x PhoneArena is hiring! Reviewer in the USA

Here's why the Apple iPhone 5s' accelerometer is off

Here's why the Apple iPhone 5s' accelerometer is off

A little over ten days ago, we reported on what turned out to be a widely publicized issue with the new Apple iPhone 5s – it's accelerometer. It's been reported and confirmed beyond a shadow of doubt that the sensor is way off, though the exact cause split the community of unhappy Apple customers in two: some suggested that it's a software-side thing, while the rest argued it's the hardware.

According to RealityCap CEO and developer Eagle Jones, its the latter that had it right all along. As we noted in our previous report, Apple has gone ahead and switched its partner for the precision device from STMicroelectronics to Bosch, which would've been a-okay, but as it turns out, the new part actually appears inferior. Before you understand why, it's helpful to understand that accelerometers rely on two key measurements to showcase the level of their accuracy: bias and variance. The former describes the level of constant inaccuracy stemming from the hardware process, while the latter simply rates the consistency of an accelerometer's readings. Jones further explained this in an e-mail to Gizmodo:

Let's say we're just measuring the strength of gravity, which should be 9.81 m/s^2. 

low bias, but high variance sensor might output: 9.85 9.75 9.81 9.90 9.79 9.77 

high biaslow variance sensor might have output like: 9.60 9.61 9.59 9.60 9.62 9.58

All fine and dandy, but it would appear that while variance for the two products is almost identical, the bias values differ greatly:

This is where we find the problem: the typical bias for the ST part is +/- 20mg, while the Bosch part lists +/-95mgThis almost 5x greater offset range is confirmed by our measurements, and is absolutely consistent with the failures being reported by users and the media. Specifically, a +/- 20mgoffset range would translate to around a +/-1 degree accuracy range in tilt detection, and a +/-95mg offset translates to +/-5 degrees in tilt.

Meaning that iPhone 5s' equilibrium measurements can be off to a pretty wide margin if you're unlucky and get a more extreme case. What's more, Jones told Gizmodo that Apple could have actually compensated for this change with further factory calibration, but for whatever reason – they didn't. Luckily, according to Jones, this can still be salvaged, though the work-around and this whole issue will definitely serve to tarnish Apple's image of shipping products that just work from the onset:

It is possible to work around the problem by incorporating a calibration procedure into apps. This procedure would ask the user to place the device in different orientations to determine the accelerometer bias. Apps can then subtract this measured bias from the data coming from the accelerometer to get a corrected reading.

What's most disappointing, though, is that we still haven't heard of Cupertino on the issue, which is probably smart on their end. That's not to say that it's not important – anything from gaming, to use of any tools and apps that take advantage of your iPhone 5s' accelerometer, down to the touted dedicated M7 chip can and will be affected.

source: RealityCap via Gizmodo

  • Options

posted on 17 Oct 2013, 02:25 1

1. GeekMovement (unregistered)

I was wondering how the iPhone 5S review score got changed to 7.5, but then realized that link took me to the retro case review instead.

posted on 17 Oct 2013, 02:27

3. Chris.P (Posts: 567; Member since: 27 Jun 2013)

That's odd. Thanks for bringing it up.

posted on 17 Oct 2013, 02:32

4. GeekMovement (unregistered)

no problem!

posted on 17 Oct 2013, 05:10

16. ghaniosman (Posts: 78; Member since: 06 Jun 2013)

Comment was low 'bias'ed

posted on 17 Oct 2013, 19:45

25. GeekMovement (unregistered)

actually it wasn't a biased comment to begin with. I was just stating an error that I ran into. but i'm not surprised with your comment as this website is full of one-sided fanboys trying to start flame wars. To clear up any misunderstandings with the profile icon I have put in place, I have appreciation for every smartphones on the market and while I do prefer android at the moment as opposed to other platforms, I still think that ios, windows, bb, and other newer platforms are quite awesome.

posted on 18 Oct 2013, 07:19

26. Jack1192 (Posts: 35; Member since: 30 Mar 2013)

He was just joking. low biased, high variance thing mate. Chill out.

posted on 17 Oct 2013, 02:25 3

2. james004 (Posts: 486; Member since: 15 May 2013)

well people can stop complaining about phone arena not posting bad things about iphone.

posted on 17 Oct 2013, 02:35

5. Dingy_cellar_dweller (Posts: 265; Member since: 16 Mar 2013)

Apple will fix the problem with io8, just like the weak Wi-Fi with io6.

posted on 17 Oct 2013, 02:50

8. woodshop20 (Posts: 459; Member since: 14 Sep 2013)

But then they will break a million other things. Even Android nightly build developers aren't "skillful" enough to do that.

posted on 17 Oct 2013, 03:28 2

10. PhoneArenaUser (Posts: 5498; Member since: 05 Aug 2011)

iPhone 5S still has weak Wi-Fi:

"iPhone 5s Wifi Performance - Android Vs iOS 7 (Test 1)"

"iPhone 5s Wifi Performance - Android Vs iOS 7 (Test 2)"

posted on 17 Oct 2013, 07:44 2

20. Shatter (Posts: 2036; Member since: 29 May 2013)

Apple: Your wifing wrong.

Heres another oen about lag on the iphone 5s/browser.


Either IOS 7 is really unoptimized or apple is faking benchmarks for the A7. If they are faking benchmarks hopefully the jail breaking community exposes them.

posted on 17 Oct 2013, 08:22

23. PhoneArenaUser (Posts: 5498; Member since: 05 Aug 2011)

I know about that video, I saw all videos in Armando YouTube channel. :)

posted on 17 Oct 2013, 03:48 2

13. rusticguy (Posts: 2828; Member since: 11 Aug 2012)

Apple with fix the problem by Cook saying "Sorry we meesed it up yet again" publicly and issuing some iTunes gift cards.

posted on 17 Oct 2013, 03:56 1

14. cc3493 (Posts: 158; Member since: 06 Jul 2013)

IOS8 is a software update. This is a hardware issue... Get your facts right. Did you even read the article?

posted on 17 Oct 2013, 07:55

21. Tuxedo (Posts: 356; Member since: 19 Mar 2013)

The FACT is that every measuring instrument has some inherent error and that has to be accounted for via calibration. These days calibration correction factors are mostly done in software. The new part may be inferior from a HW perspective, but it can still be calibrated.

posted on 17 Oct 2013, 08:01

22. AliNSiddiqui (Posts: 382; Member since: 19 Sep 2012)

Did you not read the part where apps can ask to calibrate the hardware and improve the bias? iOS8 could have something like this in initial set up maybe.

posted on 17 Oct 2013, 02:36 2

6. Sniggly (Posts: 7305; Member since: 05 Dec 2009)

First: you misspelled "accuracy." Not a big deal, and I'm sure you'll fix it, but it tickles my "irony" sensor and made me smile.

Second: Apple is probably figuring out how best to handle this before going public with a response. With all the competition out there they would suffer from another "Antennagate" issue: a relatively minor design flaw/defect that got blown way out of proportion by the ineptitude of Apple's responses. They really, REALLY don't want to mishandle this issue as the holidays are coming up like a freight train, as Apple already has enough trouble figuring out how to sell its 5c stock.

posted on 17 Oct 2013, 02:49 11

7. woodshop20 (Posts: 459; Member since: 14 Sep 2013)

What's happened to Apple quality control? This is what happens when you spend too much time suing others and not enough time improving your own products.

posted on 17 Oct 2013, 03:36 2

11. ScottSchneider (Posts: 336; Member since: 06 Dec 2011)

Good one... High time they start focusing to kiss their own a$$ on Invention and Quality of products they sell instead of trying to sue OEM's A$$e$ with Law suits...

posted on 17 Oct 2013, 07:44

19. Finalflash (Posts: 3535; Member since: 23 Jul 2013)

This is what happens when you desperately try to cut costs to increase margin and end up decreasing value. Welcome to the real post-Jobs Apple, where cool stuff takes a back seat to the money in a big way.

posted on 17 Oct 2013, 03:13 6

9. Tsepz_GP (Posts: 1043; Member since: 12 Apr 2012)

You're holding it wrong.

posted on 17 Oct 2013, 03:45 6

12. rusticguy (Posts: 2828; Member since: 11 Aug 2012)

MyxTyxPyx says -- oh this is trivial issue.
Phone still works.

Golden iPhone 5S will compensate for this going forward :)

posted on 17 Oct 2013, 04:33 2

15. itsdeepak4u2000 (Posts: 3718; Member since: 03 Nov 2012)

What a mess they have made! Previously there were not so many issues, now-a-days iOS7 and iPhone are coming up with all these things.

posted on 17 Oct 2013, 07:06

17. Googler (Posts: 813; Member since: 10 Jun 2013)

If Apple had the procedures for users to calibrate the sensors, the impression of things working perfectly out of the box would be dead. Other than that, I'd like to know the reason they changed supply companies for the sensor. If it was just to get a cheaper priced product over getting a quality product, shame on them.

posted on 17 Oct 2013, 07:20

18. AfterShock (Posts: 3698; Member since: 02 Nov 2012)

Go figure Apple is silent on the issue.

posted on 17 Oct 2013, 11:08 1

24. -box- (Posts: 3991; Member since: 04 Jan 2012)

Surprise! A cheaply-made device made with cheap components and cheaply-made software isn't very accurate!

Want to comment? Please login or register.

Latest stories