Become a MacRumors Supporter for $50/year with no ads, ability to filter front page stories, and private forums.
It certainly is nice to know that this should be an easy software fix. Would've been really disappointing for people to know that they are just stuck with unfixable faulty hardware in their brand new phone.

Apple should be able to implement something at the system level, perhaps a tweak of the API such that devs don't have to worry about adjusting their code, right?

If the variance (error) in output is as large as the article suggests, then nothing in software can fix it. The suggestion is that individual components have accelerometer reading errors that can vary from anything to plus or minus 95mg. Sure, you can set any offset in firmware that you like, but unless it can be calibrated against each individual device, that would be useless. Just as bad is that the output can drift over time and would need recalibrating. As far as I see it from the given facts, the only solution would be to allow the user to calibrate his or her device, and live with the fact that it would need recalibrating from time to time. Not good at all.
 
Last edited:
You know what most people will still take from this news?

"Oh, hey it comes with a level!"
 
scary

Scary, they call it measurement bias. If I had a 3 degree measurement bias on my trip to the north pole, I would have ended up dead somewhere on the Kamchatka peninsula!

FIX IT APPLE! or replace them.
 
The "Carpenter" app I installed on my iPhone 4 over 2 years ago is dead-on, but it doesn't have a compass. I'll just install it on my 5s when it arrives.
 
Nobody who seriously uses a measuring stick or compass uses an iPhone to do those tasks...

The wonky measurements are fine for casual iPhone users I guess
 
The problem exists with the iPhone 5 and iOS 7 also!
Compass and level are not accurate !!!!!
Making them worthless.

The thing is they are precise (Meaning the same number of degrees off every time) but not accurate. Using the precision they could add a software calibration that fixes it. But you would have to have another working compass/level to do it with.
 
Strange. Mine is perfect


I didn't say they didn't work, they are just inconsistent. I have checked this on 4 different iPhone 5 and even took up to the Apple store where they reconfirmed the inaccuracy on multiple phones.
Try turning on the compass doing the calibration roll, lay it on paper. Mark N.
Then quit the compass and do this over again. It's off a few degrees and I have seen it vary as much as 10 degrees.
 
It's amazing how someone can figure out what the problem is in a few weeks, while all of Apple engineering let this one get by them.

This is the same kind of issue as the iPhone 4 antennagate was. Apple has billion dollar instruments, 12 months development and testing time, and yet they can't find blatantly obvious mistakes, which every customer can see right away without any special tool. It's odd, especially from a company that is obsessed with the slightest detail, the color of an icon, the thinness of a font, and every single pixel.

As a software and electrical engineer myself, I know that subtle issues can remain hidden, which is very hard to prevent entirely. Perhaps it was too late to fix it, and they had no choice but to release it broken. Maybe their supplier gave them working samples during the beta testing, but the final mass shipment was defective.
 
Is there any new information here that isn't on the 70+ page thread in the iPhone forum?

Is there any information in the 70+ page thread that's not in the handful of paragraphs above?
 
calibration

Set up a table at all Apple Stores worldwide. Have the surface of the table calibrated to be dead flat in all directions.
...

You don't need a calibrated table, just a flat surface. Put the iPhone on its side, start calibration, turn 180 degrees, finish calibration.

This should be done as part of the system settings, not per-app. Then the calibration adjustment can be incorporated within the system APIs.
 
I have tried to use the compass in the past, before this became a public issue, and I can confirm it was not accurate, even after calibration. I hope Apple fixes this soon.
 
It's amazing how someone can figure out what the problem is in a few weeks, while all of Apple engineering let this one get by them.
Whatever Apple's priorities are, quality control is not currently one of them I'm sorry to say. iOS7 has been kind of ridiculous. All the inconsistencies between icons, some one way some another, some yes/no toggles in settings are actual toggles, some are a sub-page where there is a single yes/no toggle... why? All very basic stuff that *one* human factors engineer could detect in a matter of hours. What the heck? It's hard to throw crap at Android when iOS7 is turning out to be the less stable, less consistent OS. :(
 
Ok here.

I used my iPhone 4s with OS 7.0.2 and my super accurate level... no problem. I got a perfect reading.
 
I used the iPhone's level to build my house. There might be some truth to that report...

leaning_house_2.jpg
 
You don't need a calibrated table, just a flat surface. Put the iPhone on its side, start calibration, turn 180 degrees, finish calibration.

This should be done as part of the system settings, not per-app. Then the calibration adjustment can be incorporated within the system APIs.

You are right. This should be in the Setting app.
 
Well this proves its a hardware issue, caused by what seems to be Apple switching sensors due to size contstraints. (Another article posted that this Bosch sensor is smaller.) Now the question is can Apple come up with a permanent software fix. I really don't want to have to calibrate every single time I use the level, play a game, take a pano photo, etc......

You wouldn't need to calibrate it on each use of your software. The article states that the bias doesn't change much, if at all, over time so you would calibrate it once on a system level and then all your apps would have access to the correct values.

I don't understand what all the whoop-de-doo is all about as calibration was suggested in an *old* app I had for my iPod touch 2.
 
I suspect :apple: to stay mum on this flaw. Tim burned an apology last year with Maps so I don't think another admission of oops is forthcoming.

However, what I do think will happen is an update that addresses a "Very small" portion of iPhone 5s users who see this behavior. Say less than .01%? :rolleyes:
 
As far as I see it from the given facts, the only solution would be to allow the user to calibrate his or her device, and live with the fact that it would need recalibrating from time to time.

I think I could live with this. It's certainly better than having no recourse for faulty hardware in my device. Of course, I would rather they have gotten it right the first time, but I'll take what I can get. I do actually make use of ability of my phone to be a digital level.
 
Register on MacRumors! This sidebar will go away, and you'll see fewer ads.