[mistercow]

> It seems obvious that a higher perceived brightness can be achieved for any given pixel if using greyscale instead of using colors, just because less of the backlight's area is occluded.

When converting to grayscale, you typically calculate the value of the pixel and then set all color components to that value. The point of this is to keep the luminance the same as it was in the original color pixel. If you’re doing this correctly, the perceived brightness stays the same.

And just as a smell test: have you ever converted an image to grayscale and flinched away because it seemed twice as bright? Of course not; it just loses its color.

The only way you would get more perceived brightness at lower backlight intensity would be if you physically removed the color gels that overlay the LCD matrix. Which is obviously not what they’ve done here.

I’m pretty sure the increase in battery life they observed is simply because they’re using their phone less, which is very much the main upshot of the other benefits they listed. The idea that color pixels drain more energy is just obviously nonsense.

[relaxing]

> When converting to grayscale, you typically calculate the value of the pixel and then set all color components to that value. The point of this is to keep the luminance the same as it was in the original color pixel. If you’re doing this correctly, the perceived brightness stays the same.

This is precisely wrong.

To maintain perceived brightness, you want to set the color channels to the original value weighted by a scale factor that accounts for the eyes’ sensitivity to each color channel.

[mistercow]

Fine, luma, not luminance. But what you're describing is exactly that calculation. This does not change my point. Again: If you’re doing this correctly, the perceived brightness stays the same.

[ssl-3]

> And just as a smell test: have you ever converted an image to grayscale and flinched away because it seemed twice as bright? Of course not; it just loses its color.

Of course that is the way it is normally done. But it does not have to be done that way; does it?

Guidelines and norms are meant to be bent.

To demonstrate: I may be old, but I've definitely owned monochromatic "paper white" VGA displays that only responded to one channel of RGB and ignored the remaining two; the other two pins weren't even present in the connector. (These were trash for displaying color images without special care in software-world, but they were cheap.)

> The only way you would get more perceived brightness at lower backlight intensity would be if you physically removed the color gels that overlay the LCD matrix. Which is obviously not what they’ve done here.

We can evaluate that.

Suppose we have one ideal backlit pixel displaying FF0000, and that to achieve FF0000 66.6% of the backlight's total output for that pixel are being blocked and only the remaining 33.3% gets transmitted. Two subpixels are occluded; one subpixel allows transmission.

Suppose that with a backlight intensity of 100%, this pixel has a luminous output of 1 unit. (1 of what unit, you ask? For our purposes, it doesn't matter -- it's just 1 unit.)

Now, suppose we double the area of exposed backlight by instructing our pixel to display FFFF00. Our backlight intensity remains 100%, but we have twice as many subpixels allowing that backlight to be transmitted. The backlight stays the same, but our measured luminous output for this pixel is now 2 units.

To continue: FFFFFF. All 3 subpixels allow light transmission. Our luminous output is now 3 units for the same level of backlight.

Thus: With this result of 3 units, we've got 3 times as much light as we had at the beginning -- for the same pixel, with the same backlight, and about the same energy use.

---

To get back to the same 1 unit luminous intensity as we had with FF0000 @ 100% backlight, we can run the backlight at 33.3% for FFFFFF. This saves power.

Our pixel is producing 1 unit of luminosity with FFFFFF, but with only one third of the backlight required to display FF0000 with 1 unit of output.

> The idea that color pixels drain more energy is just obviously nonsense.

That's not obvious to me at all for a backlit LCD.

FFFFFF is always going to be brighter than colors like FF0000, 00FF00 and 0000FF are for a given backlight intensity, which permits the opportunity to reduce the amount of backlight provided, use less energy, and still provide the same luminosity as a color would.

And it accomplishes this without stripping layers out of an LCD panel, or using magical thinking.

Is that what was proposed? Fucked if I know. I find that articles like this (and approximately anything else that has ever been published with the words "you" and "should" juxtaposed in the title) are meant to make people bicker about dumb shit, and I try to avoid poisoning my brain by reading them.

[mistercow]

So specifically on a backlit LCD screen with dimming zones, and in the specific case where an entire zone would have originally been blue or red (which are perceptually dimmer), you could plausibly get a small amount of energy savings. But “twice the battery life” from this is not plausible.

[ssl-3]

Specifically backlit LCDs. Maybe with dynamic backlight; maybe even with zoned dynamic backlight -- but definitely backlit LCDs. It only works in corner cases where important information would normally be displayed as bright saturated subpixels (colors), and even then it only really gets good with special dynamic processing (which probably doesn't exist).

So in terms of practicality: It has none.

And at best, all that these hypotheticals can do is stretch out a battery a little longer by reducing the intensity of backlight that is necessary in order to succeed at using the device in a given level of ambient light.

It has never been my intent to demonstrate "twice the battery life" ["with this one simple trick!"]. That's the path of nonsense, and of chemtrails. :)