Mavic 2 Pro Altitude recorded on photos

[axident]

Hi all, I bought a Mavic 2 Pro 2 days ago and while I was impressed with the camera I had no end of problems with the altitude shown on my photographs. When taking the photos my controller showed the correct altitude but when I download photos to PC via any program the altitude showed anything from -7sm to + 72m even if only 2-5m to +94m from take off. After many upgrades of firmware, IMU, and compass calibration nothing changed. so there went 2 days of my life trying to fix it. I gave up and DJI support said go to retailer and get a swap which I have done. I will try out the new one with updated firmware tomorrow morning South Australian time. The other issue I had that the number of times charged on each battery remained at 0 charges..
The question is has any one else had a similar issue? Best regards to all

 

[sar104]

There are four different altitude fields in the photo EXIF data, two of which are mislabeled:

Absolute Altitude : +2849.93 (Barometric absolute altitude based on a standard atmosphere)
Relative Altitude : +73.00 (Barometric altitude relative to the take off point based on a standard atmosphere)
GPS Altitude Ref : Above Sea Level (A pointless field that actually just notes whether the absolute barometric altitude is above or below mean sea level)
GPS Altitude : 2849.9 m Above Sea Level (Also barometric absolute altitude - not GPS altitude)

The relative altitude field will be the same as the altitude shown on the controller and GO 4 app. The altitude displayed by your computer program could be any of the above - you need to examine the full EXIF data to see what it is actually quoting.

 

[Meta4]

The question is has any one else had a similar issue?

Everyone has.
There's nothing wrong with your drone, that's how they all are.
DJI support people probably have no understanding of the issue and just told you to go to the supplier to get rid of you.
The exif altitude shown is never the relative altitude that you see on your screen.

Why are you interested in the exif altitude?

 

[sar104]

Everyone has.
There's nothing wrong with your drone, that's how they all are.
DJI support people probably have no understanding of the issue and just told you to go to the supplier to get rid of you.
The exif altitude shown is never the relative altitude that you see on your screen.

Why are you interested in the exif altitude?

That's not quite correct - the EXIF "Relative altitude" field is the OSD_height value shown on the controller and app.

 

[Meta4]

That's not quite correct - the EXIF "Relative altitude" field is the OSD_height value shown on the controller and app.

The other heights are hidden in there somewhere but I was referring to the simple single altitude displayed by most common exif reading programs

 

[sar104]

The other heights are hidden in there somewhere but I was referring to the simple single altitude displayed by most common exif reading programs

Right - a lot of programs will just display the absolute barometric altitude from either the absolute or misnamed GPS fields.

 

[gnirtS]

The inaccurately labelled fields and data has been an issue since the pre 1.0 firmware and not been addressed by DJI.
Its also worth noting that GPS altitude can be extremely inaccurate due to the way its derived. Its easily possible to be out by 60m+ at times.

 

[sar104]

The inaccurately labelled fields and data has been an issue since the pre 1.0 firmware and not been addressed by DJI.
Its also worth noting that GPS altitude can be extremely inaccurate due to the way its derived. Its easily possible to be out by 60m+ at times.

I have never seen errors of that magnitude with a good view of the constellation. Theory indicates that vertical error is approximately 1.5 times the horizontal error, and that's what I've seen.

 

[gnirtS]

Its quoted as 1.5 times but i rarely if ever see the differential that low even on proper (ie not a mobile phone etc) GPS receiver.

Sat here now (20+ satellites in fix) im seeing a 30m error in my altitude (from a known, surveyed point) with 8m forecast horizontal accuracy.

I see 50-60m difference from properly surveyed altitudes a lot.

Its rare, if ever i see an altitude more accurate than 20m off.

 

[Meta4]

I see 50-60m difference from properly surveyed altitudes a lot.
Its rare, if ever i see an altitude more accurate than 20m off.

Not doubting you but that's not relevant to the OP's question.
Although the name of the field in exif info suggests the height is derived from GPS data, it actually comes from barometer data.

 

[sar104]

Its quoted as 1.5 times but i rarely if ever see the differential that low even on proper (ie not a mobile phone etc) GPS receiver.

Sat here now (20+ satellites in fix) im seeing a 30m error in my altitude (from a known, surveyed point) with 8m forecast horizontal accuracy.

I see 50-60m difference from properly surveyed altitudes a lot.

Its rare, if ever i see an altitude more accurate than 20m off.

I have no idea why you are seeing such large errors. What is your actual horizontal error? I can't remember - are you in an area with WAAS?

 

[axident]

Thanks for your replies. I will give an update after setting up the new M2Pro replacement

 

[Vitamin-J]

Re: Battery Charge Count: I’ve noticed this as well early in my ownership of the M2. I now believe the battery charge counter only increments up a number when the battery has accumulated 100% worth of charge. As this is the meaningful number when it comes to Lithium ion batteries life span. Thus, if you were to fly and burn 30% of the battery, recharge, then fly again burning 20% of your battery, recharge, then fly a 3rd time consuming 65% of your battery. When you’re done recharging after this 3rd flight, your charge count will read as “1” as the 3 recharges you’ve conducted sum up to 115% of your battery.

 

[Meta4]

I have no idea why you are seeing such large errors. What is your actual horizontal error? I can't remember - are you in an area with WAAS?

I don't think WAAS is available outside North America

 

[GadgetGuy]

Re: Battery Charge Count: I’ve noticed this as well early in my ownership of the M2. I now believe the battery charge counter only increments up a number when the battery has accumulated 100% worth of charge. As this is the meaningful number when it comes to Lithium ion batteries life span. Thus, if you were to fly and burn 30% of the battery, recharge, then fly again burning 20% of your battery, recharge, then fly a 3rd time consuming 65% of your battery. When you’re done recharging after this 3rd flight, your charge count will read as “1” as the 3 recharges you’ve conducted sum up to 115% of your battery.

Makes sense, as each partial charge is cumulative towards 100%. I haven't paid attention to it myself, as all my uses are to below 10% remaining, where I would only see a difference of one charge count decrease every 10 flights.

 

[GadgetGuy]

I don't think WAAS is available outside North America

Confirmed. From Garmin: "WAAS satellite coverage is only available in North America."

 

[sar104]

I don't think WAAS is available outside North America

Confirmed. From Garmin: "WAAS satellite coverage is only available in North America."

That's correct - I just don't remember where @gnirtS is. That said, when I was recently over in the UK the GPS altitude data were also accurate to within 5 or 10 meters.

 

[axident]

Re: Battery Charge Count: I’ve noticed this as well early in my ownership of the M2. I now believe the battery charge counter only increments up a number when the battery has accumulated 100% worth of charge. As this is the meaningful number when it comes to Lithium ion batteries life span. Thus, if you were to fly and burn 30% of the battery, recharge, then fly again burning 20% of your battery, recharge, then fly a 3rd time consuming 65% of your battery. When you’re done recharging after this 3rd flight, your charge count will read as “1” as the 3 recharges you’ve conducted sum up to 115% of your battery.

Thanks for your explanation - I will watch this carefully after a few charges.

 

[gnirtS]

That's correct - I just don't remember where @gnirtS is. That said, when I was recently over in the UK the GPS altitude data were also accurate to within 5 or 10 meters.

As above, no WAAS but horizontal error (and related gdop) is often quoted at 10ft or so maximum. I suspect its more of an error due to geoid distortion as opposed to anything else. Its simple not as accurate in some areas over others.
My particular home location the ASL altitude is known with precision purely because it was a survey point, properly marked and measured prior to the building of this street so high degree of confidence in that measurement.

Im also based in various asian countries and its quite common to take off at or very near sea level and have altitudes showing 20m or more on GPS alt. Mavic and M2 likewise. Even accounting for AMSL variations and so on its far higher than 1.5x horizontal. As i said, i suspect its geoid variation not satellite variation causing it.

 

[sar104]

As above, no WAAS but horizontal error (and related gdop) is often quoted at 10ft or so maximum. I suspect its more of an error due to geoid distortion as opposed to anything else. Its simple not as accurate in some areas over others.
My particular home location the ASL altitude is known with precision purely because it was a survey point, properly marked and measured prior to the building of this street so high degree of confidence in that measurement.

Im also based in various asian countries and its quite common to take off at or very near sea level and have altitudes showing 20m or more on GPS alt. Mavic and M2 likewise. Even accounting for AMSL variations and so on its far higher than 1.5x horizontal. As i said, i suspect its geoid variation not satellite variation causing it.

If you are seeing a systematic error rather than a random error then that's a completely different issue that has nothing to do with VDOP. The GPS receivers use a reference geoid (EGM96 I think) that should be better than that, but I guess there could be problems in some areas.