When I started this project, I thought sourcing e-ink displays would be simple. Buy screens, wire them up, done.
I was wrong. Here's everything I've learned.
E Ink Corporation has a monopoly
First surprise: almost all e-ink displays come from one company — E Ink Corporation, based in Taiwan. They hold the key patents and manufacture the panels that end up in Kindles, Kobos, reMarkables, and basically every other e-reader.
This matters because you can't just "shop around" for e-ink panels like you can for LCDs. The supply chain is narrow, and options are limited.
Grayscale levels matter more than I thought
E-ink displays come in different grayscale capabilities:
- 2 levels (black and white only) — very fast refresh
- 4 levels — good for text, some dithering for images
- 16 levels — proper images, manga-readable, slower refresh
The screens I'm using (same as Diptyx) are 4-level grayscale. This is fine for text — novels, articles, anything word-based. But it's not great for images. Manga would look rough. Photos would be a mess.
This was a deliberate tradeoff. 16-level displays are more expensive, need more complex driving, and have slower refresh. For a text-focused reader, 4 levels is the sweet spot.
Refresh modes are complicated
E-ink doesn't work like LCD. The "ink" particles physically move when you change the display. This means:
- Full refresh — flashes black/white, clears ghosting, slow (~1 second)
- Partial refresh — updates only changed areas, faster, but ghosts build up
- Fast modes — even quicker, but more artifacts
Getting this right is a software challenge. Refresh too often and reading feels sluggish. Refresh too rarely and the screen gets muddy with ghost images.
Where to actually buy them
This was harder to figure out than expected.
Good Display is where I bought my GDEY0583T81 screens — they're a reliable source for e-ink panels with decent documentation. Waveshare has development-friendly displays with breakout boards. AliExpress has raw panels, but quality and compatibility are hit-or-miss.
For prototyping, Waveshare is easiest — they document their stuff well and include example code. For production, you'd need to go closer to the source, which means dealing with distributors and minimum order quantities.
The display I'm using
I went with 5.83" e-ink displays — same as Diptyx. Specs:
- 648 × 480 resolution
- ~140 PPI (not retina-sharp, but readable)
- 4 grayscale levels
- SPI interface
- No touch, no frontlight
Two of these side by side give you roughly the reading area of a paperback spread. Not huge, but comfortable.
What I'd upgrade to (eventually)
If Verso succeeds and I make a V2, here's what I'd want:
- Higher resolution (300 PPI like Kindle Paperwhite)
- 16 grayscale for image support
- Frontlight for reading in low light
- Touch for navigation (though buttons work fine honestly)
Each of these adds cost and complexity. The current spec is deliberately minimal — prove the concept first, then improve.
Here's the reality though: the best e-ink panels — the ones in Kindles and Kobos with 300 PPI, perfect contrast, and fast refresh — aren't available to small buyers. E Ink Corporation works through partnerships, and getting access to their premium panels requires volume commitments and real money. That's the kind of thing that only becomes possible if Verso gains enough traction to justify the investment.
The takeaway
E-ink is fascinating technology with real constraints. It's not just "a slow LCD." The physics are completely different, and understanding that shapes every design decision.
If you're thinking about building something with e-ink, spend time learning the display technology first. It'll save you a lot of confusion later.
More technical posts coming as I figure things out. Follow on X.