LG Philips with first flexible 14.1" color E Ink

[nekokami]

Right now they're using separate particles for black and white, with one being positively charged, the other negatively charged, and they apply a charge to the top and bottom of the column the particles are in to cause the two types to move to the top or bottom. When they used 2-color particles, one side had a positive charge, the other a negative charge. I don't know how they'd manipulate particles with more than 2 sides. (I don't think we have that much direct control over quarks yet. )

[yvanleterrible]

If my memory serves me right gray levels are achieved by partly rotating bi colored particles.
Putting multiple colo(u)rs on the same particle would defeat the whole concept of colo(u)r displaying because each colo(u) has to be dosed grayscale style to work effectively; making this thought next to impractical.

An other concept against multiple colo(u)r particles is that if you think at it closely again, when a bicolored particle displays a gray tone it is rotated partly upwards ie. if a colo(u)r particle was there instead, it would show.

[nekokami]

What I meant about cells with different colored particles, btw, was that in one cell there would be white particles and magenta particles, in another white and yellow, in another white and cyan, and in another white and black (the current mode). The cells would be grouped together in blocks, and each block would represent one pixel. You'd need the equivalent of 2x the DPI you want (since DPI is measured linearly), and manufacturing adacent cells with different colored particles would likely not be possible using the current process. But the result would be similar to current 4-color printing on paper.

If the cell method of white particles in black oil were modified, you'd need 4 colors of oil in adjacent cells, which would be at least as difficult, if not more so, than 4 different colors of particles in adjacent cells.

Two-sided particles in white and 4 different colors would probably be equally difficult to manufacture.

[NatCh]

And the difficulties attached to color e-ink begin to come into sharp focus.

[yvanleterrible]

Sorry about the mixup neko I was anwering Jadon.

[Jadon]

Originally Posted by yvanleterrible

If my memory serves me right gray levels are achieved by partly rotating bi colored particles.

If you can partly rotate the particle, you can rotate it so any side is up. If the particle is a cube, say, you can show red, green, yellow, blue, black, white. Each side pure, each side showing only that side, no partial sides exposed. (Not that cubes would be best, since if you pack them close, you can't rotate them. A three-sided prism could have a flat side always up, and the other two sides not showing. Even better are more complex shapes, like three circles intersecting at right angles. Except hexagons work better for packing purposes, with them rotating along point-to-point diameters, always showing a single color plane face on, with intersecting planes edge-on [and you could color the edges to match the planes]). You never show mixed colors within a single cell, never half-black half-white to make grey. Each cell is pure, with mixing only intercell, although if the particles have enough colors, sides can come premixed.

[nekokami]

Originally Posted by Jadon

If you can partly rotate the particle, you can rotate it so any side is up.

Not so. If your ability to rotate a particle cannot control which axis it rotates on, but only the degree of rotation (which would be the case with electrical polarization), then your options are much more limited. If you were able to manufacture perfect cubes at that size (and this would be a lot harder than the approximately spherical particles I suspect they are using now), you could possibly get 3 colors: top, bottom, side. A prism shape allowing more colors would also not hide the colors not in use well enough.

I'm sure there are ways to manipulate the particles along more axes, but I don't see a way to do it using the top/bottom electrodes in the current generation of e-ink.

[yvanleterrible]

Also each colo(u)r has a thickness to it that will show even at 90 degrees. To be hidden it has to be under, and even then it is not effective enough, this is why our eink screens are kind of grey, the black shows through the particles.

The only practical solution now is smaller more numerous pixels,

[Jadon]

Originally Posted by nekokami

If your ability to rotate a particle cannot control which axis it rotates on, but only the degree of rotation

Then constrain your particles. If your particles are twice as long as they're wide, and are in cells slightly larger than them, they're only going to rotate on their long axis, never flip end over end. A three-sided prism, with a positive-charged flat-face and a negative charge opposite, on the line between the other two long faces, will rotate as you want. Two charge sites beneath can make it rotate as desired.

On the more complex shape, 3D intersecting planes, seeing any plane face on means seeing two others edge on, but obviously the edges are thin enough to be negligible (so instead of seeing a solid blue plane, you see on quartered by thin grey, say, lines), or thick enough that they can have a clor, and be colored to match the plane behind them.

As abstract geometry, it works. As fiddly electronics, it might require too-fine control, or too-expensive particles and chambers, or something else. That's just reality being disappointing: business as usual.

[nekokami]

Originally Posted by Jadon

Then constrain your particles. If your particles are twice as long as they're wide, and are in cells slightly larger than them, they're only going to rotate on their long axis, never flip end over end. A three-sided prism, with a positive-charged flat-face and a negative charge opposite, on the line between the other two long faces, will rotate as you want. Two charge sites beneath can make it rotate as desired.

You're right, but this doesn't change the number of options. You get one effective rotation axis either way.