From my early discovering of the way my P4P was doing it's stuff it has been my interpretation that Altitude in the EXIF information of the pictures was the height above sea level, had prefered it was displayed as altitude from takeoff point just like seen on the screen, but what ever if you know the takeoff altitude from sea it can be calculated anyway.

Resently i discovered that some of my latest pictures has some strange values in the EXIF information and thanks to the caption possibilities in the movies files i just resently discovered i has been able to compare that information against the EXIF values but i just can't figure out what is going on even after the use of several EXIF tools.

Can anybody please tke a look and help me to find the altitude from the EXIF itself or confirm that there is something strange with even my pictures or in general with the P4P pictures taken with latest firmwares. controller is GL300E but i think that does not matter as the pictures is taken and stored in the AC itself.
Is it windows that has start displaying information from a wrong field as being the altitude and from where comes the strange values of altitude above sea level from?

EDIT: Why is the GPS format different in captions compared to EXIF and why is it displayed as Long/lat, i thought the normal way would be Lat/Long but also the site where i found the height of location would have it as Long/Lat?

Takeoff point placed around 20 meters above sea, ground on location about same height.

Map with coordinates (www.krak.dk)

height above sea acording some geo map

1) screenshot from .mp4 along with a .jpg taken while the movies was recorded during horisontal (sideways) flight and some gps information extracted with other tools from EXIF:

information from captions, like seen in app during the flight

Windows EXIF information

EXIF information displayed in Gimp

Other tools:
North or South Latit|N
Latitude            |55, 55, 43,1382
East or West Longitu|E
Longitude           |11, 33, 30,0297
Altitude Reference  |Sea level reference
Altitude            |40,322

GPS Altitude Ref                : Below Sea Level
Absolute Altitude               : -40.32
Relative Altitude               : +61.80
GPS Altitude                    : 40.3 m Below Sea Level

2) same flight a little higher:

information from captions, like seen in app during the flight

Windows EXIF information

EXIF information displayed in Gimp

Other tools:
North or South Latit|N
Latitude            |55, 55, 43,0696
East or West Longitu|E
Longitude           |11, 33, 29,7230
Altitude Reference  |Sea level reference
Altitude            |1,822

GPS Altitude Ref                : Below Sea Level
Absolute Altitude               : -1.82
Relative Altitude               : +100.30
GPS Altitude                    : 1.8 m Below Sea Level

3) The day after, AC have had a factory reset when i discovered the above strange information, same takeoff point and same location.

information from captions, like seen in app during the flight

Windows EXIF information

EXIF information displayed in Gimp

Other tools:
North or South Latit|N
Latitude            |55, 55, 44,7958
East or West Longitu|E
Longitude           |11, 33, 29,2025
Altitude Reference  |Sea level
Altitude            |11,388

GPS Altitude Ref                : Above Sea Level
Absolute Altitude               : +11.39
Relative Altitude               : +75.10
GPS Altitude                    : 11.3 m Above Sea Level

Your Phantom uses its barometer for all flight functions but since back in the P3 days, DJI uses GPS for altitude in Exif data.
They also hide the barometer altitude in the XMP data inside the Exif info and some readers can display this.
GPS data (absolute altitude) is height above sea level.
Relative altitude is the barometer dtat - height above launch point.
BUT GPS is very inaccurate for altitude and can commonly show an error of +/- 60-80 metres which you can easily check by taking photos close to the sea and seeing altitudes that are well below sea level.
The GPS altitude error also changes over short time periods which makes it even more useless.

I can't find a reason for the Long/Lat format.  It's just the way it gets recorded.

I would agree with all that Labroids has said, flight data from barometer, EXIF from GPS/GLONASS.

Thank you both, it seem that windows is not able in anyway to retrieve the value from the "Relative Altitude" that other software can extract, i have now tried every documented GPS attribute in windows related to altitude without getting that value out of it. was a piece of cake to have it show the altitude doing a mouse hover on  the files but clearly that value is of no use.
I find it strange that the P4P GPS can't do better from the GPS information, i think my ordinary GPS units can do a way better measurement of the altitude, even the very old Garmin Street Pilot III.

It must have been a total coincidence that I have noticed it display a altitude of 120 meters, knowing that the images was taken at 100 meters and after that concluded that the information in windows told the altitude above sea level. Now noticed that this is far from reliable and my conclusion is that there is no reliable way to determine the actual altitude as also the barometer has its flaws in changing temperatures and barometric pressure changes.

KedDK Posted at 2018-5-13 15:13
Thank you both, it seem that windows is not able in anyway to retrieve the value from the "Relative Altitude" that other software can extract, i have now tried every documented GPS attribute in windows related to altitude without getting that value out of it. was a piece of cake to have it show the altitude doing a mouse hover on  the files but clearly that value is of no use.
I find it strange that the P4P GPS can't do better from the GPS information, i think my ordinary GPS units can do a way better measurement of the altitude, even the very old Garmin Street Pilot III.

I find it strange that the P4P GPS can't do better from the GPS information, i think my ordinary GPS units can do a way better measurement of the altitude, even the very old Garmin Street Pilot III.
Take your Garmin down to the sea and see how close it gets you to sea level.
Then do it an hour later or the next day.  
It's altitude reading will be all over the place.
I've been on a sailboat with my Garmin telling me it was a submarine 200 ft below.
That's why later models have a barometer to give usable altitude readings (unless you are in a pressurised airplane).

Garmin explain it this way:  The main source of error has to do with the arrangement of the satellite configurations during fix determinations. The earth blocks out satellites needed to get a good quality vertical measurement. Once the vertical datum is taken into account, the accuracy permitted by geometry considerations remains less than that of horizontal positions. It is not uncommon for satellite heights to be off from map elevations by +/- 400 ft. Use these values with caution when navigating.

my conclusion is that there is no reliable way to determine the actual altitude as also the barometer has its flaws in changing temperatures and barometric pressure changes.
You can add the relative altitude to the home point altitude.
The barometer changes very little during the short time a battery lasts and the error should be quite small.

Labroides Posted at 2018-5-13 16:12
I find it strange that the P4P GPS can't do better from the GPS information, i think my ordinary GPS units can do a way better measurement of the altitude, even the very old Garmin Street Pilot III.
Take your Garmin down to the sea and see how close it gets you to sea level.
Then do it an hour later or the next day.  

" barometer changes very little during the short time a battery lasts and the error should be quite small"
Once more we have to agree to disagree, everything is relative as it is said but i really would say a five meter of is quite a miss and far from specs. This is the P4P just above a lawn and just beside and below the takeoff/landing box after 23 minutes of flight.

Just a few days ago i had a flight some hours before a rain front was supposed to hit. Took off and after a short flight ascended to go free of a line of trees that normal would be passed at 20 meters. When i looked down after the pass to adjust the gimbal tilt the altitude was reported as being 60 meters, very far from the conditions and why i suspect that the atmospheric pressure can mess things up quite a lot.

All i wanted was a way to actually know the height, don't really matter if showed above sea or takeoff, but it seem that it is not easy, specially as the apparently most reliable source, the barometric information is stored in a way that is not easy read on windows without use of specific software.

Thanks again.

KedDK Posted at 2018-5-13 15:13
Thank you both, it seem that windows is not able in anyway to retrieve the value from the "Relative Altitude" that other software can extract, i have now tried every documented GPS attribute in windows related to altitude without getting that value out of it. was a piece of cake to have it show the altitude doing a mouse hover on  the files but clearly that value is of no use.
I find it strange that the P4P GPS can't do better from the GPS information, i think my ordinary GPS units can do a way better measurement of the altitude, even the very old Garmin Street Pilot III.

The GPS vertical position (Civilian band) can be +/- 4 times the horizontal error added to the SV’s via the GPS controllers in Colorado. (A old rule of thumb)

KedDK Posted at 2018-5-14 00:25
" barometer changes very little during the short time a battery lasts and the error should be quite small"
Once more we have to agree to disagree, everything is relative as it is said but i really would say a five meter of is quite a miss and far from specs. This is the P4P just above a lawn and just beside and below the takeoff/landing box after 23 minutes of flight.
[view_image]

Unless you fly around severe cold fronts the atmospheric pressure isn't going to change enough over a twenty minute flight to make any appreciable difference.
If you are saying that yours is out by 40 metres (?), it's seriously out of spec.

I'll happily accept a barometer that might drift a metre or two over a flight over a GPS that can swing  +/- 30 metres in a few minutes.

RMJovo Posted at 2018-5-14 03:57
The GPS vertical position (Civilian band) can be +/- 4 times the horizontal error added to the SV’s via the GPS controllers in Colorado. (A old rule of thumb)

I don't remember what is said to be the accuracy of the civil GPS system, think it is either ±2.5 or ±5 meters, this should then in worst case be within ±20meters of in altitude reported by the GPS information, i see around 100 meters off..
Strange enough the spec say that vertical hover precision is better that horizontal with ±0.5 versus ±1.5 meters

Labroides Posted at 2018-5-14 05:33
Unless you fly around severe cold fronts the atmospheric pressure isn't going to change enough over a twenty minute flight to make any appreciable difference.
If you are saying that yours is out by 40 metres (?), it's seriously out of spec.

I still think my GPS's seem do a decent work, admitted i didn't reach the beach to drop it at the waterline.

With that said i have two Quest2's, the other one was no good at all and changed a lot up and down all the time even both was put on the same place so there might be a big difference from item to item in this regard. My Street Pilot3 is shelfed on other location so that could not be tested right away.

For me to see the barometer in the phantoms seem to be too slow to adabt for changes in pressure and temp and might be better used in a weather station or other barometric display hanging on a wall than on a drone doing a lot of altitude changes, perhaps they have shut a short on the quality versus costs here, i don't know.
All i can say is that i see a lot of strange things with the altitude measurement both things that are off within a single flight and things that seem to differ from one day to the other.

KedDK Posted at 2018-5-20 15:34
I don't remember what is said to be the accuracy of the civil GPS system, think it is either ±2.5 or ±5 meters, this should then in worst case be within ±20meters of in altitude reported by the GPS information, i see around 100 meters off..
Strange enough the spec say that vertical hover precision is better that horizontal with ±0.5 versus ±1.5 meters

The civil band The US Government commits to the horizontal GPS signal in space with a global average of </= to 7.8m 95% of the time user range error (URE) however, the  real average of late has been on average 2.3 ft 95% of the time. As for the vertical accuracy it depends on the position of the SV’s the best config is one overhead with at lest 3 SV’s as close to the horizon as you can get. The vertical average is about 6m with times of 10-20m. All of this is depending upon how good your receiver is. Also what the manufacturer specifications on there receivers are. Please refer to GPS.GOV website.

KedDK Posted at 2018-5-20 15:34
I don't remember what is said to be the accuracy of the civil GPS system, think it is either ±2.5 or ±5 meters, this should then in worst case be within ±20meters of in altitude reported by the GPS information, i see around 100 meters off..
Strange enough the spec say that vertical hover precision is better that horizontal with ±0.5 versus ±1.5 meters

I don't remember what is said to be the accuracy of the civil GPS  system, think it is either ±2.5 or ±5 meters, this should then in worst  case be within ±20meters of in altitude reported by the GPS information,  i see around 100 meters off..
See what I quoted from Garmin back up in post #5.
GPS altitude can be +/- 200 feet
My observations confirm that errors like that are common.

Strange enough the spec say that vertical hover precision is better that horizontal with ±0.5 versus ±1.5 meters
That's because DJI use a barometer for altitude flight data and it's accurate to better than 1 foot
I still think my GPS's seem do a decent work, admitted i didn't reach the beach to drop it at the waterline.
Sometimes it will be close, other times it will be a long way off.
It all depends on the geometry of the spread of satellites you catch.
For me to see the barometer in the phantoms seem to be too slow to adabt for changes in pressure and temp
How much does your atmospheric temp and pressure vary in the time of a flight (20 mins) - not much.
Not enough to make any appreciable difference.

As for slow ... just watch your Phantom's altitude indicator as you climb.
It's not slow at all and will click over every foot you climb.

If you're getting dodgy altitude readings, try recalibrating the IMU.

Labroides Posted at 2018-5-20 16:24
I don't remember what is said to be the accuracy of the civil GPS  system, think it is either ±2.5 or ±5 meters, this should then in worst  case be within ±20meters of in altitude reported by the GPS information,  i see around 100 meters off..
See what I quoted from Garmin back up in post #5.
GPS altitude can be +/- 200 feet

You have to remember that the WGS 84 datum used for GPS  sea level surface is an oblate spheroid with a ellipsoid range from -105m to about +58m  so your GPS can read negative at sea level depending where on earth you are.

Accurate elevations are determined via
H=h-N
H=orthometric height
N=geoid height
h=ellipsoid heigh

Labroides Posted at 2018-5-13 04:40
Your Phantom uses its barometer for all flight functions but since back in the P3 days, DJI uses GPS for altitude in Exif data.
They also hide the barometer altitude in the XMP data inside the Exif info and some readers can display this.
GPS data (absolute altitude) is height above sea level.

"GPS data (absolute altitude) is height above sea level.
Relative altitude is the barometer dtat - height above launch point."

If I understand this correctly:
The height above homepoint which I'm seeing on my screen (not in exif data) is calculated by the barometer?

Labroides Posted at 2018-5-13 04:40
Your Phantom uses its barometer for all flight functions but since back in the P3 days, DJI uses GPS for altitude in Exif data.
They also hide the barometer altitude in the XMP data inside the Exif info and some readers can display this.
GPS data (absolute altitude) is height above sea level.

Lon/Lat is X/Y. (Yes, they are spherical coordinates, not cartesian but in terms of axes that is why they are in that order)

patiam Posted at 2018-6-14 09:11
Lon/Lat is X/Y. (Yes, they are spherical coordinates, not cartesian but in terms of axes that is why they are in that order)

Is it only used in that order in certain uses then or ?
As it is shown in the OP almost everywhere else the Latitude would be the first listed, same goes for other systems i have used. When not consequent used this could lead to some serious misplacement if interpreted the wrong way.

KedDK Posted at 2018-6-14 23:17
Is it only used in that order in certain uses then or ?
As it is shown in the OP almost everywhere else the Latitude would be the first listed, same goes for other systems i have used. When not consequent used this could lead to some serious misplacement if interpreted the wrong way.

The exif position data is saved as Long/Lat.
It took me a while to work out why when I entered the data for ground control points in a mapping project, I got an error of 4000+ kilometres.
Once I changed the GCP data to Long/Lat everything came together.

KedDK Posted at 2018-6-14 23:17
Is it only used in that order in certain uses then or ?
As it is shown in the OP almost everywhere else the Latitude would be the first listed, same goes for other systems i have used. When not consequent used this could lead to some serious misplacement if interpreted the wrong way.

Lon/Lat is very common in mapping and GIS, probably more common that Lat/Lon, which most other folks are more accustomed to simply by nature of it being how it is often spoken.

Depending on where you are in the world, which value is Lon may or may not be unambiguous... If it is over 90° it can't be Lat.

Confusion can also occur regarding the sign ± of Lon, which is used to denote on which side of the PM the location is, rather than alpha E/W.

Labroides Posted at 2018-5-13 04:40
Your Phantom uses its barometer for all flight functions but since back in the P3 days, DJI uses GPS for altitude in Exif data.
They also hide the barometer altitude in the XMP data inside the Exif info and some readers can display this.
GPS data (absolute altitude) is height above sea level.

Waking up an old thread ;-)
There is this thread about GPS/Barometer and how the Mavics read altitude at Mavicpilots.
How can we find out if or which DJI drones use GPS or barometer?
See post #23:
https://mavicpilots.com/threads/how-does-the-mavic-measure-it%E2%80%99s-height.44525/page-2#post-561901

RMJovo Posted at 2018-5-14 03:57
The GPS vertical position (Civilian band) can be +/- 4 times the horizontal error added to the SV’s via the GPS controllers in Colorado. (A old rule of thumb)

GPS operators are not allowed to introduce errors into Sat data used to determine position except in times and areas of conflict / war / grave necessity.

SA (Selective Availability) was switched off by Presidential decree many years ago. It was basically a worthless 'instrument' because of the use of Differential GPS to overcome the effects of SA.

OK - lets move onto barometer use for altitude in many units as well as DJI. It is not the actual reading itself that determines altitude ... it is the CHANGE of reading that determines CHANGE of altitude based on the average constant for pressure change per height change.
In effect - home point before take-off is regarded as its zero mark ... and then any increase / decrease as model rises / drops causes the altitude displayed to change.
It would take a very quick change in weather pattern ... such as a strong narrow front to pass through to give significant error after take-off ...
It would be assumed that any sensible pilot is not going out there when a serious front is imminent ... the winds and possibility of precipitation would be enough to say ... hang on till after !

Nigel

Eric13 Posted at 2018-10-1 04:54
Waking up an old thread ;-)
There is this thread about GPS/Barometer and how the Mavics read altitude at Mavicpilots.
How can we find out if or which DJI drones use GPS or barometer?

How can we find out if or which DJI drones use GPS or barometer?
All DJI drones and most airplanes of any kind use a barometer for flight altitude.
The post you refer to is from a guy who was mistaken.

RMJovo Posted at 2018-5-20 16:57
You have to remember that the WGS 84 datum used for GPS  sea level surface is an oblate spheroid with a ellipsoid range from -105m to about +58m  so your GPS can read negative at sea level depending where on earth you are.

Accurate elevations are determined via

The Oblate Spheroid basis for earth shape has been progressively attacked by later determinations, because of GPS. But it is still the most practical representation.

First we had the inaccuracies of  mapping / charts and the need to note in Titles the correction to apply to land masses / objects to bring into agreement with GPS data. Mostly based on the methods used by such as Royal Navy officers and their sextants !! Brilliant deductions in their time, but as we know now with errors.
Then we had the shape of the earth and its assumed equatorial band being reasonably uniform radius from the centre- which it is not.

The world of geographical determination is continually evolving and GPS / Satelite mapping is changing it all. WGS84 is now an old established format ... new ones are out there but not widely used or known.

Just thought I'd throw it in the discussion !!

Nigel

solentlife Posted at 2018-10-1 05:10
GPS operators are not allowed to introduce errors into Sat data used to determine position except in times and areas of conflict / war / grave necessity.

SA (Selective Availability) was switched off by Presidential decree many years ago. It was basically a worthless 'instrument' because of the use of Differential GPS to overcome the effects of SA.

The +/- 4 times is a old rule of thumb, the least amount of error in altitude is when there are a number of SV’s (GPS sat) at or just above the horizon of earth. Even with the SA gone, the altitude will drift depending  on position of the SV’s, the receiver being using for the position  fix and the atmospheric conditions at the time. You may find more info at www.gps.gov

RMJovo Posted at 2018-10-1 12:17
The +/- 4 times is a old rule of thumb, the least amount of error in altitude is when there are a number of SV’s (GPS sat) at or just above the horizon of earth. Even with the SA gone, the altitude will drift depending  on position of the SV’s, the receiver being using for the position  fix and the atmospheric conditions at the time. You may find more info at www.gps.gov

Fundamental error there .... Sats that are at low to the horizon are automatically rejected by sat rcvrs ... due to the too large errors inherent in the calculations resulting from their data.

Before my move into Fuel and associated 'biz' - my professional career was Navigator and I came up through conventional Sextant ... Transit Sat Nav ... and GPS. I also worked Doppler, Omega, Decca, Loran A and C ... to name just a few.

I have worked in my present work with Trimble, Garmin, Lowrance, Raytheon and many military / civilian units.

I may have a bit of understanding of their accuracy and use .... I don't know all the answers, but I know where to find most ...

Nigel

solentlife Posted at 2018-10-1 12:31
Fundamental error there .... Sats that are at low to the horizon are automatically rejected by sat rcvrs ... due to the too large errors inherent in the calculations resulting from their data.

Before my move into Fuel and associated 'biz' - my professional career was Navigator and I came up through conventional Sextant ... Transit Sat Nav ... and GPS. I also worked Doppler, Omega, Decca, Loran A and C ... to name just a few.

Your work experience is excellent Nigel, I to have been around since GPS’s conception also. As you would know the low SV’s would give you theoretically the best results for altitude however, real world conditions are different then the mathematical calculations. Take Care and Fly Safe.