Is somebody able to explain why certain e-ink displays are so slow to refresh while others are much faster? For instance my Garmin Vivoactive HR e-ink display, that is even capable of displaying 64 colors, is like an LCD display in terms of refresh rate apparently, you can't see the difference easily, while the one that was used to build this project takes a lot of time to even show a single frame (see the Youtube video where the display is presented, following the link Jann provided in the blog post). My best guess is that they use completely different technologies.
EDIT: Vivoactive HR uses a Transreflective LCD actually. This web site explains very well how it works:
Does that really have an e-ink display? Their specs page says "sunlight-visible, transflective memory-in-pixel (MIP)" which seems to be a type of LCD that has characteristics like e-ink.
Yes, you are right... different tech indeed! Works as a very good approximation of an e-ink frankly (I've a few kindle readers for comparison), but the contrast in artificial light is not as good as e-ink. Thanks! (and you are welcome).
Depending on what updates you need, you don't need to do a full refresh of the e-paper display (there will be some ghosting, though). Check out this video from papirus: https://twitter.com/gregeric/status/849218168224702465 That updates pretty nicely. And thank you for making redis!
> Is somebody able to explain why certain e-ink displays are so slow to refresh while others are much faster?
Sometimes it's the interface rather than the eink. The one in the article is a serial SPI interface. If the interface is parallel, instead it can be 8x, 16x, etc. faster
The tech can do better, but as I've heard it told, the market for it has been eink readers. Who needs to refresh their book at 120hz? is the thinking. Folks that want to game or have a "retina" resolution buy a tablet proper.
Upping refresh has steep power costs too. Ereader marketers like their "recharge once every 6-8 weeks!" message.
> The tech can do better, but as I've heard it told, the market for it has been eink readers. Who needs to refresh their book at 120hz? is the thinking. Folks that want to game or have a "retina" resolution buy a tablet proper.
Personally, I'd love to read most web-pages on an e-ink display, due to the high contrast, minimal power usage, and great readability even in direct light. I don't care hugely about usability for dynamic content, but I'd want fast scrolling with no update artifacts.
Related question: what's the state of color e-ink displays? Seems like it should be possible to produce a bi-stable display with a little bit of color depth, even if just through filters and dithering.
That's really impressive. I look forward to seeing that in production.
(Also, some searching suggests that it's "tens of thousands of colors" rather than true color, but that's still a major improvement, and more than enough for the majority of the web sans photographs.)
> "tens of thousands of colors" rather than true color
My meaning by true color was color achieved through actual pigment pixels versus older E-Ink products like triton that achieved color through LCD-like RGB color filters on top of reflective/non-reflective pixels. The difference between the two is quite stark.
They stopped bothering. They had some demo screens a few years back, mira-something I think, and a few products that used it, but nobody bought it because the colors were terrible due to the entire "e-inks generally don't require backlights and so massive contrast issues" I think the tech got bought up and pretty much shoved in a corner somewhere
Mirasol display is I think what you're trying to recall. I still have one; a Kyobo Android 'tablet'. In typical tablet usage conditions (i.e. indoors, probably at night on a bed) then yes, mirasol displays and color e-ink screens never matched LCDs and SAMOLEDS. Mirasol was decent/serviceable imo while color e-ink has been truly terrible.
However in outdoor conditions (e.g. at the beach or a park) I still haven't seen any iPad or Samsung display as pleasing as my old mirasol display without backlight. If someone released a new mirasol display ereader today, I would buy it in a heartbeat
Incredible work jayniz -- I guess everyone and their mother is suggesting improvements, are you preparing a new version with the rpi zero W and non-cut lego blocks?
Disclaimer: resin.io founder, we're so happy you chose resin for this awesome project ;)
Ha, exactly, but it's all good :) One guy actually made an LDD model that would require no cutting. I'll post it as a follow up on my blog.
Also to everybody reading this: I'd respond to the other comments but hacker news doesn't let me ("you're submitting too fast" even though I didn't write anything since half an hour or so)
I'd also love something like that. It seems there is a lack of demand though, as larger e-ink displays get expensive [0] and the only kits designed for the Raspberry Pi are small (<3"), so you'd need to figure out the protocol yourself.
That combined with a small keyboard and a phone for a hotspot would be awesome, I'd love to be able to sit in a park and code in the summer...
Hey, that's my thread! In case anyone's wondering, I ended up with a $25 Nook Simple Touch. If you just want the screen and no touch capability, then look through the rest of the thread. Unfortunately, I have not had the time to play with it much yet, but I can tell you the rooting process works very smoothly for the NST.
It looks like there's a business opportunity here for someone to make a really slick browser based LEGO editor that does cost estimates and orders all the correct components for you when you're finished. I'm curious how large the market for such a thing would be.
I remember using it in middle school and the prices were quite high. It was a great idea but no parents wanted to pay an extra premium on top of an already expensive product.
BrickStock is an OSX application that takes your Lego Digital Designer files and breaks them out into a manifest, which it then uses to reference the BrickLink marketplace for pricing.
I also had a Mac Plus with an external 20MB hard drive (this was late 1988). It wasn't the Apple HD20 [0] (which I had never heard of until now) but a 3rd party model, it used the Plus's SCSI port and had the same footprint so it sat underneath. I see Apple also made an HD20SC that used SCSI.
I didn't understand walrus01's comment to mean an external, rather than internal, hard drive.
Yes, though I opted not to get one because of the form factor and all the sacrifices they made in service of their goal (which was great but not for me).
In the colourful prototype picture there's a bottle cap for scale. It's a breach of Internet etiquette, which would require a banana, but it works for the purpose.
Don't worry, I discussed this with the folks at LEGO already and we're good: http://imgur.com/dOkyl4S
But yeah, after this post some nice folks showed me how a <2mm wall can be achieved without slaughtering bricks (special wall pieces, mounted sideways to allow for the left side of the display's board to disappear in in the Mac's frame).
Would anyone else chose a different software solution rather than Docker with resin.io? I love working on projects like this but I've stayed away from Docker so far. Docker plus a third-party service to manage it seems like it could be overkill, but it obviously got the job done.
Speaking as a resin fan, you need to know 0 about docker to use it. It does virtually everything for you, the fact that it's docker underneath is hidden entirely except in that you can write a dockerfile.
It's like how you don't need to provision hardware with Heroku. You just git push. And it works.
EDIT: Vivoactive HR uses a Transreflective LCD actually. This web site explains very well how it works:
http://t17.net/transflectiveTFT/