A 60fps e-ink monitor prototype in action | Hacker News Reader

A 60fps e-ink monitor prototype in action

twitter.com

55 points_Microft19 comments

Nice!

I often dream about using a 24 inch color e-ink display as my daily driver and just have an extra side monitor for everything that does video. Most of what we do is to manipulate text yet im literally at 120hz currently to render this post.... so the the cursor can blink ;)

You would regret it every time you scrolled the text on that display, because it would turn to gray mush. Your eyes would not be able to follow it as it moves and you would lose your place every time. Clarity matters, especially for text.

Pg up/down?

Scrollable media is still "new". There's no reason we couldn't or wouldn't redesign for it.

Yes, the traditional way to implement paging was to simply redraw the screen once at the new position. As long as it always moves a consistent amount, you’ll get used to the discontinuity and you won’t lose your place. Emacs still works that way by default, but even it gets more smooth scrolling options with every release.

Even so, on an EInk display you will still have the old text fading away for eight frames while the new text fades in for eight frames. It will not be instantaneous at all, and the text will be low–contrast and overlapped for that whole time. And this happens as you edit the text too. Adding or removing lines will shift later text up and down, turning it into a blurry mess. With word wrapping it will be even worse, as every character you add shifts the later text to the left as well as further down.

Still, they work ok for static text that you move by a whole page at a time.

Nice, but pause the video while it’s playing Bad Apple and you can measure the pixel response times. It looks like 8 frames or 133ms, which is pretty average for Eink and terrible compared to the average LCD. Compare that with a review of a good monitor: https://www.youtube.com/watch?v=xDB9sONfJ5Q

It's an ordinary old e-ink screen, just driven in a different way.

So what? Someone made progress in an area I find interesting. That's much more useful to me than yet another comment that is crapping on someone else's work.

But it isn’t progress. The refresh rate of the screen is still 130ms but now it’s being updated 8 times faster than that, before the pixels have changed color. Thus they continue changing color for the next eight frames before finally reaching either black or white. If your LCD panel did that you would toss it in the trash; most people cannot tolerate more than two or three frames of latency before complaining about motion blur and ghosting.

So they are basically feeding high frequency data into the e ink which low pass filters it?

That's still progress, some of the high frequency information gets through, it's just blurred.

No, that isn’t what is going on. There’s no filter. It just updates the display at 60 Hz. If you simply pause the video you will see the result: the last eight frames are all visible at the same time, getting fainter as they get older. It makes every moving object look like it has a trail of ghosts following behind it.

All LCD monitors have this problem to some extent, but cheap LCD monitors are the worst. This display is about 4× worse still.

Update: Here, a screenshot: http://db48x.net/temp/Screenshot%20from%202022-06-09%2000-59...

This is taken right during the arm movement as she throws the apple up at the beginning of the video. Since she is drawn in black on a white background, the image of her arm from the current frame (at the top) is super faint. The pixels in the apple will never reach full black because they were supposed to be black for a frame and then go back to white next frame. They started darkening, but they are already being told to lighten again. The apple essentially disappears during this motion, and it’s supposed to be the focal element of the scene!

The inertia of the e ink itself is the low pass filter, i think.

That’s just not a useful analogy.

It wasn't meant as an analogy. Am i wrong in that the mass of the ink particles is the root cause for the low refresh rate of i ink displays?

I have no idea why these displays are so slow, but the concept of a “low–pass filter” just doesn’t apply here.

Okay, now an analogy: Imagine a container vessel and a rowboat. Which of them is more stable, when you alternate their steering angle with a high frequency?

At this point I would say your incorrect. The e-ink is acting as a temporal low pass filter, for the desired signal that is the brightness of that pixel. Perhaps you’re thinking of a spatial low pass filter?

I thought it was useful

It's not an analogy. It's literally what is happening. This is apparent by your own admission that each pixel displayed is an average of the last eight frames of that pixel. That is the definition of a convolution filter. The display acts as a physical convolution filter, which is implementing a low-pass filter.

Looks as bad as STN. Now slap an active matric on it and we are back at square one.

Reply on news.ycombinator.com