[Contest] ROM Battery War

[scottylove]

Which ROM/kernel has the best battery life?

This is the most-asked and most obsessed over question on these forums BY FAR. So I created this thread in order to compile a comprehensive guide to ROM battery life, based on AWAKE TIME. Up Time means nothing in this competition because according to HTC, the Incredible has the ability to go 146hrs of standby time (aka. up time) ON THE STOCK BATTERY!!!

The premise is this, show off your best battery stats on your favorite ROM/kernel combo by posting in this format:

[Rom Name & version] ex. Virtuous 3.1
[Kernel Used] ex. Stock HTC v4
[Battery Used] ex. HTC 2150mah
[Up Time] ex. 14hrs
[Awake Time] ex. 5.5 hrs*** What this competition is really looking for.
[Screenshot of Battery Use] (only way i know how to verify actual time since unplugged)

Just in case people don't know this, if you plug in your phone, IT DOES NOT RESET YOUR UP TIME OR AWAKE TIME. So be aware of this. So far the only way i know to check this is by requiring a screenshot of the battery use page, and verify that the "Time since unplugged" matches the Up Time (within 5 mins is fine). If anyone can think of a better/more efficient way to check this, please let me know

I will compile the results in this OP to come up with the results and rank in order of AWAKE TIME. So show off your best battery day and let the ROM Battery War begin!

 

[Nayners]

Relocate....

To General?

 

[phoenixs4r]

To General?

Sent from my ADR6300 using XDA App

 

[Epicardium]

Which ROM/kernel has the best battery life?

This is the most-asked and most obsessed over question on these forums BY FAR. So I created this thread in order to compile a comprehensive guide to ROM battery life, based on AWAKE TIME. Up Time means nothing in this competition because according to HTC, the Incredible has the ability to go 146hrs of standby time (aka. up time) ON THE STOCK BATTERY!!!

The premise is this, show off your best battery stats on your favorite ROM/kernel combo by posting in this format:

[Screenshot of Battery Use] (only way i know how to verify actual time since unplugged)

Just in case people don't know this, if you plug in your phone, IT DOES NOT RESET YOUR UP TIME OR AWAKE TIME. So be aware of this. So far the only way i know to check this is by requiring a screenshot of the battery use page, and verify that the "Time since unplugged" matches the Up Time (within 5 mins is fine). If anyone can think of a better/more efficient way to check this, please let me know

I will compile the results in this OP to come up with the results and rank in order of AWAKE TIME. So show off your best battery day and let the ROM Battery War begin!

Ok, so I think this is an interesting question. It's valid as well. BUT, I must stop short of saying it's scientific and would produce meaningful results. It will not. I admire your enthusiasm and initiative, but a casual investigation of empirical results on this forum suggests strongly that no two people with the same combination of kernel/ROM are likely to get exactly the same results. One of the biggest factors, often ignored, is your baseline signal strength. CDMA phones are constantly in communication with the towers, and the strength of those signals impacts directly the duration of a battery cycle.

A more scientific study may go something like this: (and you would have to get everyone to support this unified protocol)
1. Charge you battery completely (be it bumped, not bumped, turned on, turned off, what have you...)
2. Turn the phone on (if it's off) and be sure it's in airplane mode. Now we've taken towers and radios out of the equation.
3. Using a task-killer, be sure all non-essential tasks are dead.
4. Ensure that background data and autosync are off. They may try to communicate at set intervals even in airplane mode (and eveyone has different intervals set)
5. Now, so that everyone doesn't have to wait around for three days as their battery slowly drains, you must begin a standardized battery drain protocol. I would suggest setting up a live wallpaper (one of the more intensive ones) and then setting the 'display timeout' to 'NEVER'. Just let it sit there and check the battery level at certain intervals. It doesn't matter what exactly the intervals are, just that they are recorded. Then, we could produce some graphs.

This would produce more meaninful results, but good luck at getting many people on board. Until then...just stick with the method of trying the combos out for yourselves. Don't waste time reading a bunch of other people's experiences that are completely irrelevant to you. And yes, I realize this is a potential thread killer. I'm just too passionate about science to let these things get very far

 

[Ianxcom]

I dont believe AOSP roms show awake time, only up time.

 

[scottylove]

Ok, so I think this is an interesting question. It's valid as well. BUT, I must stop short of saying it's scientific and would produce meaningful results. It will not. I admire your enthusiasm and initiative, but a casual investigation of empirical results on this forum suggests strongly that no two people with the same combination of kernel/ROM are likely to get exactly the same results. One of the biggest factors, often ignored, is your baseline signal strength. CDMA phones are constantly in communication with the towers, and the strength of those signals impacts directly the duration of a battery cycle.

A more scientific study may go something like this: (and you would have to get everyone to support this unified protocol)
1. Charge you battery completely (be it bumped, not bumped, turned on, turned off, what have you...)
2. Turn the phone on (if it's off) and be sure it's in airplane mode. Now we've taken towers and radios out of the equation.
3. Using a task-killer, be sure all non-essential tasks are dead.
4. Ensure that background data and autosync are off. They may try to communicate at set intervals even in airplane mode (and eveyone has different intervals set)
5. Now, so that everyone doesn't have to wait around for three days as their battery slowly drains, you must begin a standardized battery drain protocol. I would suggest setting up a live wallpaper (one of the more intensive ones) and then setting the 'display timeout' to 'NEVER'. Just let it sit there and check the battery level at certain intervals. It doesn't matter what exactly the intervals are, just that they are recorded. Then, we could produce some graphs.

This would produce more meaninful results, but good luck at getting many people on board. Until then...just stick with the method of trying the combos out for yourselves. Don't waste time reading a bunch of other people's experiences that are completely irrelevant to you. And yes, I realize this is a potential thread killer. I'm just too passionate about science to let these things get very far

I agree this is not a laboratory experiment with variables and controlled environments. I did not intend it to be this way, this is a statistical survey that would seek to find maximum and mean REPORTED battery times by respected members of this forum. Surveys and polls are used to predict presidential elections, determine product interest for advertisers, and for scientific research.

I appreciate your input but do not discredit valid research methods because it cannot be controlled. That is not the point!

 

[Epicardium]

lies, damn lies, and 'you know what'

I agree this is not a laboratory experiment with variables and controlled environments. I did not intend it to be this way, this is a statistical survey that would seek to find maximum and mean REPORTED battery times by respected members of this forum. Surveys and polls are used to predict presidential elections, determine product interest for advertisers, and for scientific research.

I appreciate your input but do not discredit valid research methods because it cannot be controlled. That is not the point!

If this is an opinion poll, then so be it. But, I hope you're not insinuating that opinion polls are reliable; particularly political ones. Rather, I would say this kind of data collection is more likely to lead people astray, or confuse them all the more. And we all know what they say about damn lies, err, statistics...

No harm no foul. Carry on.

 

[ToyTank]

This should be posted in general.

 

[bluekornchips]

Feather's already got this:
http://androidforums.com/incredible-roms/199848-rom-feather-v1-7-a.html

 

[larsrya8]

This should be posted in general.

Then report the thread and request that it be moved to General. I don't know if I was the only one reporting threads, but since I stopped a few weeks ago the Development forum has become a complete mess. If you (or anyone else) want to take up the mantle, feel free.

 

[scottylove]

If this is an opinion poll, then so be it. But, I hope you're not insinuating that opinion polls are reliable; particularly political ones. Rather, I would say this kind of data collection is more likely to lead people astray, or confuse them all the more. And we all know what they say about damn lies, err, statistics...

No harm no foul. Carry on.

I'm not over serious about this thread, or statistics. Never did i say anywhere this would provide conclusive, anal, scientific facts. I just thought this could be kinda fun getting people trying to best each other in battery life and let the ROM fanboys go at each other, while it also might provide some interesting info to boot. Didn't know some people took this so serious!

 

[kungfuturkey]

I think everyone is being a wee bit to serious on this one... I will post screen shots this weekend, but in the past I have made it to 15hr 46min on one charge with moderate use... I will see if I can duplicate, but This very well can show people which Rom/Kernel combinations are best... or atleast worst think about it if you have a 6 hour difference average between two ROM/Kernel combos you would know which one to pick right? Give the guy a break it is a good poll to have!

 

[Ianxcom]

This is a very real concern which has been plaguing the Dinc since the beginning. Im all for it if we can knock out as many variables as possible and make sure each phone tested has all the info available.

For example.

Rom
Kernel
Radio
Battery
OV or not


Maybe like do a different combination of the above each couple of day or week. Like in the thread title put [Now testing skyraider with kkbfs#5 9.01 radio].

Or ya know like whatever

 

[unreal2k]

I think everyone is being a wee bit to serious on this one... I will post screen shots this weekend, but in the past I have made it to 15hr 46min on one charge with moderate use... I will see if I can duplicate, but This very well can show people which Rom/Kernel combinations are best... or atleast worst think about it if you have a 6 hour difference average between two ROM/Kernel combos you would know which one to pick right? Give the guy a break it is a good poll to have!

EDIT: HA! as i was writing my post lanxcom above posted the same suggestion


I agree, give the guy a break.

My observation on battery is that it does play a huge role on what rom/kernel you have. That's a given.
The problem is that there is no standard on usage; moderate, heavy etc. means different things to different people.
Another problem is conditions and differences in process power from apps. Angry birds for example uses a lot of battery but some other games do not. Same goes for Skype that doesn't seem to die and always lurks in the background.
Finally, you are running into all sort of other situations such as different battery brands and mah, different types of amoled/lcd screens might be an issue, bump charging etc.

Well, you get the point. There are so many variables and unless you want an actual statistical analysis that is going to be as close to accurate as it will ever get, you will have to standardize the situations as much as possible.
A group of people willing to test rom/kernels on their phones and keeping the variables constant will be the ideal case i can think of.

I'm not trying to bash anyone, i'm trying to help. Cheers!

 

[setexascustoms]

Here's what I got, all the device info's in my sig. Typical day for me, nothing special. I do have the Seidio 1750mAh battery and bump charged.

 

[ToyTank]

This is the best battery life I ever had, other issues made me switch though. I don't use that set up anymore, had to reboot every 24hrs or phone would crawl.

I am using 9.01 radio now, and I am seeing good battery life still.

Screenies are current setup. Very light use while hunting, but great uptime.

-HTC1500 Batt. External charger to 4.223v
-Adrynalyne experimental SS UV Kernel.
-Skyraider 3.3.1 Sense
-9.01 Radio
-No SetCPU
-Autobrightness
-GPS ON
-Wifi ON
-Sync ON

Sorry if post is sloppy, I haz need sleep.

 

[mikeymop]

This is my best battery life:

Consistant texting, a few phone calls. I think two pictures, 1:30-2:00 of angry birds and 1:30 or gameboid. They disappeared from battery stats after awhile.

I turned off 3g at night when I was sleeping, no Juice defender, no setcpu. Cyanogen 6.0.2 ril fix and King's CFS #2 (AOSP)