• Welcome to the all-new Windows Central Forums! - We're still moving some things around, so you may see a few quirks here and there, but we're working on getting things fully completed as soon as possible. For now, take a look around, and if you run into any major issues, please let us know in this thread!

How Accurate is your Digital Compass?


Active member
Jan 12, 2013
Most modern smart phones are now equipped with what they call a "digital compass" or a "magnetometer". A neat feature, actually, and was (and is still) one of my major criteria in choosing a phone. Phones with this neat thing basically have three coils (very small ones) or something similar that are oriented perpendicular with each other such that each one of them is at right angles with the other two, giving them a three-axes Euclidian space orientation... imagine three donuts, one placed flat on a table and the other two standing on the edge of the first donut.

While the conceptual construction may be straightforward, the performance isn't as straightforward. Many users, especially in Windows Phones, have reported (and complained) being asked to re-calibrate the compass frequently, at times almost every time a compass-using app (such as Here+ Maps) is launched... at least frequent enough to start trigger a complaint. Often times the cause of the problem is having a case with a magnetic lock where the obvious solution is not to use such a case. Yet the issue of calibration isn't really limited to those with a magnetic case - just changing locations can prompt another calibration session.

If the three-donut construction of the magnetometer is so straightforward, why are we users being asked to recalibrate more often than what we feel is needed? Isn't this piece of nano-composite technology high-tech enough to be always accurate? Let's find out.

Where I live, there are at least five devices that have a magnetometer: my L1020, an iPhone5, an iPhone5s, an iTouch, an iPhone4s and an HTC One - only the L1020 is mine. The digital compass of each device were compared against two "real" compasses (a simple under-$1.00 compass and a professional navigator's compass). Calibration was performed as needed or as requested by the device. The phone was aligned with the compass needles, and the variation in the heading (which should be 360 or 0) is noted.

So, first off, the L1020. Upon launching the native Maps app and tapping on the teardrop shaped icon on the upper right corner of the screen, I was prompted to calibrate the compass (as usual). But instead of performing the recommended figure 8 motion (which would make me look like I'm playing some child's airplane game), simply slowly rotating the phone about all three-axes calibrated the compass. The result was a consistent +7-degree deviation from magnetic North at home and in the office. However, once the phone is calibrated while inside a car, the deviation is sometimes reduced to just 1 degree (barely noticeable) - but results in moving vehicles aren't reliable since the range of the deviation is from 1 degree to as much as 20 degrees.

Second, the iPhone 5. No calibration required. Compass deviation: +4 degrees.

Third, the iPhone 5s. No calibration required. Compass deviation: -3 degrees.

Lastly, the HTC One. No calibration required. Compass deviation: +11 degrees.

Wasn't able to test the iPhone4s at this point in time.

Interestingly, the iPhones had less deviation compared to the L1020. The HTC One had the largest deviation, but for practical purposes such a deviation would barely affect the performance of a compass-using app. Does this suggest higher manufacturing standards being implemented in iPhones? I don't think so. I tested another iPhone 5s a few months ago and it gave a different deviation value (+10 degrees if I remember correctly); even two phones of the same model do not produce the same reading deviations.

So, that's for the devices around me. How about your digital compass? Ring in below with your findings. :cool:

Geo Hutchings

New member
Mar 1, 2015
I never had to calibrate my nokia 930, I did with 925 which came out about same time as 1020, the 930 is always basically on point


Ambassador Team Leader
Nov 4, 2013
I have a large military binnacle mount compass that I can use for magnetic north which is very accurate in the mechanical sense.

My 925 is accurate within 2 degrees after calibration and settling down.
(This is outside away from any known interference sitting on a wood platform)

My Xperia Z1s is dead on with the mechanical compass (same criteria)

If I remember correctly for this local we have a 12 degree +/- variation for navigational purposes.


Active member
Jan 12, 2013
I do not have a real compass to compare but before I saw this video I made the wrong 8 and a calibration took a long time.

I wish my L1020 could be calibrated that quickly using the Figure-8 method. It takes me at least three Figure eights, sometimes eight or more, to have the compass calibrated.

There's another and more discrete method, though: hold the phone parallel to the ground and rotate it slowly 360 degrees - faster than the F-8 method for my L1020.

Members online

Forum statistics

Latest member