits your applications and process. At $3,499, it is too expensive to be a consumer level LCD, but compared to other Color Critical displays (which can cost as much as $15,000 and $25,000) this is a real bargain. This display was a joint venture between HP and DreamWorks animation. When I talked to the executives of DreamWorks, they were very excited about this display because it solved a huge problem for them."
I WANT IT. I don't really know why, though...
"Most people, I think, don't even know what a rootkit is, so why should they care about it?"
http://gizmodo.com/5014879/hp-dreamcolor-lp2480zx-shows-off-its-one--billion-colors
It doesn't look like anything special to me. I guess I don't need to upgrade my current monitor.
This issue is a bit more complicated than you think.
It might be better to avoid stories from people (justechn, roland p, etc) that just link to their websites. Especially those that require registration.
Slashdot should not be giving these guys (and their like) the free publicity that they figure they deserve.
Lots and lots of links for your perusal. Google makes all computing simple
They make it sound like out-of-the-box you're going to get the best image possible. But that's not the case. The color profile for the monitor needs to be adjusted to match reality (using something like ColorVision's Spyder2)before you can make that claim. There's no point in having billions of colors if they're all wrong.
I don't particularly want 1 billion colors, I actually just want 1 new one: black.
Not a very slightly gray-black, but silver-print-face-of-the-half-dome black.
"Sacrifice for the good of The State" - The State
That's not quite right.
CIELAB colour space codes colours as L (lightness) with a 0 - 100 range, and a/b (red-green / yellow-blue) each with about a +/- 100 range for physically realizeable colours. A pair of colours which are just distinguishable are a unit apart, so we can distinguish very roughly 100 * 100 * 100 colours, or a million.
However those are surface reflectances under a single illuminant. In a natural scene, your eye is adapting constantly as you look around. Your iris changes size, your retina changes sensitivity, and so on. The range of lightnesses in a natural scene is up to about 10 billion to 1 if you compare direct sunlight to deep shadow. You can distinguish a million colours at each of these points of adaptation.
If you want a display that can show a full range of dark colours and a full range of light colours, you need more than a million to 1.
And I, as a man, fear anything and anyone that can handle more than the 16 colors I can differentiate and all the marital skirmished derived from that fact.
My 0.02 cents
1) Open photoshop.
2) Make a gradient from 0-0-0 RGB to 255-0-0 RGB. This covers every possible variation of the red channel in a 16.7 million color space. Draw the gradient across your whole screen.
3) Look at the color banding and say, "Oh, I guess I can see why 30 bit color would be noticeable."
"I zero-index my hamsters" - Willtor (147206)
This display might work for reliable color matching, but not for the reasons supplied.
The main problem with getting color on one object, say a display monitor, to look exactly the same as on another object, say a magazine page, is mostly the problem of gamma, a nonlinear contrast range in different light levels. And, of course, the differing illumination of the two objects in different places, which is the actual source of the possible range of colors that can be seen coming from the object.
The human eye is very sensitive to different spectral content of light detected coming from objects. Sunlight starts out with different colors than the light shining on a display monitor or generated by the display. The magazine in the sunlight filters a range of colors through its ink, then reflecting off the paper (which is itself some color, even if that color is "close" to "white"), back through the ink, and to the eye. The display monitor's light starts out a different color from the sunlight, then is filtered through and reflected from very different materials than ink and paper. By the time the light reaches the eye from each object, they're very different. And each instance is a little different, owing to manufacturing quality variations.
And then gamma has to be factored in, which tends to dominate the color content reaching the eye. The gamma is a kind of nonlinear "contrast" (as in a TV control) in different frequencies, varying as the intensity of the same illumination is increased. But even that illumination generally isn't just the same color at all intensities, because it's emitted from some manufactured material that has its own gamma (or emission equivalent) and "color temperature" bias. Which is in turn different from sunlight, which is more stable in its source color range than most manufactured materials (except lasers, a completely different kind of illumination that looks completely different from sunlight).
Color calibration works best when there's a feedback loop of the data passed between different output objects (like paper/ink and a display monitor), linked by a video sensor (that has its own color calibration problems). It's an extremely hard problem. When I was a member of the Joint Photographic Experts Group (JPEG, who created the image file format - I helped with the color spaces spec), we spent a lot of time getting it close enough for commercial use. But we knew enough to tell that "solving" the problem 100% was not going to work. And even now, almost two decades later, it's still not solved. But every few years new tech makes it affordable for industries to add another "9" to what was once 99.999% accurate. The 30 bit gamut of this display monitor means that it doesn't constrain the range of colors as much as have old technologies. But the calibration requries sophisticated processes and software to automate them, as well as a method for comparing to actual outputs. And it still can't account for variances in manufacturing the target output media.
For Hollywood, this problem might be close to solved, though. Because movies are moving to digital projection, which can be manufactured to high precision of consistency in materials and their interaction with light, and from the same parts as the production display monitors. If all the theaters used the same DLP chips, LEDs and image surfaces (or to the precisely same standard specs) for their projectors as the studios did for all their display monitors and as all people did for their home TVs, then colors would be pretty close to identical in all those environments (except for that variable ambient lighting). These display monitors might flexibly replicate a lot of different environments to match, but the matched objects are still highly variable. For $3500, they better deliver something good.
--
make install -not war
One... BILLion colours...!
If you mod this up, your slashdot background will turn into a beautiful sunset!
I bought a 50" plasma some years ago, and was showing a few of my friends SDTV channels versus HDTV channels. Now, this was a very high-end plasma, properly calibrated, showing some of the prettiest content on Discovery HD, so we are talking a KICK YOU IN THE FACE improvement that anybody with half a brain should have been able to appreciate.
One was suitably impressed. The second said that she could kind of see a difference, but didn't really care. The third said she couldn't even tell.
I suspect these are the same people that buy a nice 24" LCD and then run it in 800x600 resolution. Sadly, I have seen this. After fixing it, I have then seen these same people maintain that aside from the aspect ratio change, they couldn't tell the difference.
Evidently a lot of people desperately need glasses and have absolutely no idea how bad their vision is. The weird part is that even when this is pointed out to them -- "Wait, you seriously can't tell the difference between 800x600 and 1920x1200? Please, for the love of Zeus get your eyes checked!" -- they generally act completely nonplussed and never bother to see an optometrist. I just don't get it. Why do so many people not care about having sharp eyesight?
ZFS: because love is never having to say fsck
If you take a photo of the sun and look at the image on this monitor, you can blind yourself.
Because to not do so is problematic for the computer which is controlling it. There's also the issue that what we REALLY see the best is greys. If you have a different number of bits per channel, you'll run in to the problem of not being able to do truly neutral greys (as was a problem in 16-bit 5-6-5 colour mode). Because of our grey perception, there's already been 10-bit black and white medical displays out there. Finally, it would be silly to artificially cripple the display.
LCDs function by filtering light through red, blue and green filters, and then blocking part or all of the light to specific sub pixels. So if you can have 1024 driving levels for one sub pixel, you can have it for all of them. No reason to restrict the pixels that happen to have red and blue filters instead of green.
So this display is 10-bits per primary colour channels, giving 1024 steps for grey, 1,073,741,824 total possible different colours.
Frankly, most daily tasks don't require good eyesight. I don't even bother wearing my glasses unless I'm reading signs or driving or something. And my level of eyesight actually requires correction; a lot of people have less-than-perfect eyesight that's still legal to drive with.
When I go to the movie theater or watch a DVD on a big screen or something (if I'm watching on my laptop, I can already see every pixel at a comfortable viewing distance), I do put on my glasses so I can enjoy the sharpness (if it's that sort of movie; some movies are better without being pixel-perfect sharp).
However, for everyday life, it provides marginal benefit. And corrective lenses inevitably introduce other kinds of distortion, which I find give me a headache. Certainly if I want to make sure something is straight and level, I take off my glasses, because I can't trust my lenses to match what my brain has been wired over the years to perceive as straight.
We *DO* have very strong sensitivity to greys. But that mostly happens in our peripheral vision. Our foveolla is richer in cones, rather than rods and thus has very big colour sensitivity, but sucks at distinguishing very dark levels of grey.
This can easily be illustrated when looking at the sky, at night, when there are no cloud and no light pollution from a nearby big city : you see a lot of stars (when getting a global picture with all your visual field including peripheral vision) but if you try to look at some region in detail, some star seem to disappear (you're looking it with the high resolution / high color / but bad grey region of your retina), and then are visible again if you stop looking at them.
There's no such thing as a single resolution or a single sensitivity to colours/greys in eyes. More likely those parameters depends on the region of the retina considered. Because of our grey perception, there's already been 10-bit black and white medical displays out there. Well... not exactly. Those displays are grey, simply because most of the picture produced in radiology are, indeed, grey scale. Thankfully we happen to have good sensitivity to grey contrasts so doctors in radiology can read them (with the help of monitors that have a wide enough dynamic range of light intensity and enough steps in between to mimic the quality of actual radiology films).
On the other hand, you could imagine obtain similar visibility to fine details by using pseudo colours. The problem is that no doctor is used to to analyse rainbow coloured pictures (...I tend to be the only one liking pseudo colour scales...) and if you move the window around (the mapping of data to intensity of grey) colours completely shift around (dark region may have been cyan with one window and orange with another), whereas with a grey scale darker region are always darker grey than lighter regions.
So the reasons are not only because of compatibility with our retina, but even more so because of practical considerations (looks like the original medium, simpler to manipulate, etc...)
Pseudo colours on the hand may be very popular in engineering printout because, well, once it's printed, it's hard to play with a display window, so you better find a way to cram as much possible information even on a medium that offers not such a big dynamic range of shades.
Note that then you have scale problems, which are happily abused for example by charlatans trying to sell snake oil to lower the radiation of your cell phone : the picture with snake oil looks much less redder than the one with snake oil. But that's because the pseudo colour mapping is different between the two pictures. Not because putting a sticker on the back of the phone suddenly stops it from frying your brain.
"Sufficiently advanced satire is indistinguishable from reality." - [Tips: 1DrYakQDKCQ6y52z6QbnkxHXAocMZJE61o ]
Spellcheckers are for the week.