Comment by adolph

1 day ago

Wow, I'm glad to see that person is getting some more recognition for this work.

A claim in the video that I can't verify but makes economic/logistic sense is that the speed problem isn't the panels but the controllers. The current crop of controllers are optimized for low power, which fits the e-reader use case but that is not optimal for the interactive use case.

> A claim in the video that I can't verify but makes economic/logistic sense is that the speed problem isn't the panels but the controllers.

I don't understand the claim. It is lacking in specifics. Are they claiming that electrophoretic materials (meaning the panels) can actually switch (meaning move pigments) faster than say x.y micrometers per second? I don't think that is true. The article shows that what Wenting did ("binary transition") is pretty much the same as what companies like Dasung did. Instead of trying to have grayscale, it is faster to hit somewhat-black and somewhat-white and give the illusion of fast movement than actual fast movement.

  • > Are they claiming that electrophoretic materials (meaning the panels) can actually switch (meaning move pigments) faster than say x.y micrometers per second?

    No, I think the claim is that the controllers are slower that what the panels can theoretically support.

> The current crop of controllers are optimized for low power, which fits the e-reader use case but that is not optimal for the interactive use case.

Why try to contort the technology for something it's not good at, instead of using a more appropriate technology like transflective LCDs? Eink isn't the only option for reflective displays. If you increase the power use of eink to get better refresh rates, I imagine you'd end up using more power than (and still end up with lower refresh rates than) an MIP display.

I don't understand the growth of the market as a whole for eink monitors, when tLCDs exist and are disappearing from the market.

  • I'm pretty sure e-ink has a much higher ceiling for reflectance than TLCDs/RLCDs, so you'll be able to use it comfortably without a frontlight in a lot more situations which could more than make up for increased power usage. I think they are also naturally better in terms of glare compared to any type of LCD.

    • Viewing angles are also fantastic compared to almost all T/R LCDs - they tend to be fairly directional. It's a great display tech for many things that don't need 60+fps.

      1 reply →

  • I'm currently reading your post on a transflective LCD monitor. The problem with them is the very low contrast ratio which requires very high ambient illumination to make them readable or other workarounds like what they did for the Daylight DC-1.

  • It isn't clear to me that eink's underlying display technology isn't good at the interactive computing use case so much as the implementations aren't optimized for it. There could be a position where more power than an eink reader is used but still far less than traditional active displays since unchanged pixels aren't driven.

    • That's how I think about it too.

      E-readers are vertically integrated devices: the hardware, software, UI, and refresh behavior are all tailored around reading. E-ink tablets like reMarkable are similar, but optimized around writing and annotation.

      A traditional monitor is much more general-purpose, so it doesn't get the same kind of end-to-end optimization for the display medium. I think there's room for an in-between category: a more interactive e-ink device where both the hardware and software are designed around the strengths and limits of the panel.

      There's some related work happening in this direction:

      https://nlnet.nl/project/epd-wayland/

    • It should be good enough for interactive use, but not for watching movies.

      In TFA it is said that for these new faster panels the transition time of a pixel is around 50 ms. This is comparable with some old LCDs.