I have on my Folio found a way to tweak the nvrm_daemon, to increase the frame draw rate.. on folio its was limited to 51 frames (tested with app: fps2d) and this tweak increases it to 78 and it shows in the fps2d app too.
Just so everybody is clear on what this hack does.. its changing the framebuffer rate, NOT screen refreshrate.
see this link if you are not sure what that means
testing this with NeoCore benchmark we hit 56 frames.. so im guessing Viewsonic Gtab also could benefit from it, but as i don't have one i cannot verify if the screen type works with this change..
Anyone with just a small linux knowledge and knows how to replace a binary, please verify this , and you might have another enhancement if it works.
download the attached zip and put the nvrm_daemon into the /system/bin and do a chmod 755 /system/bin/nvrm_daemon on the file too.. (remember /system need to be r/w)
reboot and test with "fps2d" from Market, and let me know if you gain 78 frames too ... (i hope so)
it should work on all custom roms, as im sure, you all use the same Gtab binaries for your hardware here..
The updated nvrm_daemon ran successfully TnT Lite 4.1.0
Stats:
Done avg=79 stdev=6.47
1000 iterations: touch to repeat.
Butch1326 said:
The updated nvrm_daemon ran successfully TnT Lite 4.1.0
Stats:
Done avg=79 stdev=6.47
1000 iterations: touch to repeat.
Click to expand...
Click to collapse
so the questions is, what was the previous lock rate on Gtab?
Dexter_nlb said:
so the questions is, what was the previous lock rate on Gtab?
Click to expand...
Click to collapse
Done avg=51 stdev=3.59
Pershoots 0112 kernel on zpad 3 rom G Tablet
Dexter_nlb said:
so the questions is, what was the previous lock rate on Gtab?
Click to expand...
Click to collapse
I put the original nvrm_daemon file back in a rebooted the gtab. Here's the results I got:
Done avg=51 stdev=4.32
1000 interations: touch to repeat
Butch1326 said:
I put the original nvrm_daemon file back in a rebooted the gtab. Here's the results I got:
Done avg=51 stdev=4.32
1000 interations: touch to repeat
Click to expand...
Click to collapse
if you're willing to test, maybe use it for a day or 2, and let others know if its stable with the change..
its primarily a 2D and 3D games improvement, for regular use i doubt you see differences..
In GLBenchmark its not visible (probably best 3d engine testing tool), but other benchmarks does indeed increase, as they probably is not pushing CPU limits, where GLB does...
Dexter_nlb said:
if you're willing to test, maybe use it for a day or 2, and let others know if its stable with the change..
its primarily a 2D and 3D games improvement, for regular use i doubt you see differences..
In GLBenchmark its not visible (probably best 3d engine testing tool), but other benchmarks does indeed increase, as they probably is not pushing CPU limits, where GLB does...
Click to expand...
Click to collapse
I will report back with an update over the weekend. I have Raging Thunder2, and I'm addicted to Angry Birds, so I'll get some game time in...
Thanks!!!
Butch1326 said:
and I'm addicted to Angry Birds
Click to expand...
Click to collapse
aren't we all
getting 55fps on neocore used to be 33 or so.
Thanks a bunch, will test this out in Dungeon Defenders.
FYI:
chown 755 /system/bin/nvrm_daemon
should be
chmod 755 /system/bin/nvrm_daemon
you could also:
chown root:root /system/bin/nvrm_daemon
rothnic said:
should be
chmod 755 /system/bin/nvrm_daemon
Click to expand...
Click to collapse
was too quick on the keyboard trigger
nice find... what exactly was holding our devices back? 50Hz refresh and vsync?
Sent from my VEGAn-TAB-v1.0.0B5.1 using Tapatalk
TheTruth said:
nice find... what exactly was holding our devices back? 50Hz refresh and vsync?
Click to expand...
Click to collapse
honestly, i dont know.. i was actually trying to fix the resolution to 840x480 which i accomplished so i could compare performance with a 7" elocity. and get same score, but i then experimented with some other values and rebooted a few times watching the bootup and i noticed the init of the framebuffer device changed,
NvGrFbDevOpen: Opened fb0 on disp 0 size 1024x600 format 1 rate 78.000000
and i tested it with fps2d and got info confirmed...
so actually it was pure luck
not sure of the effects yet. maybe a battery drainer ? but the hack its only a few days old, and kernel dist does not include nvrm_daemon, so im not sure where to find exact documentation on it yet.
The greatest advances seem to be pure blind luck of unintended consequences. Nice catch on your part.
Thanks Dexter. Question. Can I also use this on my Elocity that at the moment is running your 1.0c.
Thanks in advance
manfly9884 said:
Thanks Dexter. Question. Can I also use this on my Elocity that at the moment is running your 1.0c.
Click to expand...
Click to collapse
no, this driver has 1024x600 native resolution which your device will not handle.
But i can take a look at it tomorrow to see if its the same with ElocityA7 .
There is alot of tweaks to offer for ElocityA7, but sadly i still havent got one :-( , to verify it.
Dexter_nlb said:
honestly, i dont know.. i was actually trying to fix the resolution to 840x480 which i accomplished so i could compare performance with a 7" elocity. and get same score, but i then experimented with some other values and rebooted a few times watching the bootup and i noticed the init of the framebuffer device changed,
NvGrFbDevOpen: Opened fb0 on disp 0 size 1024x600 format 1 rate 78.000000
and i tested it with fps2d and got info confirmed...
so actually it was pure luck
not sure of the effects yet. maybe a battery drainer ? but the hack its only a few days old, and kernel dist does not include nvrm_daemon, so im not sure where to find exact documentation on it yet.
Click to expand...
Click to collapse
ahhh that explains alot... it seems the refresh is set to 78... thats why my evo when set to 50 or 60 Hz antibanding mode didnt do squat to remove the bands zipping by on the screen...
roebeet was very interested in the refresh rate of this device... perhaps you could talk to him about this tweak
Sent from my VEGAn-TAB-v1.0.0B5.1 using Tapatalk
TheTruth said:
roebeet was very interested in the refresh rate of this device... perhaps you could talk to him about this tweak
Click to expand...
Click to collapse
ill PM him about this as well..
. Works great on g tab. 78fps on fps2d and 54.1 fps on neocore. I'm sure quadrant will jump to. Thanks!
Update: Quadrant stayed the same which is odd as I could see that fps was much higher in 3d tests, like hitting 80 fps. Sweet!
nswenson said:
.
Update: Quadrant stayed the same which is odd as I could see that fps was much higher in 3d tests, like hitting 80 fps. Sweet!
Click to expand...
Click to collapse
your 3D score might jump, but if the rest drops just a little , you don't notice it much, it weighs not much in the complete test. I/O really makes a difference when tweaked..
But i got just another 200+ here, running it a few times, so its not a major but it did show improvement on 3D.. (a little strange as the fps2D also shows 2D increase?) but thats probably why perf. tests are not always reliable..
i tried some flashplayer tests on top of this, and those also increased
Related
What is more of a priority to you: Uncapped framerates or HDMI-out ?
Where's option 3 with automatic toggling if an hdmi cable is connected?! I keed I keed
netarchy said:
What is more of a priority to you: Uncapped framerates or HDMI-out ?
Click to expand...
Click to collapse
Have the devs found the actual cap? Not some dirty LCD hack.
jerryparid said:
Have the devs found the actual cap? Not some dirty LCD hack.
Click to expand...
Click to collapse
The main cap is probably being set in the radio, which leaves us with having to work around it to uncap framerates. Ideally if it's possible to get both uncapped framerates and hdmi-out working at the same time that's the solution that will be used.
I vote framerate for now... but hopefully there is some long term solution to both
I'll take framerate. HDMI would put up more of a fight if we could get full output of the android system instead of just selective video.
I still don't understand why the cap is a big deal. What can't you do that requires double the framerate?
I'd go with the fps because the hdmi is still restricted with output options anyway
acme64 said:
I still don't understand why the cap is a big deal. What can't you do that requires double the framerate?
Click to expand...
Click to collapse
Gaming without annoying lag and stuttering. Smooth transitions between home screens. Better live wallpaper performance. Feel as smooth overall as the HTC Aria
acme64 said:
I still don't understand why the cap is a big deal. What can't you do that requires double the framerate?
Click to expand...
Click to collapse
the device being capped at 30 fps wouldn't be a problem if the device wouldn't drop below 30fps but it often does during 3d games (the human eye typically perceives anything over 19 fps as fluid motion, without choppiness or studdering). so the main reason to uncap the framerate is because it gives you more wiggle room for the fps to drop without you noticing choppiness.
so if you were capped at 30 fps and a game caused a performance drop of 12 fps then you would now be at 18fps and you will notice choppiness,
but if the fps were unlocked to 44 (which is the novatek right now i think) and that same game dropped the 12 fps, you would now be running at 32 fps and you will notice no choppiness.
it can be hard for someone to understand sometimes as most people dont properly explain WHY having higher fps is nice instead of just sounding like iphone users "its so smooth!" like the guy above me lol.
I vote frame rate. Less than 1% of the time would be spent using HDMI for me anyway.
Didn't someone post this already ?
Sent from my PC36100 using XDA App
I couldn't care less about HDMI, disable it permanently to uncap the frame rate for all I care.
HDBaseT>hdmi
http://www.pcworld.com/article/200301/new_technology_could_replace_hdmi_with_ethernet_cables.html
Sent from my PC36100 using XDA App
What about just maintaining 2 different kernels and if u wanna use hdmi you could easily flash one kernel then flash right back to the fps fixed kernel when your finished seems like the most logical way to go
GHOST99K said:
Didn't someone post this already ?
Sent from my PC36100 using XDA App
Click to expand...
Click to collapse
Considering the OP is spear heading development on a dual Nova/Epson kernel, I think it is okay for the question to be asked on a very real basis...
the poll was our idea. we needed to see who wanted what. it came from a discussion we were having on #HTC-EVO (<~shameless plug) about the issues were facing and what to do about the community kernel. we figured the real answer would come from... well... the community. anyhow they both need to be fixed for there to actually be a fix. imo a broke isnt fixed so that was the discussion. anyhow hope that settles ths little debate.
Unless a future work-around includes being able to use the HDMI output for use with the actual phone interface and not just videos, I see no real point of using it at all.
So, definitely uncapped.
My only complaint about the Evo has been the laggy touch response. If uncapping the framerate fixes that, which the op's kernel that I flashed today has on my phone, I am willing to do without the hdmi. While I would prefer to have everything working, getting rid of the touch lag is the highest priority to me.
wfrandy said:
My only complaint about the Evo has been the laggy touch response. If uncapping the framerate fixes that, which the op's kernel that I flashed today has on my phone, I am willing to do without the hdmi. While I would prefer to have everything working, getting rid of the touch lag is the highest priority to me.
Click to expand...
Click to collapse
Using the 'dirty' fix will pretty much fix the input, go check out the 3.6.3 fix, it will almost remove all lag
Yes, this is my first ever post here! but the fact is, that I never saw the need to post and Q/A because most of the stuff that came to mind were asked and answered. Until now...
First off, I am a happy Transformer user and never had reasons to complain. But this kinda upset me..
I had a chance to play with a buddy of mines Galaxy Tab 10.1 yesterday for the first time, and I noticed something...scrolling and just moving around the tablet felt smoother than my Asus (Mind you, I have a custom rom installed, OC kernel with many optimization's)
So I figured, since his was bone-stock, and it pretty much being the same spec hardware (Tegra 2 etc..) I would run some benchmarks.
As soon as I ran Quadrant I noticed his was smoother...but, my scores were HIGHER!
Then i noticed on the bottom left hand side FPS were locked at 30FPS?!?! on mine!! and his were going well over 30!! actually maxing out at 60!!!
So what gives!?!?!
Are our Asus Refresh rates stuck at 30??? If that is indeed the case that would suck! BIGTIME!
Hopefully somebody can shed some light
THANKS GUYZ
aurena said:
Yes, this is my first ever post here! but the fact is, that I never saw the need to post and Q/A because most of the stuff that came to mind were asked and answered. Until now...
First off, I am a happy Transformer user and never had reasons to complain. But this kinda upset me..
I had a chance to play with a buddy of mines Galaxy Tab 10.1 yesterday for the first time, and I noticed something...scrolling and just moving around the tablet felt smoother than my Asus (Mind you, I have a custom rom installed, OC kernel with many optimization's)
So I figured, since his was bone-stock, and it pretty much being the same spec hardware (Tegra 2 etc..) I would run some benchmarks.
As soon as I ran Quadrant I noticed his was smoother...but, my scores were HIGHER!
Then i noticed on the bottom left hand side FPS were locked at 30FPS?!?! on mine!! and his were going well over 30!! actually maxing out at 60!!!
So what gives!?!?!
Are our Asus Refresh rates stuck at 30??? If that is indeed the case that would suck! BIGTIME!
Hopefully somebody can shed some light
THANKS GUYZ
Click to expand...
Click to collapse
Download this off market and u should get atleast 58 frames per second
https://market.android.com/details?id=com.edburnette.fps2d
Unlocking the frames would be an awsome thing in a custom kernel but I don't believe it can be done. However ill ask again
so could it be that 2d is able to run at 60hz and 3d cannot? This still doesnt quite answer the general smoothness advantage of the tab.
Im going to try to get my hands on his tab and do a side by side comparison
Btw thx for the quick reply
aurena said:
so could it be that 2d is able to run at 60hz and 3d cannot? This still doesnt quite answer the general smoothness advantage of the tab.
Im going to try to get my hands on his tab and do a side by side comparison
Btw thx for the quick reply
Click to expand...
Click to collapse
Np that just tells u the fps as a whole I believe. But a video of a side by side would be ultra helpfull
I got an average of 40fps...
St8 from the cows teet
clemsyn said:
Just an update on JRCU implementation in the kernel. It seems to give us highes scores in 3d benchmarks in antutu and seems more responsive BUT the dreaded sleep of death and intermittent reboots have made me decide to drop it. It seems to be related to wifi, if it is off, there are no reboots or sleep of death but with it on and then turning the tab off, then reboots or sleep of deaths happens.
Click to expand...
Click to collapse
In short its possible but then it would fk stuff up
Please note that it's a known fact that Tegra 2 based tablets have a worse graphics performance when running in portrait mode. Did you run these tests in portrait or in landscape mode?
Also, I've not seen this other fact reported anywhere, but I have the dock, and I have noticed that scrolling the home pages or any other item via the touchpad does produce a way lower refresh rate than using the touch screen. Although it would be very silly if true, I think the problem is related to the touchpad mouse cursor sprite, which seems to be implemented by overimposing a full screen transparent layer on top of the normal flipping animation, thus affecting performance greatly.
I would really like this being addressed by ASUS our Google, since I use my Transformer way more in NetBook than in tablet mode...
With virgin stock HC 3.2, my TF101 scored 62.50 fps in Nenamark1 3D benchmark. The previous record was 44.40 fps... The trick is to run the benchmark in 'Zoom to fill mode' Now my device is the 3rd fastest TF101 on the net .
Ref :
http://nena.se/nenamark/view?version=1&device_id=424
cowballz69 said:
Download this off market and u should get atleast 58 frames per second
https://market.android.com/details?id=com.edburnette.fps2d
Unlocking the frames would be an awsome thing in a custom kernel but I don't believe it can be done. However ill ask again
Click to expand...
Click to collapse
Why would this be awesome? Running anything higher than the native refresh rate of the screen doesn't do anything except waste CPU and GPU cycles.
I am just guessing here but the native refresh rate is probably 59 or 60 Hz.
Just installed this apk from android police and type a search for 1440P videos. Go in the to settings for the video and switch the resolution and you'll now see an option for 1440P. Watched a video on it and it looked very detailed and crisp. Crystal clear. Enjoy!
Article below
http://www.androidpolice.com/2014/0...roid-delivers-number-playlist-related-tweaks/
Download the apk below:
https://www.androidfilehost.com/?w=...731847fd95a6f5463e9fa2bfd8e492e0ec58d86d98799
Sent from my Nexus 7 using XDA Free mobile app
Also here is a beautiful video playing in 1440P. The detail is incredible. I appreciate this screen more and more. Just compared the video on my wife's iPhone screen compared to the g3 and it really blows it away.
4K Demo UHD 1440p:
http://youtu.be/11zf7Y5tUbM
Sent from my LG-D851 using XDA Free mobile app
bustinstugots24 said:
Also here is a beautiful video playing in 1440P. The detail is incredible. I appreciate this screen more and more. Just compared the video on my wife's iPhone screen compared to the g3 and it really blows it away.
4K Demo UHD 1440p:
http://youtu.be/11zf7Y5tUbM
Sent from my LG-D851 using XDA Free mobile app
Click to expand...
Click to collapse
Not sure why but I can't get the 1440p option to show up on my G3
shook187 said:
Not sure why but I can't get the 1440p option to show up on my G3
Click to expand...
Click to collapse
You need to make sure that the youtube app version is on 5.9 if it is. On a video that supports 1440p/2k you need to click on the 3 dots top right of the video and then click on the cog. Should allow you to change it to 1440p
With the display and the sound on the G3 I could actually imagine watching a movie on it - which I've never bothered to do on any mobile before. It's absolutely GREAT!
no need to double post ...
http://forum.xda-developers.com/lg-g3/themes-apps/youtube-app-support-2k-video-t2841657
thank you, this basically confirmed my suspicions there is zero different between 1440 and 1080 on this screen size at least between the G2 and G3 as I played both next to each other today and out of the 5 people in my house I asked, none of them could tell a different including me. I think im going back to 1080p with 540dpi as I didn't notice any difference between that or the 1440p w/ default DPI. Battery life was just improved enough that I liked it. Thanks for posting the new apk!
Koroz said:
thank you, this basically confirmed my suspicions there is zero different between 1440 and 1080 on this screen size at least between the G2 and G3 as I played both next to each other today and out of the 5 people in my house I asked, none of them could tell a different including me. I think im going back to 1080p with 540dpi as I didn't notice any difference between that or the 1440p w/ default DPI. Battery life was just improved enough that I liked it. Thanks for posting the new apk!
Click to expand...
Click to collapse
For changing the dpi i use texdroider_dpi, what do u use for changing the resolution?l
qum said:
For changing the dpi i use texdroider_dpi, what do u use for changing the resolution?l
Click to expand...
Click to collapse
NOMone
Koroz said:
NOMone
Click to expand...
Click to collapse
Well i tried and seems I'm doing something wrong. Can u pls tell me each setting u set?
qum said:
Well i tried and seems I'm doing something wrong. Can u pls tell me each setting u set?
Click to expand...
Click to collapse
Width = 1080
Height = 1920
DPI = 530+ (anything under has been reported to cause problems. I like 540)
It's 4:45 am here so I am heading to bed after spending a night drinking so if you need more help and I don't respond its because I passed out and ill respond tomorrow when I come back to reality.
I
Koroz said:
thank you, this basically confirmed my suspicions there is zero different between 1440 and 1080 on this screen size at least between the G2 and G3 as I played both next to each other today and out of the 5 people in my house I asked, none of them could tell a different including me. I think im going back to 1080p with 540dpi as I didn't notice any difference between that or the 1440p w/ default DPI. Battery life was just improved enough that I liked it. Thanks for posting the new apk!
Click to expand...
Click to collapse
qum said:
For changing the dpi i use texdroider_dpi, what do u use for changing the resolution?l
Click to expand...
Click to collapse
I was tempted to try the dpi modification but was a bit hesitant to. I was under the assumption most apps don't work properly and widgets on the home screen get cut off and just don't appear to look right. Maybe I'll try it. What do you guys recommend? Does it really make THAT much of a battery difference? How much more SOT do you get by making this modification?
Sent from my LG-D851 using XDA Free mobile app
bustinstugots24 said:
I
I was tempted to try the dpi modification but was a bit hesitant to. I was under the assumption most apps don't work properly and widgets on the home screen get cut off and just don't appear to look right. Maybe I'll try it. What do you guys recommend? Does it really make THAT much of a battery difference? How much more SOT do you get by making this modification?
Sent from my LG-D851 using XDA Free mobile app
Click to expand...
Click to collapse
Sorry for the late response was in a bourbon coma.
So as long as you don't go down below 530, you *SHOULD* have zero problems. I ran mine at 540, but I left it at 530 first to see as that was what most say is the "breaking" point for most apps while keeping them on screen. ALL "LG" themed apps I tried worked fine, contacts, phone, messaging etc. I run Buzz Launcher w/ Zooper, and the only issue I had was with Naxos Flat "Circle Clock w/ Weather and batt stats showing the circle background wrong. I am sure if I had removed this widget, then reconfigured it most likely would have worked.
As for the battery test, I dunno I "felt" the difference but since I didn't really use any formal testing before I sent back my sprint version I don't want to tell you it does or doesn't. It "appears" to me it makes a big difference as I am usually recharging after a single movie streaming over LAN w/ ES File Manager played with Archos Video Player. Now I can watch a couple of movies. That doesn't really satisfy my "scientific" mind and it shouldn't yours either. ;^)
I think though what I will do is just run a test tonight when I get home, then tomorrow put the phone back to 2k and default dpi and re-run those same exact tests. I'll take advice as I have never "tested" battery consumption on a smartphone before. I figured I would just do a few different tests of playing movies. I don't "game" on my phone but if you have an opinion on a good one to test with I could run that too. Any experience with battery tester from Antutu, I am going to check that out also.
Koroz said:
Sorry for the late response was in a bourbon coma.
So as long as you don't go down below 530, you *SHOULD* have zero problems. I ran mine at 540, but I left it at 530 first to see as that was what most say is the "breaking" point for most apps while keeping them on screen. ALL "LG" themed apps I tried worked fine, contacts, phone, messaging etc. I run Buzz Launcher w/ Zooper, and the only issue I had was with Naxos Flat "Circle Clock w/ Weather and batt stats showing the circle background wrong. I am sure if I had removed this widget, then reconfigured it most likely would have worked.
As for the battery test, I dunno I "felt" the difference but since I didn't really use any formal testing before I sent back my sprint version I don't want to tell you it does or doesn't. It "appears" to me it makes a big difference as I am usually recharging after a single movie streaming over LAN w/ ES File Manager played with Archos Video Player. Now I can watch a couple of movies. That doesn't really satisfy my "scientific" mind and it shouldn't yours either. ;^)
I think though what I will do is just run a test tonight when I get home, then tomorrow put the phone back to 2k and default dpi and re-run those same exact tests. I'll take advice as I have never "tested" battery consumption on a smartphone before. I figured I would just do a few different tests of playing movies. I don't "game" on my phone but if you have an opinion on a good one to test with I could run that too. Any experience with battery tester from Antutu, I am going to check that out also.
Click to expand...
Click to collapse
You won't too much of a battery difference because the display is still pushing that many pixels no matter what resolution you set. You'll likely see some performance improvements but that's probably it.
---
I just tried watching a few gaming vids on 1440p and its absolutely gorgeous!
4ndroid99 said:
You won't too much of a battery difference because the display is still pushing that many pixels no matter what resolution you set. You'll likely see some performance improvements but that's probably it.
---
I just tried watching a few gaming vids on 1440p and its absolutely gorgeous!
Click to expand...
Click to collapse
Can you cite a source please I want to read it and see how they explain that, because this statement makes zero sense to me. Regardless if the panel is still pushing the native amount of pixels the amount of energy(power) needed to power the panel lowers. You can view this by the heat generated on your CPU/GPU using different resolutions as the device may still use the default panel the smaller "image" size even with required scaling will still use less power then a full size image at native. Same goes with CPU.
Koroz said:
Can you cite a source please I want to read it and see how they explain that, because this statement makes zero sense to me. Regardless if the panel is still pushing the native amount of pixels the amount of energy(power) needed to power the panel lowers. You can view this by the heat generated on your CPU/GPU using different resolutions as the device may still use the default panel the smaller "image" size even with required scaling will still use less power then a full size image at native. Same goes with CPU.
Click to expand...
Click to collapse
I'm not too sure either. I've been viewing threads on this for quite a while now and that's the general consensus. You can find quite a few threads here on XDA with ppl discussing this.
http://forum.xda-developers.com/lg-g3/general/amazing-battery-life-tweaks-t2806078
4ndroid99 said:
I'm not too sure either. I've been viewing threads on this for quite a while now and that's the general consensus. You can find quite a few threads here on XDA with ppl discussing this.
http://forum.xda-developers.com/lg-g3/general/amazing-battery-life-tweaks-t2806078
Click to expand...
Click to collapse
I've read them none of them cite sources either, its all people talking conjecture. I'll do some actual testing and post results here. For example a lot of people are saying "From my past experience", that usually equals "I haven't tested it, but take my word for it". Maybe the results will surprise me and show there is minimal battery life change, but I doubt it
EDIT: So I played around with the AnTuTu test suite's battery tester. Seems legit enough that it should give us a good baseline set of results to compare. I put my phone back to default right now and I am charging it to 100%, then Ill run the test. Should take a couple hours at least. Then I'll switch back to 1080/540dpi and do the same battery test with a full charge again and see what the results are. Probably won't be done until tomorrow morning my time but if they get done sooner (Pacific timezone) ill post them up here.
I'm half tempted to restore factory to do this testing to make it even more a set of baseline results that most VZW users can guess they will have. Right now though im running Greenify Paid w/ Xposed plugins and other things etc. Really wish we had unlocked bootloader so I could just back my stuff up but alas, I find QA/Testing fun so I dunno, I might, might not reset.
Koroz said:
I've read them none of them cite sources either, its all people talking conjecture. I'll do some actual testing and post results here. For example a lot of people are saying "From my past experience", that usually equals "I haven't tested it, but take my word for it". Maybe the results will surprise me and show there is minimal battery life change, but I doubt it
EDIT: So I played around with the AnTuTu test suite's battery tester. Seems legit enough that it should give us a good baseline set of results to compare. I put my phone back to default right now and I am charging it to 100%, then Ill run the test. Should take a couple hours at least. Then I'll switch back to 1080/540dpi and do the same battery test with a full charge again and see what the results are. Probably won't be done until tomorrow morning my time but if they get done sooner (Pacific timezone) ill post them up here.
I'm half tempted to restore factory to do this testing to make it even more a set of baseline results that most VZW users can guess they will have. Right now though im running Greenify Paid w/ Xposed plugins and other things etc. Really wish we had unlocked bootloader so I could just back my stuff up but alas, I find QA/Testing fun so I dunno, I might, might not reset.
Click to expand...
Click to collapse
Ok, I await your results. It'd be interesting to see.
ran the update. showing youtube 5.9.0.10. searched for 1440p videos and still only showing 720 option. Anyone know whats up?
rollerpig said:
ran the update. showing youtube 5.9.0.10. searched for 1440p videos and still only showing 720 option. Anyone know whats up?
Click to expand...
Click to collapse
try clearing data and cache on youtube app, then re-installing. mine works fine.
Hi, under product/etc/hwpg there are files for powersaving modes.
If you edit it, you can test your mate9 world first 1ghz gpu for mobile pgones with your favorite games(emulators)
Modify normal cpu policy, or super cpu policy.xml (for powersaving settings) and reboot
Under is example of usage: app added to modified normal_cpu_policy.xml strings is -package name com.dolphinemu.dolphinemu
I found that the max gpu frequency is normally not recheable by phone due to this .xml strings. Which blocking it to 960000000hz
<scenemode name=".*?(hw.performance|dolphinemu*?).*" id="10000">
<cpu_minprofile>1805000</cpu_minprofile>
<cpu_a15_minprofile>2304000</cpu_a15_minprofile>
<gpu_minprofile>1037000000</gpu_minprofile>
<ddr_minprofile>960000000</ddr_minprofile>
<threshold_up>300</threshold_up>
<threshold_down>150</threshold_down>
</scenemode>
Warning, setting max gpu and loading heavy tasks might harm your device.
how long it takes to overheat ?
WieserZ1C said:
how long it takes to overheat ?
Click to expand...
Click to collapse
If my readings of gpu temp sensor ar true, it is about 60°c in 10 min of course cpu usage will also affect this value
Would love to try this but you know warranty and root don't go well together. Anyway how did you managed to run dolphin and what games, i can't seem to get playable fps.
arminbih said:
Would love to try this but you know warranty and root don't go well together. Anyway how did you managed to run dolphin and what games, i can't seem to get playable fps.
Click to expand...
Click to collapse
Zelda wind waker native resolution constant 30fps, even on large arenas
2x resolution sometime little lag on large areas
Fifa 06 2x res. playable after set override cpu to 34
Native res and cpu override to 34 near constant 35fps sometimes sound lagg
My settings: place my ini file to dolphin-emu/config
I just got it to work thx alot. Yaaaaaaay
id suspect over-voltage / temperature protection are in play so i dont see any negative side affects in doing this
virtyx said:
id suspect over-voltage / temperature protection are in play so i dont see any negative side affects in doing this
Click to expand...
Click to collapse
Thats true, if temp gets too high phone just shut down itself. My mate9 never go off after high cpu/gpu usage like mate7 with this mod and its bit colder than mt7
I suppose that this have effect on Android games also, right?
If the GPU throttles due to heat after a gaming session as is, I dread thinking how horrible it will be, when you overclock it.
So please tell me, could I use this to underclock the GPU of this phone? Cuz I play game and it's too hot for me but it's not a heavy game, just a casual game. If yes, please tell me how. Thanks and have a nice day!
lemonade747 said:
So please tell me, could I use this to underclock the GPU of this phone? Cuz I play game and it's too hot for me but it's not a heavy game, just a casual game. If yes, please tell me how. Thanks and have a nice day!
Click to expand...
Click to collapse
Try paste this txt to product/etc/hwpgnormal_cpu_policy or super_cpu_policy this add new profile to your applications
Sample com.yourappname.secondname for another game com.game.game2
<scenemode name=".*?(yourappname|game2*?).*" id="10020">
<cpu_maxprofile>1402000</cpu_maxprofile>
<cpu_a15_maxprofile>1805000</cpu_a15_maxprofile>
<gpu_maxprofile>533000000</gpu_maxprofile>
<threshold_up>300</threshold_up>
<threshold_down>150</threshold_down>
</scenemode>
This will underclock your gpu by half and cpu little bit
gahu777 said:
Try paste this txt to product/etc/hwpgnormal_cpu_policy or super_cpu_policy this add new profile to your applications
Sample com.yourappname.secondname for another game com.game.game2
<scenemode name=".*?(yourappname|game2*?).*" id="10020">
<cpu_maxprofile>1402000</cpu_maxprofile>
<cpu_a15_maxprofile>1805000</cpu_a15_maxprofile>
<gpu_maxprofile>533000000</gpu_maxprofile>
<threshold_up>300</threshold_up>
<threshold_down>150</threshold_down>
</scenemode>
This will underclock your gpu by half and cpu little bit
Click to expand...
Click to collapse
Thanks! I will try this weeken and report soon! :fingers-crossed:
P/S: excuse me but what does "id" number mean? I have to write exactly like that or I could find somewhere? And regarding the app name, I have only one game, the format like this com.myappname.secondname so I write this
<scenemode name=".*?(myappname*?).*" id="10020">
Is it correct? And btw could you please give me the name of the app you're using to show cpu, gpu on the screen? I downloaded a lot of app like that but nothing shows my gpu info, only cpu or I have to run sth heavy then gpu should work?
New id is required for new group off apps, but you can also edit existing id like 10015
Method with new id can be aplied to just one app
The app for monitoring is cool tool
gahu777 said:
New id is required for new group off apps, but you can also edit existing id like 10015
Method with new id can be aplied to just one app
The app for monitoring is cool tool
Click to expand...
Click to collapse
I tried many times but it's not working, gpu doesnt show in any app for me, and cpu doesnt change anything, use app Cool tool as you said but only show like this: CPU:xx%, xxxx mhz and M: xxx mb where x are numbers. It didnt show multi core cpu, didnt show gpu and after I added the new profile, I rebooted then cpu still goes up to 1800mhz. So I think it's not working right?
P/S: my game name is: en.co.atm.unison, just that, it doesnt have any "com" or second name like others.
Check out wind waker wideo with 2x res and gpu set to 800mhz-1ghz, playing it about hour, no slowdowns, this phone is a beast!
https://youtu.be/k2CFkDjGbco
gahu777 said:
Check out wind waker wideo with 2x res and gpu set to 800mhz-1ghz, playing it about hour, no slowdowns, this phone is a beast!
Click to expand...
Click to collapse
Very very cool
I can't found this on my Mate.
albertobom said:
I can't found this on my Mate.
Click to expand...
Click to collapse
The path to profile files is incorrect in my previous post. /product/etc/hwpg/normal_cpu_policy.xml or super_cpu_policy.xml
gahu777 said:
The path to profile files is incorrect in my previous post. /product/etc/hwpg/normal_cpu_policy.xml or super_cpu_policy.xml
Click to expand...
Click to collapse
Bro can you share what app you use to monitor gpu frequency?
Edit: found it but don't know how to use it.
I see that samsungs and v30 got the option to change screen resolution to save battery. Is that option in the G6? (I currently got an LG G2 and want to upgrade)
No all we got is the option to change the acpect ratios for the apps
Sent from my LG-H873 using Tapatalk
I wouls like to know if its possible to do so somehow.....for gaming my phone overheats and starts throttling.... If i would run 1080p it should run smooter..
I am not rooted
Hunteres said:
I wouls like to know if its possible to do so somehow.....for gaming my phone overheats and starts throttling.... If i would run 1080p it should run smooter..
I am not rooted
Click to expand...
Click to collapse
That's exactly why Im asking. Its a shame if you cant tune down the resolution and loose performance. Perhaps a custom rom may steal that function from the v30 now that is released.
My LG G2 never get hot. I wonder why new cellphones got this problem.
Yeh my phone geta hot after 15 min and loose fps....almost to unplayable state.... I need underclock cpu or lower resolution.... To keep it little more cool
shagratt said:
That's exactly why Im asking. Its a shame if you cant tune down the resolution and loose performance. Perhaps a custom rom may steal that function from the v30 now that is released.
My LG G2 never get hot. I wonder why new cellphones got this problem.
Click to expand...
Click to collapse
Just tried the otpion from battery settings...game battery saver..custom. And for the game i play (mobile legends) i lovered resolution to minimum and i lovered the fps as well.. Its much much better.... But will test it more... It says resolution lovered .... So i asume it must be doing something... As i am sure we cant notice 2k and 1080p drop on such a small screen....but cpu should stay cooler... Test it as well and let me know how you get on...please
Note i launch the game from the menu itself...by choosing option play....
Hunteres said:
Just tried the otpion from battery settings...game battery saver..custom. And for the game i play (mobile legends) i lovered resolution to minimum and i lovered the fps as well.. Its much much better.... But will test it more... It says resolution lovered .... So i asume it must be doing something... As i am sure we cant notice 2k and 1080p drop on such a small screen....but cpu should stay cooler... Test it as well and let me know how you get on...please
Note i launch the game from the menu itself...by choosing option play....
Click to expand...
Click to collapse
I think that only works with 3d applications but not affect daily use in wich the cpu/gpu need to move full screen resolution.
shagratt said:
That's exactly why Im asking. Its a shame if you cant tune down the resolution and loose performance. Perhaps a custom rom may steal that function from the v30 now that is released.
My LG G2 never get hot. I wonder why new cellphones got this problem.
Click to expand...
Click to collapse
shagratt said:
I think that only works with 3d applications but not affect daily use in wich the cpu/gpu need to move full screen resolution.
Click to expand...
Click to collapse
Doesnt say anything about 3d
..did you test it mate?
It seems that it defo makes a difference
Hunteres said:
Doesnt say anything about 3d
..did you test it mate?
It seems that it defo makes a difference
Click to expand...
Click to collapse
I still dont own the phone so cant test it but see a few videos/reviews about it and it render games at lower resolution. But only that, games, so all other drawing functions from normal aplications are not afected. It works as intended and indeed it help a lot in fast demanding games (like racing, shooters ,etc)
Yes like you said...did today hevay gaming over 1 hour.... No heat like before....warm yes...but no even close to it.... The resolution is noticeably lower..but i went to max fps and min res....
Defo a good feature...better then lower all res....
Cheers
I havent had any heating issues with g6. But resolution change would a thing nice to see with oreo update.
shagratt said:
I see that samsungs and v30 got the option to change screen resolution to save battery. Is that option in the G6? (I currently got an LG G2 and want to upgrade)
Click to expand...
Click to collapse
I bet we gonna get this in oreo update
https://goo.gl/UKaif3 This will help for the resolution but you have to be rooted
Oreo should bring resolution change option.
Non-rooted option: search Google for adb shell wm size and adb shell wm density commands. I haven't tried the size command but density gave me little bit nicer look and feel.
Bit of an old thread, but right now I'm running this res for better game performance:
adb shell
wm size 1080x2160
wm density 480
Notice that this causes the viewfinder in the stock camera app to be zoomed in and you won't be able to see the full image being captured. But I use the ported Google Camera which doesn't suffer from the problem.
https://forum.xda-developers.com/lg-g6/how-to/guide-best-battery-life-t3735720
here it is step by step with example for our loved G6... I didn't try it tho, and if you do, do it at your own risk!
can't wait to see if any of you isn't such coward, tries it, and tell us that it makes wonders for battery and heating....
edit: found the post here on XDA with the step-by-step guide!! give it a try someone!