Valid HTML 4.01 Transitional

Samsung Galaxy S III
Checkout Details

Jim Carter, 2012-10-23

To-do list:

The checkout page is a summary list where long discussions will not fit. This page is for additional information about the various checkout steps.

Removing the Back Cover

On the Galaxy S III the back cover seems securely attached, but it's easier to get off than the Droid-3 was. In the center of the north edge is a screwdriver slot for prying; a (male) fingernail will also work. I found it most effective to attack the gap toward the east, i.e. the side where the LED flash is, away from the wired headphone jack. When you get the northeast grabber open, go down the east side peeling off the grabbers, and then similarly do the west side.

To replace the cover, engage the two south grabbers first, then the sides, and the north grabbers last. In all of this, be careful not to touch the power button and accidentally turn on the pocket computer.

Setup Steps -- Stock Image

Activating the Phone (Verizon)

This section applies if you did not purchase your phone directly from Verizon. If you purchase from a brick-and-mortar store, the customer service people will do all this for you. If you purchase online, there is a web form to help you through activation.

A phone is generally SIM-locked to the carrier that subsidized it. If you intend to use another carrier's SIM, e.g. if travelling abroad, you will need to break this lock. The carrier should unlock it after a certain time, or if you can't make this happen there are unlocking services (not free). Unlocking the phone does not relieve you of your obligation to pay under your contract or let you avoid paying the unearned part of your subsidy (early termination fee).

Call Verizon Customer Service on another phone, e.g. a landline. Once you have picked your plan, told them your MEID (for CDMA, or IMEI for GSM), and paid, the customer service rep will establish an association between the MEID and your account, and they will send e-mail to confirm. This typically takes about 15 minutes.

Then, activate your phone. Initially, in the status bar your carrier logo is a flashing triangle, indicating no service.

*228 Option 2 is to just update the Preferred Roaming List (PRL). If the network is updated, e.g. 3G is added or a new roaming agreement is made, your phone needs to be told to use the new capabilities. Some forum posters suggest doing this update monthly, though that seems kind of frequent to me. Now the Holy Grail is to get another carrier's PRL onto your phone, because otherwise it can't even see the other carrier's towers.

Display in Sunlight

The display is excellent in a wide range of lighting conditions. It is hard to fight the sun and win, but in direct sunlight I judge the contest to be a draw.

Display in Artificial Light

The Galaxy S III's AMOLED display, viewed in artificial light, gives good color rendition which is independent of the viewing direction. When viewed at a radical angle, e.g. nearly edge-on, the brightness is maybe about half that for perpendicular viewing.

Light valve (LCD) displays generally have trouble giving good color when viewed off perpendicular.

Ambient Light Sensor

The ambient light sensor works, and influences the display brightness. It is on by default in both the stock image and CM-10, but you can turn it off, and at least on CM-10 you can tweak the control parameters. It's in Settings - Display - Automatic Backlight. You particularly need to turn on Light Sensor Filter - Enabled, to use an average rather than instantaneous brightness, and Allow Light Decrease, otherwise the brightness can only increase (dumb).

About lux: The light flux is reported in lux, from 107000 lux in (my) direct sunlight to 1 lux in a dark closet. The value can be seen on Phone Tester and on GPS Status :-) See this Wikipedia article about luminous efficiency. Lux means lumens per square meter and lumens are proportional to watts but depending on how strongly the eye responds to the various colors in the light. The denominator may be the total power into the lamp or the equivalent power of the emitted light. On the latter definition, the theoretical maximum scale factor is 683 lumen/watt; practical lamps range from 0.3 (candles) to 100 lumen/watt for the highest scoring LED lamps and plasma tubes (which are similar in efficiency).

Details About the Internal SD Card

There is 16Gb of internal NOR flash memory, where the previous generation of pocket computers have around 256Mb of NAND flash. /proc/mounts show these partitions mounted; there are definitely partitions we don't see for the booter, the kernel, and probably the radio image. Sizes here are in Mb (220 bytes). These are for CyanogenMod-10.

Mountpoint Size Used Format
/system 1024 390 ext4
/efs 13 4 ext4
/data 12000 394 ext4
/cache 826 77 ext4
/persist 7 4 ext4
/firmware 63 41 ext4
/storage/sdcard0 12000 394 (fuse)
/storage/sdcard1 1024 303 vfat

The /data partition is replicated on /storage/sdcard0 by a FUSE mount (not a bind mount); I haven't yet found out what peculiar filtering happens through this mount. I also don't know the purpose of the three small partitions /efs, /persist and /firmware.

On CM-10 the SD cards are accessed through two symbolic links: /sdcard to /storage/sdcard0 and /external_sd to /storage/sdcard1. On the Verizon image, /sdcard also points to the internal NOR flash, but /ExternalSD is to the external card (I think; this is from memory).

There is 11Gb for generic user data on the internal NOR flash. In the Verizon image this is mounted on /mnt/sdcard with a symbolic link /sdcard to this mount point. The camera app stores photos in /mnt/sdcard/DCIM. If there were OTA updates, which Verizon gives you but CyanogenMod doesn't have, the new image would be deposited in /sdcard. I don't actually know which filesystem types are acceptable on /sdcard. But it is generally believed that Recovery can only read VFAT when flashing an image.

There are suggestions in the settings that you can encrypt your /data filesystem, but encryptfs is not mounted by default, unlike in at least one previous version.

Data Speed Tests

The data speed test is to download a compressed (Vorbis) music file of 2.45e7 bytes (24.5Mb) or 1.96e8 bits, using HTML protocol (no encryption, no second compression). Note, all the data rates in this section are in bits/second, not bytes/second. Via 802.11g at 5.4e7 bits/sec the download should take 3.62 seconds. The theoretical speed is never achieved in the wild. To delete the downloaded copy, in the browser hit Menu -> More -> Downloads. Long-click on a filename and you get a menu including Delete.

Download times for this file:

Download To Protocol Carrier Seconds Bits/Sec
Server Read and checksum on server 0.86 2.27e8
Laptop 802.11g 8.2 2.39e7
HTC Dream 802.11g (CyanogenMod-6.2 / Android-2.2) 80 2.45e6
HTC Dream GSM (EDGE 2G) AT&T 253 1.78e4
Droid 3 (VZ) 802.11g (stock Android-2.3) 14 1.4e7
Droid 3 (CM9-12-24) 802.11g 11 1.8e7
Droid 3 (CM9-01-30) 802.11g + Bluetooth 138 1.3e6
Droid 3 (CM9-01-30) 802.11g (no BT) 10.3 1.7e7
Droid 3 (CM7) 2G (1x-RTT) Verizon 120 3.2e4
Droid 3 (CM7) 3G (EVDO) Verizon 600 8.2e4
Droid 3 (CM9-12-24) 3G (EVDO) Verizon 93 7.5e4
Droid 3 (CM9-01-30) 3G (EVDO) Verizon (BT) 151 5.6e4
Droid 3 (CM9-01-30) 3G (EVDO) Verizon (no BT) 104 8.0e4
Galaxy S III 4G LTE Verizon 43 4.16e6
Galaxy S III 802.11n(g?) 12 1.49e7

On the slower physical layers, the download was cancelled partway through and the reported rate is based on the time and progress up to cancellation.

If Bluetooth (music via A2DP) is running at the same time as the speed test, it slows down 802.11 radically, and EVDO noticeably.

Using the OpenNetworkMap app, it downloads a 4Mbyte test packet at 1.36e5 bit/sec, but uploads 4Mbyte at 1.4e6 bit/sec (probably bogus). On a retest, 2.2e5 bit/sec download, 2.9e5 bit/sec upload. This app reports in bytes per second, and the numbers here have been multiplied by 8 giving bits per second.

Bluetooth Testing

USB Connection

When you plug the provided USB cable between the Droid and a computer running Linux (or presumably Windows), it plays an audio notification and pops a notification item saying Connecting mass storage to USB. Both SD cards are unmounted. You can open the notification and change to Charge Only, whereupon the cards are remounted.

You get one USB major device Motorola PCS 22b8:42d7. The iProduct is identified as XT862 (correct) and a serial number of 0910E4490202200E is given (for mine). It has a bInterfaceClass of 8 Mass Storage. There is also a Vendor Specific Class 13 Motorola ADB Interface. The mass storage interface has two discs.

The internal card, labelled MOT, on /dev/sdb, has 11.9Gb of storage. Its subdirectories are DCIM (photos), download, Android, .android_secure, and slacker.

I formatted my external card with proper partitions and it is on /dev/sdc1.

I copied some local HTML documents onto /media/MOT, and a file URL is honored. Example: file:///sdcard/htdocs/leones_h.d/leones.shtml

Details of the Battery Test

The AnTuTu battery test unfortunately doesn't report objective units for its score, such as the battery's measured capacity (versus the nominal capacity of 2.1 amp-hours), nor does it report the power used by the processor and display in its standard task. Thus this test is useful for tracking how a battery degrades with time, but not for comparing between devices of different types. The test turned the display on full and ran the CPU at 100% (how many cores?).

When I did the test, getting a score of 514, here is the history of discharging and charging:

Minutes Charge Voltage
0 100% 4.2
48 80% 4.0
93 58% 3.8
113 49% 3.7
147 33% 3.6
173 18% (missing)
188 15% 3.7
0 14%
46 39%
116 86%
137 95%
148 97%

Jimc's Suggestions for Saving Battery Power


Miscellaneous items found when working on other issues: