Related
I know how to use a search button, I have been on forums for a LONG time.
I also know how to use google. That being said, I have done hours (literally) of reading, learning, and attempting to choose and run the best kernel for my phone.
I would like some personal input from community if I could.
I have looked extensively into Imnuts and Nemesis2all's kernels.
I want one that will give me a tiny speed boost, but more so one designed to give me the longest battery life I can get while still containing the ability to run at a decent enough speed.
Any input would be great.
Edit-I guess it would be nice for you guys to know as well, I am running Evil Fascination 3.6.0, with the black mods =]
Ive heard both good and bad between these two kernels. Everyones phone is different and results will vary. Imnuts has a fixed undervolted setting either 100mv or 50mv but the limit is 1000mhz so theres no overclocking. Nemesis' OTB it is customized through voltage control app on which frequency steps you want to undervolt and you can overclock the kernel past 1000mhz for faster speeds. Ive also had good experiences with Comavolt (undervolted by -75mv) on SC 2.9.2 but I havent tried it out with CR 2.0 yet. If you arent into messing around with the voltage settings too much go with Imnuts or Comavolt. I suggest giving 4 days on each kernel with a battery stat wipe prior to see which one to settle with. Hope this helps!
Thanks man.
That's the kind of reply I was looking for. I think I am going to try out comavolt for a few days.
And do a full battery stats wipe, I may try and dig and see if I can locate an app that can log, and show me in much more detail where my battery life might be going, rather then just the silly android one in our menus ( I am fan of detailing when it comes to performance. )
I will post up in case any one else has concerns or questions whether to which to choose. Or if anyone else has an personal inout I would gladly take it into consideration.
Sn4ckm4st3r said:
Thanks man.
That's the kind of reply I was looking for. I think I am going to try out comavolt for a few days.
And do a full battery stats wipe, I may try and dig and see if I can locate an app that can log, and show me in much more detail where my battery life might be going, rather then just the silly android one in our menus ( I am fan of detailing when it comes to performance. )
I will post up in case any one else has concerns or questions whether to which to choose. Or if anyone else has an personal inout I would gladly take it into consideration.
Click to expand...
Click to collapse
CPU Spy and Spare Parts are two good apps for the kind of info you are looking for on battery usage and more.
Also, being that I am on evil fascination.
IS it considered to be more a AOSP rom, or is it something else?
And give the two options pretty much explained to me, does anyone know if theses are STRICTLY zip? Or can they be had in an ODIN compatible format as well, (hate flashing from sd card if I don't have to.)
I had been using Comavolt with Comm 1.3 until I went to PBJ with the Comm 2.0. I will see how it is with PNJ. I had very good result with Comavolt; so I hope that helps out.
EF is a touch wiz rom
Sent from my SCH-I500 using XDA Premium App
I am running otb. OTB lets you change the clockspeed and voltage. Effectively, you can undervolt, overclock, or both in whatever combination. I have a voltage control app on com rom that makes it easy. I'm not too aggressive, and I've not had a single issue out of it (performance or stability).
Posted from my Samsung Fantastinate using Voodoo magic
Personally, I believe that there's no point to overclocking your phone other than being able to say "my phone is overclocked." I've always gone for battery life when choosing kernels and I've never had performance issues. I've had good experiences with all of imnuts' 100-mv UV kernels. I just put SetCPU on conservative and forget it.
Sent from my SCH-I500 using XDA App
i'm running stock, debloated ED01 and nemisis2all's OTB 1.6 on stock settings flie for my phone and the battery life is great for me. you can tweak the clock speed and voltage really easily with the recommended voltage control app that's free from the market.
Please forgive me if this is not the appropriate place for this question.
There are quite a few kernels floating around. Which kernel is best in terms of stability, performance and most importantly, battery life? At the moment, I am not too interested in flashing my ROM to CM7 or Cognition.
Thank you in advance.
tboooe said:
Please forgive me if this is not the appropriate place for this question.
There are quite a few kernels floating around. Which kernel is best in terms of stability, performance and most importantly, battery life? At the moment, I am not too interested in flashing my ROM to CM7 or Cognition.
Thank you in advance.
Click to expand...
Click to collapse
There are only three custom kernels available right now. Only two of them actually boot. (Netchip's is nonfunctional) Your choices are codeworkx's stock kernel and mine.
Codeworkx's is primarily a reference kernel at this point - as far as I know it's a straight Samsung source build. It's missing a few performance and stability fixes (interestingly enough, even ones that codeworkx himself backported.) to my knowledge.
Stability and battery life are my primary design goals with my kernel. The only reason it even has OC is because that was low-hanging fruit and was a popular feature. If you don't overclock it should be stable. You may even want to try the Experimental 10/09 release as it's been good so far for everyone that has tried it except for one report of issues with overclocking - but that could just have been bad luck or a hot room. If you don't overclock it should be stable, and if it's not, keep the 10/09 non-experimental build on your SD card to flash from CWM.
Sent from my GT-P7510 using Tapatalk
Thank you entropy! Do you mind if I ask what kind of battery improvements you are getting over the standard kernel? I am currently rooted with the CW kernel.
Any progress with BLN on either Kernel at all?
tboooe said:
Thank you entropy! Do you mind if I ask what kind of battery improvements you are getting over the standard kernel? I am currently rooted with the CW kernel.
Click to expand...
Click to collapse
Most of them should be small. codeworkx's backport of the 2.6.35.14 cpuidle fixes helps in certain cases of extreme unusual drain, but usually it's just a small improvement.
Mostly just small improvements and better handling of some odd "corner case" drain situations.
clumsywob said:
Any progress with BLN on either Kernel at all?
Click to expand...
Click to collapse
Last time I tried to implement it, CPU deep sleep was inhibited when the light was on.
I'll give another go at it in a few weeks... At least it didn't crash like it did on the Infuse. I've been sick lately, AND the past week was a major development sprint, I am taking it slow for a bit so people can beat on what's there, especially the experimental release, which seems solid so far.
Entropy512 said:
Most of them should be small. codeworkx's backport of the 2.6.35.14 cpuidle fixes helps in certain cases of extreme unusual drain, but usually it's just a small improvement.
Mostly just small improvements and better handling of some odd "corner case" drain situations.
Last time I tried to implement it, CPU deep sleep was inhibited when the light was on.
I'll give another go at it in a few weeks... At least it didn't crash like it did on the Infuse. I've been sick lately, AND the past week was a major development sprint, I am taking it slow for a bit so people can beat on what's there, especially the experimental release, which seems solid so far.
Click to expand...
Click to collapse
Awesome, sounds like an improvement! Hope you feel better! Hopefully someone gets BLN or something similar working soon. I know it's minor, but I just can't understand why Samsung didn't implement that from the beginning or at least give us a LED for notifications.
Off topic, is there a fix out for the call logs to only show "calls" instead of calls and texts?
Hi,
Does anyone have any kind of summary of the available kernel options for the SGS2?
I've been running mostly stock firmware, though
a) using the cf-root kernel (for titanium backup and wifi adhoc)
b) Updating via odin & using frankenstein firmware
For now I'm planning to stick with the stock rom as performance and battery life is good, as is software compatability - no messing, most things work.
Though I am intrigued by the other kernels that are floating around especially with respect to power saving. Compatability/currency is critical - other features are secondary. This includes gpu acceleration and the likes all working properly.
I mostly leave some sync active along with push apps so ridiculous standby, but only with all comms down isn't interesting. I don't have a power issue though as can get 12-17 hours usage on stock.
I like the idea of BLN but worry it's going to hit power too much due to wakeups. Just maybe it might be handy if it can be figured for infrequent notification
Are there a particular subset of kernels worth looking at? Trying to cut down the problem space for all of the kernels.
Someone mentioned "siyah" as a good one?
I'm on KI8 now, probably moving to KJ2. root is a given?
Or maybe I'm too conservative - are there current, close to stock roms that I'm silly in ignoring?
planetf1 said:
Hi,
Does anyone have any kind of summary of the available kernel options for the SGS2?
I've been running mostly stock firmware, though
a) using the cf-root kernel (for titanium backup and wifi adhoc)
b) Updating via odin & using frankenstein firmware
For now I'm planning to stick with the stock rom as performance and battery life is good, as is software compatability - no messing, most things work.
Though I am intrigued by the other kernels that are floating around especially with respect to power saving. Compatability/currency is critical - other features are secondary. This includes gpu acceleration and the likes all working properly.
I mostly leave some sync active along with push apps so ridiculous standby, but only with all comms down isn't interesting. I don't have a power issue though as can get 12-17 hours usage on stock.
I like the idea of BLN but worry it's going to hit power too much due to wakeups. Just maybe it might be handy if it can be figured for infrequent notification
Are there a particular subset of kernels worth looking at? Trying to cut down the problem space for all of the kernels.
Someone mentioned "siyah" as a good one?
I'm on KI8 now, probably moving to KJ2. root is a given?
Or maybe I'm too conservative - are there current, close to stock roms that I'm silly in ignoring?
Click to expand...
Click to collapse
Not being funny mate but by the time you wait for someone to reply to you, you can actually take the time and search first in this forum to see if anyone else has answered your question....Loads of threads around to see which kernel is the best...Or even actually take the time to go into the Original dev section and read about each kernel that is available there....so next time before you open a thread please press the nice search button and read
anyway go for either 3 tops kernels ninp, speedmod or siyah..
The amount of info seemed overwhelming.
I ended up going with speedmod 2-12
Siyah was also on my shortlist
Sent from my GT-I9100 using Tapatalk
Hi there,
probably that request would make more sense in the developer forum but i didnt post enough yet
Actually i get insane battery batterylife with the bali 1.8.8 UV version, therefore i tried the non UV version with even more lowerd voltages via set cpu. it was stable but drained a lot faster. don´t know why.
so my request would be "pre-undervolted" bali kernels that might have even longer batteryliftime. Any hints on that? Someone who is able to compile something like this?
Best regards,
vibrantoparanto (from germany)
Hint' Bali X+Voltage control app= win !!!
Sent from my SGH-T959 using xda premium
^^ that is your answer....currently what im using
bali-x is what i used before - i lowered the voltages up to -150mv with some frequencies but it still gave me only about half the battery lifetime (less than 24 hours with moderate use) then i get with bali uv. What did you use to manually undervolt and how is your battery lifetime?
vibrantoparanto said:
bali-x is what i used before - i lowered the voltages up to -150mv with some frequencies but it still gave me only about half the battery lifetime (less than 24 hours with moderate use) then i get with bali uv. What did you use to manually undervolt and how is your battery lifetime?
Click to expand...
Click to collapse
i hate to burst your bubble, but youre chasing a ghost. uv'ing will not produce very noticeable changes to your battery life, especially if were are talking baout different versions of the same kernel.
take a look at this: http://forum.xda-developers.com/showpost.php?p=21024393&postcount=8
Cannot agree with that. I tried *a lot* and flashed different roms and kernels for years now. With stock kernels i get up to two days standby with low/moderate use, with bali UV i get almost three days. I have no idead why, but at least the standby time differs dramatically between kernels. At least thats for sure - under load that might be another thing. But probably you´re right and its just a "good standby behaviour" which would hardly benefit from lowered voltages. Would be still interesting to try a hardcore UV kernel
TopShelf, it is funny that you posted that note/post b/c I was just getting ready to PM you that I put part of it in my guide like I mentioned a week or so ago and the apps too.
Take a look at the Guide and maybe you will find some answers to your questions.
If you want to UV to the max, then you will have to test that yourself b/c every Vibrant reacts differently to OC/UV. Step it down -25v each day (don't set on boot until you know they work) and see where it freezes. There are OC/UV examples in the guide but beware, they are both for MIUI kernels.
Is there probably any admin out there who understands what i talk about and can move that thread to the developer forum?
this thread belongs here, you posted in the right section. posting in development will not get you answers from developers, it will only get you yelled at by hall monitors.
it is a common misconception that "developers" ONLY visit the "development" section. this is far from the truth.
sorry to say, the reason you have not gotten answers isnt because you posted in the wrong section, but because requests like these are pretty much never honored. the reason is because if a developer decides to make a slight alteration to a ROM/kernel for a member, he will then get BOMBARDED by dozens of other members who have tiny little tweaks that they would like in their ROM/kernel
Still trying to bring back some life to my request with this argument: I just checked my girlfriends device (using zendroid 2.1.2 with bali 1.8.8uv) and couldnt believe it: 3 days and 14 hours since unplugged with 8% left. The battery is a stock 1500mHa. And as its my girlfriends phone its definity not "overtweaked and calibrated". Personally i care much more about standby time than super heavy usage (guess that should be fine then too).
vibrantoparanto said:
Still trying to bring back some life to my request with this argument: I just checked my girlfriends device (using zendroid 2.1.2 with bali 1.8.8uv) and couldnt believe it: 3 days and 14 hours since unplugged with 8% left. The battery is a stock 1500mHa. And as its my girlfriends phone its definity not "overtweaked and calibrated". Personally i care much more about standby time than super heavy usage (guess that should be fine then too).
Click to expand...
Click to collapse
Since you didn't like the answer you got, for clarification purposes you are basically asking for a dev to use their time for FREE an modified/debug/test a kernel with your requirements just for your own sake ?
i wanted to point out that this is a very interesting kernel with great battery lifetime. if you have a look at all the other threads of people desperatly trying to get 24h of battery then you probably wouldnt consider this request as "wasting free time of others for my own sake".
vibrantoparanto said:
i wanted to point out that this is a very interesting kernel with great battery lifetime. if you have a look at all the other threads of people desperatly trying to get 24h of battery then you probably wouldnt consider this request as "wasting free time of others for my own sake".
Click to expand...
Click to collapse
i can prove to you that UVing does not make that much of a difference. if youre willing to learn, try this little experiment:
-UV your phone using your current settings
-put your phone in airplane mode
-turn it off
-charge it to 100% while its off
-after it boots up, leave it in airplane mode and immediately lock the screen
-let your phone sit idle in airplane mode from full charge to death (record the time)
--------------
Once youve done this, repeat all of the above steps but with STOCK VOLTAGE. record the time.
now compare the two times. i would be willing to bet a large sum of money that it is not a notable difference, if any.
my point is instead of trying to convince yourself that your theory of UVing is correct, why not do a TRUE battery life test, and simply find out for yourself. you cannot run true battery tests while you are actually using your phone becuase phone usage is massively different on a day to day basis, even if you dont think it is, it is. there is background data, text messages/calls, notifications, etc. (which is why i said to put it in airplane mode, that way its just sitting there, with little to no variables)
i guarantee you that you are putting way too much thought into voltage settings. they do not help as much as people want them to. dont say im wrong until you actually try what i said above. although youre probably nto going to.
vibrantoparanto said:
Still trying to bring back some life to my request with this argument: I just checked my girlfriends device (using zendroid 2.1.2 with bali 1.8.8uv) and couldnt believe it: 3 days and 14 hours since unplugged with 8% left. The battery is a stock 1500mHa. And as its my girlfriends phone its definity not "overtweaked and calibrated". Personally i care much more about standby time than super heavy usage (guess that should be fine then too).
Click to expand...
Click to collapse
sorry, but this does not prove anything. what are you comparing it to? your phone? you cant compare her battery life to your own unless you guys have the same exact setup, same exact apps, and same exact usage habits, etc.
you are not going about this battery things the right way. you NEED to be willing to learn more instead of trying to come up with your own reasons in your own head as to why your battery gets a specific lifespan. If getting THAT MUCH more battery life was as easy as UVing your phone, dont you think the manufacturers would incorporate that and use that as a marketing strategy?? "New from Samsung, the Galaxy S III!!! This will revolutionize the mobile industry, curing the battery woes across that smartphone world...now with lower voltage for drastically improved battery life!! Pick yours up today!!"
It's not that easy, you need to read up on what affects battery life.
thank you for your thoughts topshelf, but as i posted earlier i am on that topic for years now and i really tested a lot. not as scientific as running phone on idle with different frequenzies but trust me: I know a whole lot about this topic. Also i share your thought that the voltage is not that important in the end. On the other hand i have to say you guys are quite ignorant about my observations i did earlier:
1. The bali-x kernel with manual undervolting provides (provable) worse battery lifetime than
2. the 1.8.8uv kernel.
Conclusion: As you say it is not that much about the voltage but about the way the kernels work as a whole. On the other hand it would be *very* interesting if the insane battery lifetime of bali uv could be even further improved. And no i wont put my phone to sleep for days in order to do scientific test (whoever is willing to i will donate - battery lifetime almost like astrology)
Last but not least: It is true that different roms provide significantly different battery lifetimes - for whatever reason. Some roms turn your phone into a handwarmer, others give you almost the standby times you were used from the non-smarthpones. And i think even the big companies have a hard time to write device specific code that is highly optimized - i mean you still get new devices with gingerbread instead of ics and so on. conclusion: on the software side *is* much room for improvement if it comes to battery lifetime on most devices.
just my 2 cents
vibrantoparanto said:
i wanted to point out that this is a very interesting kernel with great battery lifetime. if you have a look at all the other threads of people desperatly trying to get 24h of battery then you probably wouldnt consider this request as "wasting free time of others for my own sake".
Click to expand...
Click to collapse
Look at your post, you and "others" are asking for some one to do research and development on their free time and for free ? little selfish on you.
Look at all the kernels thread in the development section and read thru all the comments, I bet there is far more people whining and complaining about this and that, than providing actually constructive and monetary feedback to the developers.
Its not as easy as it seams, ecotox already tried this before with his ultra low power kernels and some reported great succes while others couldnt even boot, it goes down to the quality of the silicon on each particular phone. no dev has the time/money to develop such request considering all the different variables involved.
lastly if you are so fond on this idea, why dont you take the initiative and learn to compile and patch and make your own kernel ?
Diff phone but a good simple guide to read:
http://arighi.blogspot.com/2011/08/howto-custom-kernel-on-samsung-galaxy-s.html
basic read:
http://stackoverflow.com/questions/5529470/how-to-compile-linux-kernel-for-real-android-phone
basic read:
http://igottadroid.com/wordpress/?p=61
For latest source:
http://opensource.samsung.com
"Be the change you want to see in the world." Mohandas Gandhi
Simple: If you want to change the world, start with yourself.
I'll read those links. Thanks for the info.
vibrantoparanto said:
And no i wont put my phone to sleep for days in order to do scientific test (whoever is willing to i will donate - battery lifetime almost like astrology)
Click to expand...
Click to collapse
im not trying to pick on you, but this sarcastic response shows your overall attitude that jrafael is trying to make you realize. you want somebody to take time out of their day/night to do something for you...yet you arent even willing to let for phone sit for a day in order to learn a thing or two...letting your phone sit doesnt even take any effort, compiling a kernel takes a lot more effort than that
i hate to say it, but why should someone help you, if you dont want to help yourself.
thank you jrafael for the links - thats a great read.
@topshelf: still think you overreact here. For one who has developed it and has the workflow it is a matter of a few minutes - and no one is forced to do anything unless he thinks its might be worth it - maybe its inspiring? i am just asking. not everything idea must be done by the one who thinks about it. and the quote about battery life is true i would say - we are really lacking valueable info on battery drain with different setups.
at first: this is not a "whats the best rom" thread.
i have 2 very focused questions:
What are the idle times with those different roms?
Are there some Main feature differences?
Aktually im running cm11 m5 snapshot with smartassv2 Governor. its smooth but the battery life is kinda poor.
Im looking for a rom with long battery life. And i dont care about google now, talking to my phone(ok google) nor i need the special features from the standart samsung rom.
Would be realy helpfull if u can review to my questions. Greetings!
pls apologize my bad english.
1 each phone is different depending on what's installed and running.
2 read there threads or flash each ROM and find out that way. You forgot about aokp?
andrewwright said:
1 each phone is different depending on what's installed and running.
2 read there threads or flash each ROM and find out that way. You forgot about aokp?
Click to expand...
Click to collapse
2. thx for informing abaout aokp!
1. im asking for someone who did a test with the same phone and all roms. i dont ask for different situations, i sked fpr idle time and features!
Probably there is somebody out there who testet it and can present his results
twinko said:
2. thx for informing abaout aokp!
1. im asking for someone who did a test with the same phone and all roms. i dont ask for different situations, i sked fpr idle time and features!
Probably there is somebody out there who testet it and can present his results
Click to expand...
Click to collapse
There are no definitive results, I have used both and I keep coming back to Omni due to the fact that on SlimKat music stutters like crazy when the screen is locked. Also, Omni updates are small enough to update often without having to download a complete update, to many users this a huge deal due to data limitations in certain countries. As for usage, idle time etc, like stated before, that depends entirely on what is installed on your device. Install a bunch of apps that need to constantly receive data updates and you will have battery drain and very little deep sleep. Where features are concerned, you kinda learn to live with limitations due to missing features on a specific rom. If you really want, you can most likely find an app that will bring a missing feature to a rom. I see a lot of people throwing their toys out the cot because one of the other does not have a specific feature implimented, yet the developers have done so with good reason, usually either performance or battery life.
With all said, there are many roms out there, Omni and SlimKat aren't the only players. It's difficult to compare roms nowadays because you are never quite sure if your kernel tweaking or custom kernel with it's hundreds of tweaks is the reason for a device not going into deep sleep either.
Give them all a try and decide for yourself, nandroid backups and apps like Titanium Backup makes it so easy.
cornelha said:
There are no definitive results, I have used both and I keep coming back to Omni due to the fact that on SlimKat music stutters like crazy when the screen is locked. Also, Omni updates are small enough to update often without having to download a complete update, to many users this a huge deal due to data limitations in certain countries. As for usage, idle time etc, like stated before, that depends entirely on what is installed on your device. Install a bunch of apps that need to constantly receive data updates and you will have battery drain and very little deep sleep. Where features are concerned, you kinda learn to live with limitations due to missing features on a specific rom. If you really want, you can most likely find an app that will bring a missing feature to a rom. I see a lot of people throwing their toys out the cot because one of the other does not have a specific feature implimented, yet the developers have done so with good reason, usually either performance or battery life.
With all said, there are many roms out there, Omni and SlimKat aren't the only players. It's difficult to compare roms nowadays because you are never quite sure if your kernel tweaking or custom kernel with it's hundreds of tweaks is the reason for a device not going into deep sleep either.
Give them all a try and decide for yourself, nandroid backups and apps like Titanium Backup makes it so easy.
Click to expand...
Click to collapse
Thx for your long, interessting and helping text!
im missing only 1 thing. if someone flashes all roms, doesnt install any apps. Only new rom and full wipe before, restart and tracks the time the phone has battery. Dont u think this is a possibility to compare the roms battery usage? i think there is no one who does this, but i probably will try this and put down a little table here. so people like me are able to see the rom with the best battery efficiency. so if they decide this rom has not enough performance, they can switch to "battery efficiency Number 2" and so on.
greetings
twinko said:
Thx for your long, interessting and helping text!
im missing only 1 thing. if someone flashes all roms, doesnt install any apps. Only new rom and full wipe before, restart and tracks the time the phone has battery. Dont u think this is a possibility to compare the roms battery usage? i think there is no one who does this, but i probably will try this and put down a little table here. so people like me are able to see the rom with the best battery efficiency. so if they decide this rom has not enough performance, they can switch to "battery efficiency Number 2" and so on.
greetings
Click to expand...
Click to collapse
Sure you can do what you stated but bear in mind how will you compare the battery life across roms..
There should be some standardized test... Like looping video playback or some other method which does not change based on you usage habits..
Sent from my GT-I9300 using Tapatalk
Pa is another you're missing from you're list of developed roms. What you could do is test video playback, calls incoming/outgoing,one game of choice "high end battery draining game" and compere them all. As for the kernel tuning.. Read about what really happens and you will fine not a lot does happen with UV/uc .
twinko said:
Thx for your long, interessting and helping text!
im missing only 1 thing. if someone flashes all roms, doesnt install any apps. Only new rom and full wipe before, restart and tracks the time the phone has battery. Dont u think this is a possibility to compare the roms battery usage? i think there is no one who does this, but i probably will try this and put down a little table here. so people like me are able to see the rom with the best battery efficiency. so if they decide this rom has not enough performance, they can switch to "battery efficiency Number 2" and so on.
greetings
Click to expand...
Click to collapse
You can compare but you have to use that phone for more days not like just one and you have to full charge and then discharge 2 or 3 times till the phone make the correct battery stats. So i think you can test but better with daily use or get an app that drains battey faster well i think this may work