Reviews - Reader comments

GSMArena feature labs: The tests

GSMArena feature labs: The tests - read the full textThe GSMArena Labs are tests designed to give an objective account of how a device performs in real-world scenarios. In short, it's about what you can expect from a gadget you're planning to buy. Our tests are constantly being developed...


Sort by:

> In reply to Cowz @ 2012-05-23 20:10 from vxKC - click to readagree ,

the screen should view a white page and adjust brightness to obtain defined amount of lux per 1 inch (not tofavour large screens) , thatbrightness level will be the test level, then the web and the movie shouldn't be very dark nor very bright.

  • Reply
  • 2012-10-05 16:09
  • fXDF

it is impressive to see so professional comparison, I wonder is it possible to list the output impedance of compared phones? because impedance will affect the output power.

  • Reply
  • 2012-09-27 19:07
  • 4MXa

> In reply to Anonymous @ 2012-08-22 21:37 from vGkq - click to readHahaha, my thoughts exactly...! How about an Apogee Duet or something?

  • Reply
  • 2012-09-19 19:49
  • f0x9

Fast track pro? seriously?

  • Reply
  • 2012-08-22 21:37
  • vGkq

> In reply to Paul @ 2012-05-12 18:13 from 3E0p - click to readI kinda have to choose,either a PS3 or a phone(Xperia Play).I'm still in a big demlia as I like portable gaming but I also like the PS3.It can happen all type of shit till summer and just not get the console at all,let alone I can get a job for the summer and then get my console,even though I wait for the summer all year and just waste it doing something instead of finally wasting time like I like it's kinda hard for me to digest.

  • Reply
  • 2012-08-05 14:03
  • xghb

> In reply to boybawang @ 2012-05-12 07:22 from v0XM - click to read+2This is a very good move from SE. They're kind of leading the way here, psoriming updates for their whole line-ups ahead of time. I wish more manufacturers did that. But SE really needs to get their act together regarding hardware, too. They always seem to use 1 year old processors, like they're RIM or Nokia. I expect better from a top Android manufacturer. Plus, they need to make all their phones as slim as the Arc (or slimmer). Most of their phones, especially the lower-end ones are still as bulky as they were in 2005. Unacceptable. I do like that they kind of have their own design style, there, though. That's respectable. But if they really want to be taken seriously by Android customers, they need to up their game in the hardware department, too.

  • Reply
  • 2012-08-04 00:07
  • PNnV

> In reply to Anonymous @ 2012-06-05 10:00 from ftb} - click to readSamsung Galaxy S run the Ice Cream Sandwich version and in a pttrey good way as you can see in this video. To know more about it, you can go to the his blog to download anda0 tutorial install Android 4.0

  • Reply
  • 2012-07-02 13:23
  • 35ta

> In reply to EskeRahn @ 2012-04-25 19:57 from 0u3G - click to readI'm garetful you made the post. It's cleared the air for me.

  • Reply
  • 2012-07-02 12:44
  • Jjyx

> In reply to Ace Angel @ 2012-04-26 18:03 from u7m9 - click to readace angel i do agree with you hey

  • Reply
  • 2012-06-05 10:00
  • ftb}

> In reply to not happy @ 2012-05-19 13:46 from gxnw - click to readI was about to post something similar. Using 50% is an awful decision and I can't see how if any of these tests are in any way well performed that they could have come up with such a test method.

The results are worse than meaningless, they are deceptive!

Please set phones to a standard uniform lux setting. You have to also declare the lux setting that could be achieved for the test.
Not declaring the lux output in the current tests is just hiding from people how flawed this test is.

A real shame as that could have been a great feature for this site.

  • Reply
  • 2012-05-23 20:10
  • vxKC


thanks for your efforts - however, I think that results concerning the screen are utterly useless (even unprofessional)... let me explain:

First of all you set the screen brightness to 50%. That's not good enough as each smartphone will emit a different brightness at its 50% setting. Results vary immensely. By setting brightness to 50% you punish phones that offer high screen brightness; therewith punishing good outdoor visibility. And all of that totally unnecessary.

You need to choose a certain brightness level in lumens.. or candelas whatever your favorite brightness is. Choose anything that is good indoors. Then measure again.

Let me be clear: That's a complete FAIL on your part and comperability is not given.

Second: The movie. Which one is it? Is it Batman with almost all black parts therewith completely favoring AMOLEDs or is it some documentary on Antarctica - therewith destroying AMOLEDs? That's just not good enough. You need to choose a movie and state which one you use.

Anyway, that's not as much of a fail as the 50% brightness.. Or do you want very dark phones? As you those would rule your test?


  • Reply
  • 2012-05-19 13:46
  • gxnw

it would be nice to have an interactive database of all phones tested, so that one can sort the phones by web browsing,YOUR stand-by figures (and not the manufacturer's) and total endurance rating

  • Reply
  • 2012-05-12 18:13
  • 3E0p

Dear gsmarena,

Please specify the resolution and bitrate of the Video used in Battery life test.


  • Reply
  • 2012-05-12 07:22
  • v0XM

For the Battery test... I suggest do include battery endurance on GPS.
Hav a nice weekend ya~

  • Reply
  • 2012-04-26 18:03
  • u7m9

Please add detailed GPS tests. No website does that and GPS is quite important in modern phones. Most times you just check the lock speed and not much more. Would be good to have tracks displayed, to show accuracy, not only how fast it locks but how precise it is, antenna quality, etc.
That would help in GPS failures like the Galaxy 1.

  • Reply
  • 2012-04-26 10:05
  • n4Y9

Excellent article, thanks GSMArena for doing all these test for us all; they do help a lot. None of the following is in contradiction with the article, it's just a few additional points / clarifications I'd like to make:

- Display contrast test keep speaking about "black" levels. While entirely correct, this may create the impression that contrast only matters between "white" and "black" parts of a picture, which would fail to take into account that all colors on any display today are composed of distinct red, green and blue light emitters. Thus, for all fairly "pure" colors (mostly just red, just blue or just green) the two "unused" sub-pixels need to stay as "black" as possible for a vibrant result - therefore contrast is quite important even in a picture that has no obvious black parts at all.

- Acoustics are measured in decibels (ie. on a non-linear, logarithmic scale) for a good reason - because that's the scale human ears work on, too. Therefore while it is absolutely true that a 3db difference equals double the sound intensity (and a 6db one equals double the sound pressure) in a physical sense, you need a difference of TEN decibels to double the sound volume as perceived by the human ear [1]. A few dbs therefore take a lot more energy to generate but do not sound a lot louder.

- Stereo crosstalk is sound in a channel "bleeding over" from another channel, and the less of it you have the better. However, it's important to remember that zero decibel is an arbitrary level of sound intensity (about the typical threshold of hearing of an average human) and since crosstalk is very small usually, amounting to very quiet sound levels, it gets expressed as NEGATIVE decibels. Therefore, the less crosstalk there is, the bigger the db number will get, but in a negative direction (-50db crosstalk is much better than -30db). The original 'The larger this value is "the further away from zero"' is meant as "further away from zero in a negative sense", not as "larger in a mathematical sense, where -5 is a larger number than -50 is".

- Everybody knows that displaying remaining battery in most portable devices today is closer to predicting weather than factual appraisal. What not everybody may realise though is that in fact there is no direct way of telling how much energy is left in a cell. An ideal battery would keep its voltage perfectly constant - for any load - until it becomes empty, and absoultely noone could tell how much juice is left in it without taking it apart and analysing it chemically. The only reason we CAN (somewhat) tell what is in there is because cells are not perfect, and their voltage keeps dropping slightly as they discharge - emphasis on slightly. This small change is what devices are trying to detect - but considering that the change is small, dependes on the momentary load, the age of the cell and the way it has been used, the ambient temperature, and just possibly the alignment ocf the stars (just kidding), it becomes understandable that pretty much every portable device suck at telling how much energy it has left. The point where this becomes regrettable however is that these devices use the same measurement to decide they have too little energy left at some point, and switch off - possibly when there would still be quite a bit of juice in the batteries if only they would keep going. Which explains also why apparent battery capacity can depend on the firmware version of the device - even if said firmware does exactly nothing to reduce consumption, it might measure / display the battery levels differently.


  • Reply
  • 2012-04-26 08:51
  • Mvvy

This is a really good idea for the tests for the phones, and it really gives a clear idea on how they perform....AWESOME JOB!

  • Reply
  • 2012-04-26 00:57
  • Mf5U

This is why GSMARENA is > all . thank you gsmarena

  • Reply
  • 2012-04-25 20:01
  • 6QqM

Thanks for a great site.
I'm especially happy on the new display and battery tests *ThumbsUp*

  • Reply
  • 2012-04-25 19:57
  • 0u3G

Note: Sponsored advertising links are in green.