https://hackaday.io/project/11537-nekocal-an-e-ink-calendar/... (previously discussed at https://news.ycombinator.com/item?id=16140284 )
More research from others:
Certainly a number of drive waveforms and the basic concepts are presented. There are likely implementation subtleties or specific configurations that are not explicitly disclosed.
You can probably figure out a lot about both the implementation, and motivation by digging through the patents.
Additionally, as others have posted, the waveforms vary for different panels, due to manufacturing variation, so reversing a single example won’t tell you much about how to drive a different panel.
This is more or less the type of situation patents were invented for: a simple invention anyone can copy once it's invented, but difficult (I assume) to invent in the first place.
It's why I don't like most software patents, but why I think codecs, especially the modern ones, should be patentable. They're complicated engineering challenges requiring you to make numerous tradeoffs and I feel that just because the result is an algorithm doesn't mean that it shouldn't be patentable.
Other software patents I'm more dubious of, but I feel comfortable saying H.264 should be patentable.
You can buy fairly cheap displays from https://www.waveshare.com/ along with Raspberry Pi HATs and the PaperTTY project gives the power to do a partial refresh instead of full screen refresh on these screens. The waveforms are available in the Waveshare open source code and you can modify them as you will.
I am getting my screen next week and hoping to make my own little emacs only laptop using a Rpi Zero.
It has a few different ways of display. For example:
Floyd–Steinberg dithering: essentially only black and white with dithering, and the speed pretty good.
A16: uses 16 shades of grey, but the speed is horrible for moving images.
So I wonder if a mix of the two can do much better (e.g. moving objects using floyd-steinberg until it stops moving)
Before doing that, I need to know the exact limit of eink.
Are some operations slow because the microcontroller or because it is the limit of the display module itself?
This article shows such a question is not that easy to answer using the openly available information.
Also, I'm interested if there are algorithmic problems to be solved in this space. (I do research in theoretical computer science).
If you are in NYC and knowledgeable I'm very interested to chat (I know little, but I have the monitor you can play with).
It's because the technology is a physical one --- tiny bits of reflective/absorbant particles have to physically move up and down to change the state of the pixel. It's almost like a miniaturised flip-dot display:
The response speed increases with temperature and voltage, but temperature is not really controllable and high-voltage high-frequency drivers are not easy to make in the sizes required. Here's an article about someone making his own segmented eink display:
In the comments you can see a picture illustrating the difference the driving voltage makes.
(Also previously discussed at https://news.ycombinator.com/item?id=14124086 )
Related video of someone playing around with the eink material itself, showing how it responds to voltage changes and that it's possible to "draw" on one: https://www.youtube.com/watch?v=jXxSOR1PHpE
For your case, it seems more reasonable to divide the screen into some kind of rectangular cells, keep checksums for the cells and switch them to A16 if the checksum stays constant for two visits? This would avoid the need for a full extra frame buffer while still being relatively accurate. I think one main problem will be that you'd need to modify the screen firmware to do something like that. Probably would make sense to code a software emulation on a regular screen to see what it would look like...
It is reasonable enough to see what's happening, but obviously, I'm not going to watch a movie on that.
Been waiting for an e-ink laptop specifically for working with it outside in the sunlight, and in the evening, since blue light is supposed to be really bad for humans at night.
"Dasung Paperlike PRO vs Onyx Boox Max2 Pro Extended Monitor" https://www.youtube.com/watch?v=6pw-oCItgx8
This company https://www.visionect.com/blog/32-inch-eink-development-kit-... makes a 32" product for signage that doesn't seem to have any input for normal computer use...but its 3500eur. I don't know if they paid huge licensing fees etc or what but the pricing is mental for the tech/materials involved and seems to be purely business/supply stuff. The largest I have seen otherwise is the Dasung stuff that's 13" and 1000eur or the Onyx Boox MAX 13" that is a self contained E-reader and has terrible lag when used as a monitor. I have not found anything between the 13" and this singular 32" available and nothing but the small Dasung designed to be used as a monitor.
Another option is to place the back of a plastic screen towards the sun, using it as a backlight.
White LEDs are currently approximately 25% efficient (and have a limit of 44%), so my LCD display needs to use 26 watts of electricity to be as bright as direct sunlight landing on it.
My laptop's backlight can't actually do that, but if it could, it would exhaust 70 watt-hour battery in less than 3 hours, without accounting for any power used by the CPU, network card, RAM, or hard drive.
It is about contrast, not brightness.
e-ink has a contrast ratio of about 20:1 with 40% reflectance. A sunny day is 10000 - 30000 nits, giving at least 4000 nits of brightness.
The 2018 MPB screen has a 1500:1 contrast ratio and about 500 nits of brightness.
So if e-ink is really easier to read in the sun (which I think) then brightness is much more important.
2. Visual snow syndrome: Many had their first experience of visual snow after staring into the screen for too long. People who believe in that is the cause of visual snow would want a screen that does not emit light.
There’s also the spikey spectrum, which looks the same consciously but I’ve seen claims  that is messes with our circadian rhythm.
 I lack the skill to tell if this is real science or neo-mystic woo.
sudo su -c "echo 200 >/sys/class/backlight/intel_backlight/brightness"
Haven't found a Windows nor MacOS equivalent though.
I've used it a fair bit in conjunction with Redshift to get a screen that's reasonable for reading ebooks in low light.
I'd buy an e-ink laptop tomorrow if there was one.
It seems to be well known that it for low frequency PWM it causes headaches for some people, so I would expect that there is at least some research in that direction. That said, it seems to only be a problem for few people and only for very low frequencies.
How so? You can set any brightness you want to match the environment. Usage under direct sunlight is one edge case e-paper is clearly better at, and any usage with poor environmental lighting is detrimental to your eyes itself, regardless of technology.
I'm really sceptical about the entire "E Ink is better for your eyes" thing until I see some serious studies. People managed to ruin their eyesight centuries before computer screens. There's just too many factors at play here, from genetics to typography.
Firstly staring directly at a light source means that you are seeing most the rays directly. The only attenuation you will receive will be scattering from the medium - presumably air which will be low.
Seeing it from a second surface adds more effects - the surface may have absorb some of the light. Any real surface is also not going to have perfect reflectivity (everything from scattering, internal diffraction and destructive interference could occur). In short there will be loads of secondary/tertiary effects causing attenuation. Someone also mentioned polarisation, I also wonder about interaction between reflective surfaces which could cause additional attenuation via interference but the physics of waves has never been a strong point.
The other major factor is the physiology of the eye - we evolved from a world that didn't have abundant light sources and it's hard to know the effect of focusing on a source of light that differs from the ambient profile.
I know its not scientifically measuring nits/lumens etc but stare into even a low wattage (15w etc) light bulb for a few min and see how it leaves you feeling...seeing spots etc. Then stare at a newspaper or book or E-Reader under a 250w lightbulb or even brighter sunlight outside...and notice how you do NOT have that same problem. Even with a much brighter light source there is something different about how/how much actually enters your eye. I'd love an answer as this is a major life problem for me.
If I look at a matte LCD display (glossy doesn't work in this example due to reflection) in bright sunlight there is a narrow window of display brightness where the things are bright enough to be readable but the sunlight overwhelms the "lit" effect and it appears sort of like you would expect color E-ink too...more flat and printed like the display is now reflective rather than backlit..but it's a very narrow sweet spot that's hard to maintain and I cannot seem to replicate well with artificial lighting inside as it takes a lot more than a 60w bulb even directed right at the display and glare starts to become a factor with a more concentrated light source vs ambient, indirect light from outside. Brightness too low and it's unreadable due to contrast, too high and you get the glow again. To me the root problem is the backlighting somehow...possible combined with dithering or other rendering games that cause even static images to be moving as far as the eye is concerned. But it's not a common enough issue for industry to care about, and as with most things, until individuals suffer from something themselves, they often refuse to believe it's an issue for others.
Does it look good with white text on black background ?
The main advantage is that it's easier on the eyes
I’m unconvinced, but (obviously) to test this I would have to actually get a monitor sized black and white reflective display, and I can’t find any of those — largest I have found so far is 320x240 for about £200
That being said the segmented LCDs like those found in 4 function calculators and wrist watches usually look fantastic.
If I didn’t already have way too much on my personal to-do list, I’d be interested in converting an LCD with a backlight into a reflective one. I assume it’s not possible with all models, but also that it is possible with some.
> All of the lookup table information has been removed from this datasheet, because the industry thrives on secrecy or something
Highly recommended, as usual for him it's a great video/write-up:
One thing that I want to note is that while their documentation might be closely guarded secret, the actual waveforms are actually available out in the open.
For instance here are the raw waveforms for one of the latest Carta generation e-paper panel:
~ hexdump -C epdc_ES103CS1.fw | head -10
00000000 34 88 c4 7c 01 2d 01 00 dd 09 00 00 07 06 a4 00 |4..|.-..........|
00000010 03 58 21 1d 3c ca 01 85 02 00 00 00 40 00 00 fe |.X!.<.......@...|
00000020 47 00 00 01 00 04 0d 00 ff fc 00 00 00 00 00 54 |G..............T|
00000030 00 03 06 09 0c 0f 12 15 18 1b 1e 21 26 2b 30 28 |...........!&+0(|
00000040 00 00 00 00 00 00 00 08 3e 05 00 00 00 00 00 e8 |........>.......|
00000050 cb 06 00 00 00 00 00 c8 8d 0a 00 00 00 00 00 a8 |................|
00000060 4f 0e 00 00 00 00 00 98 00 00 00 00 00 00 00 a0 |O...............|
00000070 65 00 00 00 00 00 00 a8 bc 00 00 00 00 00 00 b0 |e...............|
00000080 0b 01 00 00 00 00 00 b8 5a 01 00 00 00 00 00 c0 |........Z.......|
00000090 aa 01 00 00 00 00 00 c8 31 02 00 00 00 00 00 d0 |........1.......|
At work I looked into integrating e-ink into a product, and the problem seems not patent fees, but availability. Granted, that could be due to patents itself, but the answers I got why it's hard to procure e-ink display were sadly "the end customers don't want it" and "e-ink is dead".
One perfectly reasonable theory of self-interest is to keep prices very high. Yes, they could drop the price and possibly make more. But they could end up making less, and they'll certainly have to do a lot more negotiation work.
As an analogy, look at Apple. They have less than 20% of the smartphone market, but make 80% of the profit.  They were way ahead of everybody and could have captured a much larger share of both the phone and tablet markets. They're using a different theory of self-interest than the other players.
The people making zillions of phones make a fraction of the margin Apple does on it's cool phones.
Granted, I do understand that from the perspective of a western entrepreneur trying to do business in China it seems to be a lawless space, especially if you've been burnt by fake products from Alibaba etc..
Small time shipments are unlikely to be looked at (eg. if someone buys a single electronic ink display off of AliExpress), but someone actually importing many displays for resale would likely eventually get caught.
You occasionally search the market for products that contain your tech (in this case eink), then look in your records to see if they bought/licensed from you.
I remember reading about "fake" Kindle replacement displays showing up on eBay a while ago, not as good as the real thing in terms of image quality etc. but certainly usable, so someone is clearly making them...
And https://hackaday.io/project/21168-fpga-eink-controller has an FPGA doing waveforms by extracting them from existing firmware. So while it might be a closely guarded secret, just replicating it is possible.
No crazy +/- 40V lines, no external driver circuitry, and extremely low power. Not as low power as E Ink, but since you don’t need to drive wacky voltages, your power budget math will come out the same or better.
Some models are different, though, and basically turn into a mirror when their pixels are turned on, rather than become matte black.
Adafruit sells a dev board for 25 bucks. Perfect for low power wearables, much better than E Ink.
How these are constructed and used are well known, and they're probably lying about "closely guarded" to make sure people don't realize it is just a clever usage of an off the shelf part, and none of the magic is in-house.
The post identifies some on market E-ink display components and discusses the design of a proprietary display driver.
Wow, I've never seen that. Is it a common thing? It also puts me off buying one of these (assuming the $900 price tag hadn't put me off already).
Erased markings are usually a sign of off-the-shelf parts; they didn't go to the effort of printing their own branding on them.
If you want to see a more complete list of waveform names you could probably go look at the gpl releases of popular readers...
At that point I would really care for software with a minimum amount of non fancy feature that doubles or triple battery life. If you make extensive use of text, and don't use high resolution image or set low limits, there are surely ways to make a more durable device.
Cheap androids are awesome, but I have no idea why they're becoming slow as they age, and I'm sure software is always to blame. I really crave minimalist designs when it comes to hardware and software.
An e-paper display would save a bit more, sure, but the type of display isn't the only limitation to a low-power smart phone. You need to design the whole thing to work together to save power; changing one component isn't gonna be enough.
Japan display has made 600 dpi controllers.
Further been thinking of building an Eink based smartphone, the eink display would have low power. Some kind of low power small system board running Linux and a 4g modem. Not sure what system board to pick. Batteries should be LifePo4 Lithium iron for battery safety. What kind of high resolution touch based eink displays are out there which are easy to drive from a small Linux system board?
Juul spent a good amount of time working out how to read the stored waveform configurations (different for each unit) and getting them to run at a reasonable speed.
Seems limited only by the voltages that just end up frying parts.
One can calibrate the brightness, white balance (e.g. Apple's True Tone), resolution, etc of a regular LCD or OLED display so that the same amount of photons hit your eye (and are more tightly controlled when you move the display around) as with an E Ink display, and get better resolution color, and saturation to boot.
Besides the idea that it was easier on the eyes was never backed by any substantial scientific studies, it was just "but it's like paper, so it should be" plus some manufacturer sponsored crap.
Well, there's the extended battery life I guess...
Other than that, it's hard to compare because nobody makes large monochrome lcd displays. Monochrome is actually a feature for the intended use: newspapers and books use a pretty limited amount of color, so you get improved pixel density and geometry without having to split pixels into colors.
OLED won't work for dark text on a light background, and tends to be rather expensive -- I can't imagine a Kindle sized device with an oled display for less than $200, so that limits it to the high end models.
Battery life is also much, much better. My tablet lasts two to three days at most, under light load, with power saving turned up to eleven. The Kindle can easily last one to two weeks. (It all depends on usage patterns, of course, but for my case those are the numbers I come up with.)
I do agree, though, that it is hard for me to think of a use case other than an ebook reader. For that particular case it is perfect, though.