A new technique to help reduce the power used by mobile device screens has been proposed (Abstract, Full text [PDF]), called FingerShadow. The technique dims the screen when the user touches it, at the touch point and the area below where the finger would be, as this area is unlikely to be visible to the user while they interact with the device (an example demonstrating this is shown on the phys.org site, and in the full paper).
In their paper, they proposed the screen-power saving technique based on user touch interactions, FingerShadow. This technique applies local dimming to the screen areas covered by user fingers. As the user cannot see the screen areas covered by fingers, FingerShadow is able to save power without compromising the user visual experience, they said. The authors reported the technique showed potential with about 12.96% in average of screen power saving, with 22.32% at most.
"We have studied 10 users' touch interaction behaviors and found that on average 11.14% of the screen were covered by fingers. For these 10 users, we estimate that FingerShadow can achieve 5.07%~22.32% power saving, averaging 12.96%, with negligible overhead."
The paper discussed the advantages of their proposed technique. "FingerShadow has several advantages. First, the dimming area is located by the finger touch/hover operation, and the sensor call back process's calculation load is ignorable. Second, the pixel rendering process doesn't require intensive computation load, the target area could be directly turned down or dimmed to the predefined level. Third, the dimming layer introduced by FingerShadow is handled by dedicated display controller rather than GPU," they wrote.
(Score: 0) by Anonymous Coward on Saturday October 04 2014, @04:16AM
if this is true, why would there be any need for "the dimming layer introduced by FingerShadow [to be] handled by dedicated display controller rather than GPU"?
(Score: 0) by Anonymous Coward on Saturday October 04 2014, @04:22AM
also, on what planet does running a dedicated graphics controller constitute "negligible overhead"?
(Score: 0) by Anonymous Coward on Saturday October 04 2014, @08:43AM
Come on, they tested this on TEN whole phones. Don't argue with the study. Ten!
(Score: 3, Interesting) by darkfeline on Sunday October 05 2014, @02:05AM
A dedicated display controller might just be a tiny bit of logic hardwired and optimized for the calculations necessary. We're not talking about running a second GPU here. For example, I could design a circuitboard that performs a custom mathematical operation faster and with less power than running a full kernel and compiled program on a general-purpose CPU. In this case, using a dedicated controller would reduce the computation load of the process even more, and remove all impact on the phone's normal CPU/GPU performance.
Also keep in mind that displays burn a lot of power. You can do a lot of computations and still come out ahead in the power department (for some value of "a lot").
Join the SDF Public Access UNIX System today!
(Score: 3, Informative) by bob_super on Saturday October 04 2014, @05:22AM
Ok, I'm the exception because my phone has a real keyboard, but even then, I'm pretty sure that an average of 13% power saving is super-optimistic.
If dimming saves 90% of the power in the finger area, you are still implying that fingers cover 15% of the screen at all times... Even with man hands and a 3.5" screen, I don't see it. With the new 5.5" craze and girly hands? not gonna happen.
(Score: 0) by Anonymous Coward on Saturday October 04 2014, @08:12AM
For many LCD displays the light is from a backlight and the backlights aren't normally high res enough to dim just the finger portion.
So this concept will probably only work for LED screens and other screens where the light is from something with enough resolution.
(Score: 1) by axsdenied on Saturday October 04 2014, @12:10PM
They are talking specifically about OLED screens which do not have backlighting.
(Score: 2) by Tork on Saturday October 04 2014, @09:40AM
🏳️🌈 Proud Ally 🏳️🌈
(Score: 2) by frojack on Saturday October 04 2014, @10:33PM
I didn't do the math either.
I just turned on my phone, placed my finger on the center of the screen, noticed ZERO shadow in a normally lit room, and assumed these guys were blowing smoke.
There are so very few areas where you would actually cast a shadow that this is worthless. Even outside in sunlight, you have to shadow your phone to see it, and if the screen goes darker when you shadow it, that would be counter productive.
No, you are mistaken. I've always had this sig.
(Score: 2) by Tork on Saturday October 04 2014, @11:18PM
🏳️🌈 Proud Ally 🏳️🌈
(Score: 2) by Bot on Saturday October 04 2014, @11:00PM
Yes, the idea seems to have potential, but as a marketing bullet point. Now, if my phone didn't remind me the battery is slow by vibrating and lighting up the screen, which drains battery even faster, and ran a proper gnu/linux, like the nokia n900 could, I'd consider it good enough.
Account abandoned.