It is my understanding that AMOLED screens consume no power for pixels that are set to black. It is also my understanding that phones are being released without a notification light on them (such as the Nexus S). It is also my understanding that Android phones expect to spend some time in the Idle state and do various update tasks, reduce overall power consumption etc. while in such a state so not allowing them to ever idle would not be a solution for any problem.
Assuming my above understanding is correct, how difficult would it be to make a change to Android such that notification information could be displayed on-screen in a few pixels (say, in the center of the screen) so that the majority of the screen is consuming no power (black pixels) but you still have a few colored pixels for notifications. These pixels could be set to black when there are no notifications so that the small power they do consume is only used when a notification needs attention.
A step beyond this would be to create an interface that would allow other developers to create various themed idle screens that would be customizable by the user, or at least by a variety of developers. This may be as simple as changing the colors, rate of flashing (if any), number of notification pixels used, etc. Also, different apps could notify in different parts of the screen to give further feedback as to what needs attention without having to touch your phone (just a quick glance).
My background is that of a C++ Software developer who has dabbled a bit in Android, writing various small apps for it. I have never worked with the OS itself before so I am uncertain if this would be a trivial task or a very complex one. Perhaps it is even something that could be done without any OS changes (though to my understanding this is not the case).
Check this thread/app: http://forum.xda-developers.com/showthread.php?t=730692
Sent from my GT-I9000 using XDA App
OK I was going through settings like I occasionally do since I sometimes find something interesting and I did today under the screen settings and the mode selection.
There are three settings:
Dynamic
Standard
Movie
The default I'm pretty sure (since I can't remember after messing with it) was standard. I'm curious as to what dynamic is. Movie seems fairly obvious to me since it adds a slight yellow hue to the screen. Dynamic seems to brighten colors it seems but why is it referred to as dynamic.
I haven't seen the screen adjust itself or anything upon using it either.
Check this thread:
http://forum.xda-developers.com/showthread.php?t=1229470
From what I've observed, dynamic gives out brighter and better colors at the expense of battery life. Standard is pretty much "standard". Then movie mode uses the least battery and would not strain your eyes as compared to the other two.
Ah thank you for that!
EDIT: Please jump to post 4 for updated test results . Come back to this post for why this thread
There are umpteen threads on battery saving tips / methods and a few on AMOLED screens and battery savings. So why a new thread, could be the Question
Because
1.Battery saving means are meaningful if backed by numbers, else it’s subjective
2.Tests below were done on N7100, but given the popularity of AMOLED (two dozen and counting per Wikipedia), across cell phones and tabs, this may be interest to battery freaks and hence posting in general section rather than N7100 threads
3.Provide a replicable methodology of testing on your AMOLED device (rooted) both manually and automated. There are threads on XDA as well that suggest testing methodology but have not come up with easy testing methods and automating it unless I missed
I am NOT a qualified guy to understand the technical details of testing screen power nor of the methods/ apps used. There is enough material on the net to show how complex it is and savings using using dark screens vary from 18 to 41 %. IMHO, the jury is out and appears to be taking a long break
All I have done is put together various pieces to show a way of testing it and come up with results that are indicative and demonstrate trend lines.
Also the apps chosen for testing is not be misconstrued as canvassing for them.
There may be better ways of doing this test using other apps or better testing ways, which you are welcome to share here.
Test Device Configuration
GT-N7100 (t03g), ROM- Deep Impact (Morph variant) (Kitkat) by @Sammy_052
Kernel Agni- 3.0.101-India-v4.4.2-OC-INTL-STOCK_HYBRID-KitKat
Apps Used for testing
3C Tool Box Pro: Has a nice feature showing screen consumption in mAh (though on devices like mine which don’t have a current sensor, it is derived from battery drain percentage). Plus, it has the advantage of modules that can be accessed and test automated. Immense thanks to @3c , developer, who helped a ton
Darker : Screen Filter that was thoughtfully provided in the ROM. Free version allows darkening up to 50% as used for test
Stay Alive : To keep the screen alive for two hours (being the test duration), since the maximum screen time out is limited to 10 minutes in KitKat. 3C Toolbox offers a way of choosing apps , which keep screen on, as long as apps are kept in foreground. For some reason, this did not work with Nova Launcher as chosen app, hence the need for a different app. There are many free apps on Play Store with similar functionality and may be used after testing that they do not impact 3C
MacroDroid : To automate the test process. This is by @UndeadCretin and is a great tool to automate without a steep learning curve. The aim of click starting the test and getting results in mail box has been achieved (barring manual intervention in the first minute of starting the test) . The macro used for this is posted on the app thread. This macro can be easily tweaked and battery bugs can have a field day trying things like Live wall paper, under volting/over volting and altering CPU / GPU frequencies to see impact on power
Broad test Methodology (details later in thread)
Screen power usage was tested in 3 cases. Cell was kept in airplane mode to get cleaner results (optional) and Stay Alive used to keep the screen on for two hours ( targeting to get a drop of 10% battery level) for each test.
Case 1: Bright wall paper
Case 2: Bright wall paper with screen filter
Case 3: Solid black wall paper
Results of Testing ( files attached )
Case 1: Bright wall paper: Screen consumption is 195 mAh
Case 2: Bright wall paper with screen filter: Screen consumption is 189 mAh
Case 3: Solid black wall paper: Screen consumption is 172 mAh
Conclusions
1. Using a solid black wall paper as compared to bright wall paper shows 11 % savings.
2.Using a screen filter on bright wall paper shows very little savings 3% . Was surprised at this result and repeated the test and the results were practically same. This disputes the accepted belief that screen filters save power but is savings as low as this?
Internet reports savings from 18% to 40% by using black screen as compared to 11 % above, am guessing are due to a bunch of reasons, like testing methods, screen size, age of device, device to device variations, ROM, kernel settings, GPU/CPU voltage and frequency to name a few at device level .At next level, is the apps used, usage pattern of phone, algorithms used for dimming the phone and many more.
Testing as above is not real life. To carry out real life testing, ideally one would need to replicate activites for a fixed duration like calling, texting, browsing, watching stored video, on line music / video. I have not come across any app or way of doing this and would be happy to give it a shot, if someone can suggest a simple way
Detailed Test Instructions - To provide consistent testing environment and get clean results
1.Disable scheduled running of apps during the test and Greenify
2.Set the wall paper which you want to test and ensure sufficient battery
3.Run Stay Alive and use the second icon (red one) in the notification bar
4.Run Darker at your preferred settings for tests with screen filter
5.3C settings ( Thanks to @3c )
a)Settings>Recorder>Continuous recording enabled
b)Settings>Recorder>Time frame and Rate- depending on how long (maximum) you want to test and how frequently you want data to be updated, choose appropriate option . I used 12 hours, 10 minutes
c)Settings>Battery>Monitoring> mA retrieval method (if your device is listed choose, For Samsung devices, on enhanced estimated or estimated on the other side is pretty stable and the same for all devices, but often slower to update.
d)Settings>Battery>Monitoring>Estimate, use estimates drain, or if your device has a current sensor and reliably works choose use current sensor
6.Create shortcuts on home screen for battery statistics module of 3C using 3C widget
7.Start the test and go to markers tab and set the marker. You can create markers from the action bar or from history tab long pressing a record. Markers shows average consumption, be it mA or %/h (tap the value to switch). Min/max shows the range for the % or mV. Remember to show full history (not just % changes) by long-pressing the options and tapping “full history”.
8.Time the test and take screen shot of markers results
9.Repeat test for different cases
Looking forward to your views
Nice
Wow, this is a totally put together review. AmoLED screens save a lot of power on black, unlike LCD backed by light from LED where the entire screen needs to be on at all times when the screen is displaying anything. LED screens do not have a back light to support them in producing light. The individual pixels themselves produce the light. Hence if part of the screen is black, it means that, that part of the screen is off.
I had thought this through when making the rom in the first place and hence was the need in keeping all the elements either complete black or white. But samsung as a company decided to go away from this beautifying and practical aspect and the later phones have begun shipping with rather different set of colors.
Sammy_052 said:
Wow, this is a totally put together review. AmoLED screens save a lot of power on black, unlike LCD backed by light from LED where the entire screen needs to be on at all times when the screen is displaying anything. LED screens do not have a back light to support them in producing light. The individual pixels themselves produce the light. Hence if part of the screen is black, it means that, that part of the screen is off.
I had thought this through when making the rom in the first place and hence was the need in keeping all the elements either complete black or white. But samsung as a company decided to go away from this beautifying and practical aspect and the later phones have begun shipping with rather different set of colors.
Click to expand...
Click to collapse
Thanks for your view. It is fitting that you as ROM developer, included this aspect in your ROM and that is what triggered me thinking.
The test above is representative of a small portion of SOT, maybe around 10%. Real life usage is what needs to be measured and am searching for an app that can run fair amount of real life use cases. Only Antutu does something similar but requires user intervention to run. Ideal is auto play so that it can be automated through MacroDroid. Your thoughts around this are welcome
Sent from my GT-N7100 using XDA Free mobile app
Updated test results
So, i was stuck for long to do real life testing and it is done now !!!
A few major changes from the previous set up
1. ROM changed to wesam.othman v8 by @wesamothman
2. Better Battery Statistics (BBS) used in place of 3C Tool Box mainly because of ease of setting up and being more popular, eliminating process described in OP. But IMO, 3C Tool Box is far more versatile and powerful in studying battery usage
3. Most importantly, discovered Battery HD app (free) by Small Tech (https://play.google.com/store/apps/details?id=ch.smalltech.battery.free&hl=en). Very hand app and well designed. This app has on line calibration features using wifi browsing, streamning music and video. I wanted to extract these modules and incorporate in the macrodroid macro (part aim of this test is automation as mentioned in OP). Requested the developers and amazingly, they jumped in to help and created separate links for each of these modules so that i could call them in the macro to execute without any manual intervention. Extremely thankful to Santiago and Vadim of the team who made it possible :good:. Without their help, automation of this test could not have been possible. Thanks again !!!
Coming to the test , real life usage to the extent possible has been simulated except for gaming, calling , texts and social media apps. Test consits of 30 minutes of browsing, 15 minutes each of streaming music and video. Screen was kept on during the test duration and darker (screen filter) when used was set to 50%. Pics attached
Results
Case 1: Bright wall paper- Batterydrain was 10 %
Case 2: Bright wall paper with screen filter - Batterydrain was 9%
Case 3: Solid black wall paper - Batterydrain was 8%
So, this may be interpreted as- if you get 5 hours SOT for instance using a bright wall paper , you would have drained 50% of battery. But with using a screen filter on the bright wall paper, after 5 hours you would have used only 45% and the extra 5 % (over usage not using screen filter) would give you additonal 30 minutes of SOT !!!
And if you were using a solid balck wall paper instead, that would have given you an additional hour , making it 6 hours of SOT !!! ( Of course, this is assuming same usage and linear battery drain)
So, it is your choice, you can test it yourself
Happy battery life...cheers
I've searched the forum and completed several google searches using various words or phrases, but nothing comes close to my question.
Anyone having screen flicker issues on auto brightness? I've had my phone 4 days and have continually observed the screen incessantly increase and decease in brightness, particularly noticeable at night, probably because I'm finally able to sit down with the device.
Along with aggressive memory management closing apps I want open and the long in the tooth notification isssue, I'm up i the air as to keeping the phone.
Kindly chime in, maybe I just got a lemon.
Yesterday i noticed something strange in my OP7p and i'd like you to check whether something like this occurs in your phones. Issue, which i will describe occurs only when i have 90hz refresh rate set and the low brightness of the screen with the auto adjustment of brightness turned off. The easiest way to see that is on youtube application with a dark theme. When we start a movie on the full screen (horizontally) and watch it for a few seconds, then go back to the vertical position, there is something like weak but visible blink of the screen, or a minimal colours change. It can be noticed in other apps, e.g. on facebook. Seems to me this is a software problem, beacause the same phenomenon doesn't exists when we set 60hz refresh. I would be very grateful if you can check it on your phones and give me a know about it.
I don't usually start videos in full screen mode and never noticed it but yeah definitely does a white flicker.
This is the highest brightness I could get it to do the flicker at.
Thank you for an answer. Personally i am able to see this flicker on a little bit higher screen brightness, but maybe its because im sitting in a dark room now. I was worried cause sometimes i see it during the normal usage of phone, and if only my OnePlus would be affected by this flickering i could still return it or exchange for a new one. It isnt very disturbing however interesting is whats cause of it.
kamkoz96 said:
Yesterday i noticed something strange in my OP7p and i'd like you to check whether something like this occurs in your phones. Issue, which i will describe occurs only when i have 90hz refresh rate set and the low brightness of the screen with the auto adjustment of brightness turned off. The easiest way to see that is on youtube application with a dark theme. When we start a movie on the full screen (horizontally) and watch it for a few seconds, then go back to the vertical position, there is something like weak but visible blink of the screen, or a minimal colours change. It can be noticed in other apps, e.g. on facebook. Seems to me this is a software problem, beacause the same phenomenon doesn't exists when we set 60hz refresh. I would be very grateful if you can check it on your phones and give me a know about it.
Click to expand...
Click to collapse
this is usual for EVERY Oled screen and is called jelly effect...
You might want to try the DC dimming option in Settings->Utilities->OnePlus Laboratory.
matze19999 said:
this is usual for EVERY Oled screen and is called jelly effect...
Click to expand...
Click to collapse
From what i know jelly effect appears while we scrolling some content and propably that's not what i meant in my first post. Also flicker in my case is only visible on 90hz refresh rate.
I have galaxy s7e as a second phone and that effect doesn't happen there.
vishal3967 said:
You might want to try the DC dimming option in Settings->Utilities->OnePlus Laboratory.
Click to expand...
Click to collapse
Thanks but i tried this option and it didn't do the job well.