Why wouldn't this work?

Why wouldn't this work?
Would there be any downsides?

Other urls found in this thread:

computer.org/csdl/trans/tp/2011/07/ttp2011071370-abs.html
twitter.com/NSFWRedditGif

got my dick stuck in a bottle

I only see parallelograms/diamonds. go back to preschool

Expense. /thread

what are the upsides?

>/threading your own post

The downside is that you'll have either shitty vertical lines or shitty horizontal lines, depending on orientation. Then consider how almost everything on your screen is practically a box.

literally everything is based off a matrix and this would fuck with that. every other pixel would be offset and you would have to correct this with software/hardware inside the monitor, because the video feed going into a monitor is standardized.
you would have shitty lines like the other user said. Think gaymen with no anti aliasing...
it would not increase pixel density.
the narrow angles of each color space would decrease light throughput compared to other designs.

1. Unnecessarily hard to manufacture. (If not impossible with current techniques)
2. Prohibitively expensive (mostly as a result of #1)
3. Lack of a clear-cut benefit. (It's much easier to just use smaller squares instead)
4. Significantly worse at representing text and straight lines
5. Stupidly hard to drive and compute with (how do you render on a hexagonal grid?)
6. Backwards incompatible with all existing technology

It would basically only be “good” for film and vidya, but the benefits would be marginal, the downsides would be massive, and the cost would make it unrealistic either way.

Programmers already can't into subpixel addressing, this would make things worse
Rastered content can't be shown without resampling.

I don't know but you can bet your ass they've done the math. You can read the shitty article on pixel geometry on Wikipedia to see that they do different setups for different sorts of displays.


Imo just rotate your screen a few deg and you'll see there's no point in Diamond pixels

What the fuck is it with you lazy faggots and not putting a proper background on images so it doeant look like this when i open it?

Forgot image

There actually exist some pretty strange pixel arrangements, although they all do largely align to some sort of grid (even if pentile is offset every other line).
Pic related is the bizarre pixel structure of the Apple watch

Galaxy S5

Interestingly, it seems that there is a paper on this pixel structure: computer.org/csdl/trans/tp/2011/07/ttp2011071370-abs.html

An advantage would be thin, repeating lines of inverse colors (yellew blue, red cyan or green Magenta). Currently these look identical to black and white lines. Very pointless though

Could you demonstrate what you're talking about with an image (please)?

Why is it that Apple is the only innovator around?

0/10 bait

He did the right thing. There was nothing more to say.

for sensors, you generally want more green spacial sensitivity than red or blue.

for LCDs, the shape of subpixels affects color control and switching speed, so you end up with things like long rectangles made of stacked chevrons for IPS, etc.

things generally are the way they are for pretty solid reasons, so you can generally assume there are cost and/or effectiveness issues underpinning designs you don't intuitively understand.

(5) is not a real problem. You would just do all the computation in a regular layout. Nothing would change. The conversion happens in the monitor with dedicated hardware (it's just a linear transformation, nothing difficult that requires any difficult algorithms).

Still, yeah, the problem is mostly that it's awkward in production without any upsides.

That would make it pretty damn hard to address.

You think anti-aliasing fonts with regular RGB pixels is hard? This would be an entirely new dimension of pain. Same when you want to display any kind of raster image. You'd need to filter anything, losing information in the process.

Seriously though, making this picture made be realize just how shitty that layout is for type.

Just make them smaller
Problem solved
:^)

>for sensors, you generally want more green spacial sensitivity

I know very little about sensors, please explain to me why they "like" green more?

If you make them smaller, any advantage you get from using that particular arrangement also vanishes.

It's not that sensors like green more.

It's that human eyes like green more. We can discern a lot more shades in the green part of the spectrum.

>man I wish my screen were hexagonal instead of square --No one ever

>using the smiley with a carat nose

Yeah but what if we make them at like 1200dpi?

I can explain it: your monitor probably has red (R/r). green (G/g) and blue (B/b) in this order, over and over again, a few million times.
For this explanation RGB would be white (capital letters mean full brightness) and rgb would be black (subpixels are off)

for vertically alternating black and white pixel-lines, you'd have RGB rgb RGB rgb RGB rgb.
but if you take Yellow (RGb) and blue (rgB), you'd have RGb rgB RGb rgB RGb... or differently written BRG brg BRG brg. this is not visually distinguishable from white and black lines. Same goes for red (Rgb) and cyan (rGB). I did make a mistake though, with green and magenta. There you'd get RgBrGbRgB etc, which doesn't have these RGB or BRG or GBR clusters.
This advantage is pretty stupid though, because you will almost never have that situation (though it might come up in other situations, I don't know)

I don't really think what he says is true. left most line is 0xFFF, 0x000, 0xFFF
middle one is 0xFF0, 0x00F
right one is 0xF00, 0x0FF

But at least on my monitor they don't look the same.

However the patterns on the right are actually
Left:
0x0F0, 0xF0F,
0xF0F, 0x0F0

Right:
0x00F, 0xF00, 0xFF0, 0x0FF
0xFF0, 0x0FF, 0x00F, 0xF00

But they all look like pretty even grey (the value depends on your monitor's gamma).

instructions unclear?