Could you have, literally, a gallery of eInk art, where all of the panels have no power. Aren't even plugged in. Someone grabbed a bunch of displays, plugged them in to a machine, rendered the picture, yanked them out and framed them and stuck them on the wall.
Will they last? As I understand it eInk is a "no power" display. Could this be done?
Sure. I have a couple eink devices, if they're not powered they just keep whatever image they had.
They're not exactly the best possible displays for all art though. The "whites" aren't all that white is the biggest problem. It's perfect for a book, and I even really like it for comics, but it's probably a bit limiting for art.
That's not even mentioning that the good ones are only grayscale. Color ones exist, but the colors look _badly_ washed out.
They're also pretty small and/or expensive.
Basically you'd be better off in every way with just a printer, any kind of printer.
I have not seen any actual proof that the colour ones do exist. Other than the crappy 3 colour displays.
I have only seen obscure demos or OEM store listings without much detail. From what I can tell, they are just regular LCD displays with matte surfaces to look like epaper.
The term is 'bistable': once the state is changed, no energy expense is needed to keep it. To change the state, a relatively considerable amount of energy is needed.
> Will they last
You have turned the stone, the white side facing up: it will last.
> Could you have, literally, a gallery
Yes, as the EPD dots are changed to the desired states, the displays are like physical content for picture frames, "you could hang them on the walls".
(«last», in the original post, was very probably only about the consumption side and the quoted reply only referred to that attempting an explanation.)
I am not informed about E-ink pigment degradation. It is official though that the dot state switching capability is not infinite: after a number of switches - around 10 million to the specifications read a few years ago - the dot breaks.
A cheap black and white laser produces pictures at least as good as most e-ink devices and in higher resolution. A toner cartridge for my Canon costs about 100 USD and will print 2000 sheets. 2000 sheets of A4 cost about 30 USD.
An e-ink display that size costs 100s of USD so I can afford to reprint the picture thousands of times before I hit the cost of the e-ink display.
And anyway, art galleries do not change the content frequently so they could afford to use a higher quality printer and higher quality paper.
> art galleries do not change the content frequently
No. :)
The art galleries the original poster mentioned trying to clarify the question do not.
Art galleries do display dynamic art, increasingly. And mentioning E-ink in the context of art installations makes little sense if the the technology were used for static content only.
That depends on the application, scale, and frequency of changes, etc.
E-ink displays can update without human intervention. If you're working with a large number of individual items or displays, then you'd have to factor in the human labour cost.
Keep in mind that smaller displays (6--8") can be obtained for ~$20--50 either used or in bulk, and if you're looking a labour costs, these would actually pay for themselves in a few days. Sooner if displays are over a widespread area, are updated often, and/or are hard-to-reach.
A few A4/letter sized notices at doors which an admin staff could update daily or weekly? Stick with paper. A few hundred or thousand over a large campus or outdoor area, updated multiple times daily? E-ink is probably a better fit, and certainly competes with other digital / electronic signage.
...And you will probably be intrigued to know about the "Advanced Color e-Paper (ACeP)" technology, which uses actual multiple pigments inside each dot.
It currently has low pixel density, it is very costly, the update may be very slow (in some models it takes over 8 seconds to change the states of its dots), and it should be using trichromy instead of quadrichromy (the declared total colours amount to 4096, which suggests 16^3, where 16 is the number of states in the traditional B/W E-Ink displays). The actual real visual effect "in presence" is to be seen - the ability to have big sized changeable printouts is there.
For smaller displays, such as the Onyx BOOX Poke Color (7.8"), the price premium is relatively slight over the B&W equivalent: $400 vs. $350 per the manfuacturer.
Keep in mind that for virtually all e-ink devices, the display properties themselves are dictated by the display vendor, Eink.com, who have a monopoly under patent. It's as if, say, all OELD / LCD displays were manufactured by a single company. (There's a high concentration within that field, but not a complete monopoly.)
Your post is more correct than mine was: the product description says «SB1452-QAA ... with color filter design».
But it is more probable that that display is kaleido, not triton (similar idea but upgraded results). Kaleido is the technology used in the colour E-ink displays shipped in the past few years.
Also the information I found to refresh my memory was incorrect: yes, 4096 is for the filter. Kaleido uses RGB trichromy.
Some information state 32000 colours for ACeP - probably rounding for 2^15. Which I cannot translate into a good guess, not having found more precise information: 5-bit state per pigment trichromy makes little sense, but also 4-4-4-3 (CMYK) is not convincing, or 3-3-3-3-3-3 (?), or 3-2-2-2-2-2 (KCMYOG). One of the relevant matters is that the difficulty for E-ink was to find a method to change the state of the single pigment in the same cell - it is not clear how many pigments per cell they can now control.
If I'm not mistaken, this is how Best Buy does their price tags now. They have eInk (or, eInk looking) displays on the end of racks that can be changed out with power but otherwise seem to work quite well.
If you want, you can think of e-ink displays like this:
- Persistence is free. Once an image is set, it remains indefinitely. Even with no power applied. I've noticed no degradation over days or weeks, and expect images would be retained for years.
- Pixels are cheap. Dot densities are on the order of 220--300 dpi. This compares with many consumer-grade laserprinters. High-definition printers approach 600 dpi, photgraphic printers perhaps 1200 dpi. At a even less than a meter distance this is undetectable. I can just make out pixels under a high-power magnifying glass.
- Repaints are expensive. It is changing the display which costs power, not sustaining it.
- Quality/responsiveness is a trade-off. At maximum frame rates, displays achieve about 4--8 Hz (fps). Quality degrades considerably with ghosting and contrast issues. High image quality comes at low refresh rates (0.5 -- 2 Hz), but is at or near paper quality. Higher-quality repaints induce greater flicker, lower-quality, less.
- Colors and shades are few. Most e-ink displays have limited greyscale palettes, typically 4--16 shades. As a consequence, even B&W imagery appears posterised. For fine art other than line drawings, you'll want a dithering or halftoning pre-processor for best results.
- Overall contrast is also lower than a black ink on white paper. It's closer to a very dark grey on a somewhat off-white, though under bright light (sunlight or artificial) it is highly readable and pleasant. More so under sunlight than an unshaded paper which tends to be too bright.
- Frontlights enable low-light viewing, though they tend to reduce the darkness of blacks whilst lightening the whites. I prefer both ambient room light and the frontlight for indoor / night-time reading, though it's possible to read using just the frontlight, of course.
Displays are exceedingly well suited to text and line art. (The Onyx BOOX line comes with pen-and-ink line art for power-off and suspend screens, by default. These look delicioius.) Stark black-and-white content (say, Picasso's "Don Quixote") also render marvelously. For photographs, I prefer a halftoned render (this is achieved automatically at certain display quality settings on my device). Mixed-colour, low-contrast gradient images, especially nonphotographic images, seem to be the worst option. Anything which leans heavily on colour for context comes out quite muddy given both the B&W and limited greyscale pallettes. Website animations are immensely distracting as they result in a whole-display flicker at high disiplay quality, and a strobing effect of the image itself at modestly lower-quality rendering.
Navigation by pagination (where the entire page changes at once) is vastly preferable to scrolling. At high display quality, the latter just disolves the screen into static, making tracking of how much you've scrolled and where the new/old content division are difficult. Tools such as the EInkBro browser default to paginated navigation and make Web browsing very much better than a stock browser (say, Firefox/Chrome, or anything based on scroll). I find myself trying to invoke EInkBro's pagination on Firefox all the time.
(The BOOX is an Android-based device. It's a tablet that happens to be an excellent book-reader. I've been using one since March.)
The Office Automation Word Processor I wrote (for use with the Onyx Max2) uses the concept of a two-columns landscape rendering in fixed pages each starting from an arbitrary point in the document - no scrolling anywhere, just render from top-left to bottom-right. Use of real estate is fully optimized. These special tools call for efficiency as a goal of their software.
Typical different paradigms make only sense in other contexts (e.g. WYSIWYG simulation of paper pages to check physical printout).
I actually want to try to implement this with ESP32 and deep sleep, where it changes the image say once a day, and it only needs to be charged on the order of once a year.
Even better would be if the top part of the frame is lined with solar panels and use that to tickle charge the battery.
Joan’s parent company, Visionect[1], also produces the Place & Play, a 13” or 36” eInk display[2]. Some guy created a rotating newspaper frame from it[3]. Looks quite amazing, and it has a 1 year battery life.
31 comments
[ 0.96 ms ] story [ 64.7 ms ] threadCould you have, literally, a gallery of eInk art, where all of the panels have no power. Aren't even plugged in. Someone grabbed a bunch of displays, plugged them in to a machine, rendered the picture, yanked them out and framed them and stuck them on the wall.
Will they last? As I understand it eInk is a "no power" display. Could this be done?
They're not exactly the best possible displays for all art though. The "whites" aren't all that white is the biggest problem. It's perfect for a book, and I even really like it for comics, but it's probably a bit limiting for art.
That's not even mentioning that the good ones are only grayscale. Color ones exist, but the colors look _badly_ washed out.
They're also pretty small and/or expensive.
Basically you'd be better off in every way with just a printer, any kind of printer.
I have only seen obscure demos or OEM store listings without much detail. From what I can tell, they are just regular LCD displays with matte surfaces to look like epaper.
https://goodereader.com/blog/product/pocketbook-inkpad-color...
I've got one of these and it's quite nice. Of course the colors are not as intense as other screen types. But still actually pretty good.
However, with the Nook I'm using, the Android subsysyem eventually realises it is out of power and will displau a static "charge me" screen.
If you built your own device, you could detatch the screen from the power and the image would stay on.
The term is 'bistable': once the state is changed, no energy expense is needed to keep it. To change the state, a relatively considerable amount of energy is needed.
> Will they last
You have turned the stone, the white side facing up: it will last.
> Could you have, literally, a gallery
Yes, as the EPD dots are changed to the desired states, the displays are like physical content for picture frames, "you could hang them on the walls".
That's only true of the white side, though. If you've got the blue side of a stone facing up, it may fade.
I am not informed about E-ink pigment degradation. It is official though that the dot state switching capability is not infinite: after a number of switches - around 10 million to the specifications read a few years ago - the dot breaks.
Only if you do not plan to change the visualized contents frequently.
An e-ink display that size costs 100s of USD so I can afford to reprint the picture thousands of times before I hit the cost of the e-ink display.
And anyway, art galleries do not change the content frequently so they could afford to use a higher quality printer and higher quality paper.
No. :)
The art galleries the original poster mentioned trying to clarify the question do not.
Art galleries do display dynamic art, increasingly. And mentioning E-ink in the context of art installations makes little sense if the the technology were used for static content only.
Sgustok is a marvellous repository, though infrequently updated nowadays: https://sgustok.org/video/category/installyacii
E-ink displays can update without human intervention. If you're working with a large number of individual items or displays, then you'd have to factor in the human labour cost.
Keep in mind that smaller displays (6--8") can be obtained for ~$20--50 either used or in bulk, and if you're looking a labour costs, these would actually pay for themselves in a few days. Sooner if displays are over a widespread area, are updated often, and/or are hard-to-reach.
A few A4/letter sized notices at doors which an admin staff could update daily or weekly? Stick with paper. A few hundred or thousand over a large campus or outdoor area, updated multiple times daily? E-ink is probably a better fit, and certainly competes with other digital / electronic signage.
It currently has low pixel density, it is very costly, the update may be very slow (in some models it takes over 8 seconds to change the states of its dots), and it should be using trichromy instead of quadrichromy (the declared total colours amount to 4096, which suggests 16^3, where 16 is the number of states in the traditional B/W E-Ink displays). The actual real visual effect "in presence" is to be seen - the ability to have big sized changeable printouts is there.
ACeP display, 31.2'', 1280x720 (50ppi) at 2300$: https://shopkits.eink.com/product/31-2%CB%9D-color-epaper-di...
https://onyxboox.com/boox_nova3color
https://onyxboox.com/boox_novaair
Yes, the colour experiene is muted. This review seems to capture the capabilities and limits well:
https://youtube.com/watch?v=aSUcnkQCa-s
Keep in mind that for virtually all e-ink devices, the display properties themselves are dictated by the display vendor, Eink.com, who have a monopoly under patent. It's as if, say, all OELD / LCD displays were manufactured by a single company. (There's a high concentration within that field, but not a complete monopoly.)
https://www.eink.com/
Display characteristics, properties, sizes, and price are pretty much determined by Eink.
But it is more probable that that display is kaleido, not triton (similar idea but upgraded results). Kaleido is the technology used in the colour E-ink displays shipped in the past few years.
Also the information I found to refresh my memory was incorrect: yes, 4096 is for the filter. Kaleido uses RGB trichromy.
Some information state 32000 colours for ACeP - probably rounding for 2^15. Which I cannot translate into a good guess, not having found more precise information: 5-bit state per pigment trichromy makes little sense, but also 4-4-4-3 (CMYK) is not convincing, or 3-3-3-3-3-3 (?), or 3-2-2-2-2-2 (KCMYOG). One of the relevant matters is that the difficulty for E-ink was to find a method to change the state of the single pigment in the same cell - it is not clear how many pigments per cell they can now control.
If you want, you can think of e-ink displays like this:
- Persistence is free. Once an image is set, it remains indefinitely. Even with no power applied. I've noticed no degradation over days or weeks, and expect images would be retained for years.
- Pixels are cheap. Dot densities are on the order of 220--300 dpi. This compares with many consumer-grade laserprinters. High-definition printers approach 600 dpi, photgraphic printers perhaps 1200 dpi. At a even less than a meter distance this is undetectable. I can just make out pixels under a high-power magnifying glass.
- Repaints are expensive. It is changing the display which costs power, not sustaining it.
- Quality/responsiveness is a trade-off. At maximum frame rates, displays achieve about 4--8 Hz (fps). Quality degrades considerably with ghosting and contrast issues. High image quality comes at low refresh rates (0.5 -- 2 Hz), but is at or near paper quality. Higher-quality repaints induce greater flicker, lower-quality, less.
- Colors and shades are few. Most e-ink displays have limited greyscale palettes, typically 4--16 shades. As a consequence, even B&W imagery appears posterised. For fine art other than line drawings, you'll want a dithering or halftoning pre-processor for best results.
- Overall contrast is also lower than a black ink on white paper. It's closer to a very dark grey on a somewhat off-white, though under bright light (sunlight or artificial) it is highly readable and pleasant. More so under sunlight than an unshaded paper which tends to be too bright.
- Frontlights enable low-light viewing, though they tend to reduce the darkness of blacks whilst lightening the whites. I prefer both ambient room light and the frontlight for indoor / night-time reading, though it's possible to read using just the frontlight, of course.
Displays are exceedingly well suited to text and line art. (The Onyx BOOX line comes with pen-and-ink line art for power-off and suspend screens, by default. These look delicioius.) Stark black-and-white content (say, Picasso's "Don Quixote") also render marvelously. For photographs, I prefer a halftoned render (this is achieved automatically at certain display quality settings on my device). Mixed-colour, low-contrast gradient images, especially nonphotographic images, seem to be the worst option. Anything which leans heavily on colour for context comes out quite muddy given both the B&W and limited greyscale pallettes. Website animations are immensely distracting as they result in a whole-display flicker at high disiplay quality, and a strobing effect of the image itself at modestly lower-quality rendering.
Navigation by pagination (where the entire page changes at once) is vastly preferable to scrolling. At high display quality, the latter just disolves the screen into static, making tracking of how much you've scrolled and where the new/old content division are difficult. Tools such as the EInkBro browser default to paginated navigation and make Web browsing very much better than a stock browser (say, Firefox/Chrome, or anything based on scroll). I find myself trying to invoke EInkBro's pagination on Firefox all the time.
(The BOOX is an Android-based device. It's a tablet that happens to be an excellent book-reader. I've been using one since March.)
The Office Automation Word Processor I wrote (for use with the Onyx Max2) uses the concept of a two-columns landscape rendering in fixed pages each starting from an arbitrary point in the document - no scrolling anywhere, just render from top-left to bottom-right. Use of real estate is fully optimized. These special tools call for efficiency as a goal of their software.
Typical different paradigms make only sense in other contexts (e.g. WYSIWYG simulation of paper pages to check physical printout).
Even better would be if the top part of the frame is lined with solar panels and use that to tickle charge the battery.
https://getjoan.com/
https://www.ionnyk.com/
https://www.eink.com/architecture.html
[1] https://www.visionect.com
[2] https://www.visionect.com/product/place-and-play-32/
[3] blog: https://gregraiz.com/i-made-an-eink-newspaper/ and video: https://youtu.be/gECj1AE9D2c and code: https://github.com/graiz/newsprint