Thread: Hisense A5C, A5 Pro CC, color e-ink phones
View Single Post
Old 06-01-2020, 02:10 PM   #75
mdp
Wizard
mdp ought to be getting tired of karma fortunes by now.mdp ought to be getting tired of karma fortunes by now.mdp ought to be getting tired of karma fortunes by now.mdp ought to be getting tired of karma fortunes by now.mdp ought to be getting tired of karma fortunes by now.mdp ought to be getting tired of karma fortunes by now.mdp ought to be getting tired of karma fortunes by now.mdp ought to be getting tired of karma fortunes by now.mdp ought to be getting tired of karma fortunes by now.mdp ought to be getting tired of karma fortunes by now.mdp ought to be getting tired of karma fortunes by now.
 
Posts: 1,481
Karma: 9010563
Join Date: Jul 2013
Device: none
Quote:
Originally Posted by ottischwenk View Post
If it is in colour, then it will be displayed in colour, but with 100 dpi - unreadable
But it is not clear how it is engineered, hence detailed pictures should be seen.
The concepts we are using here are strongly metaphors, not facts: it is certainly tinted, it does somehow build colour, but the perceptive trick there is much more unusual than RGB and CMYK: it does not even have black as K, the closest it has to K-black can be the dots partially overlapping the gaps between the tinted bars... Not to mention it does not have white (what I called "mélange")!

So: how clever will be the algorithms that detect what to render and how? What is rasterized and what is kept as vector, and how are the subpixels treated in the rasterization?

Because, just by thinking a bit on it, one could for example keep full-res vector, e.g. rendering this text, and still tint it red or mauve - by rendering anti-aliased black (meaning grey-surrounded black) and brightening it where that black overlaps the red-tinted bars of the filter (or similarly other RGB composition). Certainly this is not as clean as the rendering of a black vector graphics, first of all, but especially one should wonder whether such effect (as I proposed it) has an acceptable result.

Not all tricks are feasible, nor all tricks are employed. In fact, for example, such device will have to rely heavily on A2 to display video. Now: will such A2 be employed cleverly or not? Because, for example, I have seen devices using a Floyd-Steinberg dithering for A2 rendering: which has the gigantic drawback of using very heavy randomness in a hardware technology which has the characteristic of having a breakthrough low cost in retaining dot state, and conversely a massive cost in changing such dot state (of dots which also have a limited lifetime) - thus losing the power efficiency of E-Ink's EPD. Clever dithering on EPD not only generically looks nice: it is also supposed to change as little dots as possible. Implementation not granted at all.
mdp is offline   Reply With Quote