Develop Bugs

Starhub user come in!!

Come in discuss what rom and radio you are using and how is the 3G connection.
Maybe we can help each other find the best radio for Starhub!

ROM: cyanogenmod nightly build 130
Radio: 5.11.05.14
APN: shinternet
Reception's not bad, mostly at full bar. For surfing is quite ok, can see facebook photos, youtube videos. Not exatly fast, but tolerable.

Maybe starhub is not giving enough love to android users but mostly to iPhone users. Hehehe.. 3G connection on hero & desire can totally suck balls!
Sent from my HTC Desire using Tapatalk

Forgot bout the ROM & Radio..
Sent from my HTC Desire using Tapatalk

Moved to general as not development.

speeds seem quite alright to me, maybe its just your location?
using DeFrost 5.1
i had speeds of:
downloads
1336 kbps
1353
1690
843
1372
uploads
964
386
344
379
75

Epic4G Modem Performance.

I have used and tested different modems on my phone using official Gingerbread leak EH06 Rom. For each modem downloaded I ran Speedtest app three times.
Afterwards I ran Xtremelabs Speedtest app twice to make sure that speedtest results weren't B.S. Then, I loaded a full webpage on the browser.
Testing was done during Daytime and Nighttime for each modem Except EH06.
I only have daylight results for EH06.
Modem Download Upload Ping Giantbomb
DK28: 958 kbps 628 kbps 158ms 54 seconds
EB13: 432 kbps 666 kbps 196ms 86 seconds
EC05: 453 kbps 365 kbps 391ms 60 seconds
EE03: 655 kbps 611 kbps 168ms did'nt load
EF02: 838 kbps 423 kbps 257ms 57 seconds
EF10: 1209 kbps 596 kbps 124ms 25 seconds
EG22: 324 kbps 253 kbps 275ms 53 seconds
EH06: 729 kbps 396 kbps 144ms 47 seconds
The Results above were recorded At 3:30-4:30 PM.
Modem Download Upload Ping Techreport.com
DK28: 832 kbps 639 kbps 161ms 10 seconds
EB13: 980 kbps 567 kbps 140ms 14 seconds
EC05: 772 kbps 564 kbps 138ms 11 seconds
EE03: 849 kbps 589 kbps 154ms 10 seconds
EF02: 1033 kbps 462 kbps 164ms 9 seconds
EF10: 1300 kbps 542 kbps 160ms 8 seconds
EG22: 1045 kbps 528 kbps 153ms 10 seconds
The Results above were recorded At 3:30-4:30 AM
8/20/2011 Daytime results were later than nighttime.

Looks like EF10 was the best.... I didnt use it that much..... Was it stable enough to keep?

Damn, might consider doing that pesky OTA anyone wanna persuade me not to?
Sent from my SPH-D700 using xda premium

Wheres di18?????
Sent from my SPH-D700 using Tapatalk

xopher.hunter said:
Damn, might consider doing that pesky OTA anyone wanna persuade me not to?
Sent from my SPH-D700 using xda premium
Click to expand...
Click to collapse
Noooo dont!!!!!.... Lol

dont ef10 is unreliable
dk28 was the best one for me

theduce102 said:
Looks like EF10 was the best.... I didnt use it that much..... Was it stable enough to keep?
Click to expand...
Click to collapse
I had gps Problems with EF10. DK28 gave me better results And much faster 4g.

jat255 said:
So far I've had the best success with EC05... The gps locks almost immediately and I haven't had any of the stability problems that I had with the GB modems.
I'll be on that one until the official GB hits.
FWIW, I'm in the DC area, so that may affect things.
Sent from my SPH-D700 using XDA App
Click to expand...
Click to collapse
I was in Slidell, LA when I performed this comparision test. over there the 3g speeds are acceptable.
However, here in baton rouge the 3g sucks badly. I get about 25-100 kbps during the daytime but during the night time the 3g speeds are normal.
We do have 4g although unofficially the speeds aren't all that amazing but its good enough.
I will perform another modem comparison test over here in baton rouge using both 4g and 3g including the new EH17 modem.

+1
I approve of this thread. It's nice to see a comparison chart!

eh06 works best for me in milwaukee

undervolt frequencies?

I just downloaded incredible control and am on doomkernel v6 and was wondering if anyone is running the phone with lower voltage setting than stock and still have the phone working.

Yeah, undervolting works nicely. Just gotta find the frequencies that are stable for you, because every device differs a bit

Roughly what are the frequencies
Sent from my Xperia Play using xda premium

Well don't blame me if these don't work with your phone, but mine are
Speed : Voltage
122880 : 750
245760 : 800
368640 : 800
460800 : 875
576000 : 875
652800 : 875
768000 : 900
806400 : 900
921600 : 925
1024000 : 975
I don't like overclocking so I haven't tested much above 1GHZ
PS: Some may seem like they can go lower but start bugging out after a while

i noticed that undervolting seems to produce very little stuttering every now and then, but as for myself, i feel like battery gains a little more life, like maybe 3-4 hours. Still testing though thanks for your frequencies and yes, i know the fact that every phone seems to react differently but i wanted to check what other sexplayers are using.

[Q] What is the best CPU Frequency in my testing? and Why ?

Today, i played SNES emulator on my play with difference CPU frequencies.
I found that the CPU% is depend on CPU frequency.
CPU Frq(Mhz) Volt(mv) CPU % Gameplay
249 800 100 lag
364 800 95 playable
460 850 80 playable
652 950 75 playable
806 1000 65 playable
1018 1100 60 playable
1209 1100 55 playable
1401 1150 50 playable
My question is ..
If i am interesting on battery life issue,
what is the best CPU Frequency in my testing ? And Why ?
Testing environment.
R800a ,Rooted,BL Unlocked,Doomkernel,SNESoid(Dragon Quest VI)
Thanks.

[Q] gsm gnex performance issues?

I've had my nexus a few months now and I am dissatisfied with its performance. It seemed to work fine on stock. There were no lag issues and all videos played between 58 and 61fps. I have since rooted it and have tried multiple custom roms. All of the roms I have tried have not given me the same performance I had on stock. Videos have dropped to an average of 29-43fps and the cpu has not seemed to be consistant.
Has anyone else seen the same thing or have any idea of how to increase the performance?
I know benchmarks are not a true test of performance but there is a noticeable difference in everyday use. Both are taken under ideal conditions using XenonHD and air kernel. As you can see there was also great variance in scores especially when clock speed is taken into consideration.
I am still attempting to pull a logcat, but the usb connection does not seem to be working properly. This may be the actual cord and remains to be seen if this another issue with the device.
normal clock speeds:
antutu: 3037
cpu: 988
gpu: 1244
ram: 335
IO: 470
quadrant: 1961
cpu: 5908
mem: 1773
2D: 95
3D: 1563
overclocked to 1.5ghz
antutu: 6620
cpu: 3261
gpu: 1528
ram: 1099
IO: 732
quadrant: 2259
cpu: 5726
mem: 1808
IO: 1194
2D: 245
3D: 2320
overclocked to 1.8ghz
antutu: 3343
cpu: 1098
gpu: 1277
ram: 467
IO: 230
quadrant: 2235
cpu: 5847
mem: 1662
IO: 1122
2D: 245
3D: 2301

antutu/quadrant blows
0xbench/linpack ftw
use performance governor, no oc
or make it easier on your device and yourself and stick to stock.
Sent from my i9250

thank you. I appreciate your response
I know benchmarks are not a true test of performance but there is a noticeable difference in everyday use.
Click to expand...
Click to collapse
but i ran linpack anyway and it currently returns values of 73.589 Mflop/s multi-thread and 40.954 Mflop/s single thread
second, this thread isn't about differing benchmarking techniques, it's about a loss of usability and the inconsistant performance of the device in an everyday environment and to see if anyone seemed to be having the same problems or have developed a solution. the benchmarks were a device to show the difference in performance quantitatively through a benchmark result
running stock is not an option, i use rooted apps

thezombiehunter said:
running stock is not an option, i use rooted apps
Click to expand...
Click to collapse
Scary, rooted apps. And whats so hard about rooting stock?
Sent from my i9250

Performance and clockspeed are not a linear scale. When you start pushing clockspeeds too high without properly balancing vcc it starts to become unstable. This is why the lower scores on the extreme overclock. As far as your user experience goes, just run rooted stock if you know that is when your performance was the best.