Is there any way to embed the Barometric altimeter readings in  photographs?  The reason I ask is many times it is helpful or imperative  (in the case of measuring the height of something) to know the real  height above ground.

A little explanation here for non-real  airplane crowd.  When you view the GPS data attached to a photograph you  will see the latitude, longitude and altitude.  X, Y, and Z.  These are  dreived from the GPS satellites in our skies.  By the geometry of the  system, the X and Y positions are more accurate than the Z  altitude.   Many times I see GPS altitudes of -25 meters when I know the bird is up  at 100 feet or so.

The barometric altitude is derived by the DJI  bird based on air pressure off the prevailing recorded pressure at  pre-flight.  This is also what the bird uses to maintain hover and soft  landing.  (amoung other things)
As things are now, if I am making a  height measurement, I must snap a screenshot on my phone or tablet to  record the actual altitude above ground.  Looking at embedded GPS data  with the pictures yields wildly varying altitudes.

So the question is, is there a more straightforward method?

Thanks all in advance.

If you shoot in RAW then it uses the barometer in the EXIF data.

You are right.
GPS altitude data is really, really bad and of no use to anyone.
Why DJI decided to replace barometer data in the Exif info after firmware version 1.4 is a total mystery.
I have heard that the barometer data is still hidden deeper in the exif info but can't confirm as I have stayed at 1.4 so my photogrammetry software still works.

DJI-Ken Posted at 2016-7-6 08:06
If you shoot in RAW then it uses the barometer in the EXIF data.

Ken ... there is no good reason for DJI to have changed this.
As kirby says "Looking at embedded GPS data  with the pictures yields wildly varying altitudes."
GPS altitude data is completely useless.
And shooting Raw isn't the solution for photogrammetry, which is the primary use of altitude data anyway.
Can you bring this up with DJI tech people?

Just a note on Baro altitude. It varies with the environment. So if the temperature changes the barometric pressure will change, and thus the baro altitude will change as well. From the piloting rule of thumb a 1 in of Hg is approximately 1,000 ft.

For GPS, it normally uses WGS84 (most likely) and that stays constant no matter what the outside environment is doing. The reason why you are seeing negative numbers is because our world is not really a sphere. So there are points that are lower than what the average altitude would be.  The problem with GPS is the accuracy of values and the fact GPS has drift to it as well. Most of the time they are specified to be within +/-50 ft for X and Y. For The Z the spec is normally like 1.5 times the X and Y. So up to +/-75 ft in Z.

It is normally better to fly under a baro reference point up and away (most of the time it's > 1,000 ft) and a GPS altitude when low to the ground. The reason why I think DJI flies baro is it's easier to get a good reading of pressure altitude as you don't need to wait to for good satellite coverage, that and the fact they don't expect the pressure to change much in less than 30 mins.

DerikA Posted at 2016-7-6 10:29
Just a note on Baro altitude. It varies with the environment. So if the temperature changes the baro ...

Barometric altitude doesn't change much over the length  of a 20 minute Phantom flight and if you shoot say 100 photos for a photogrammetry job, flown at a fixed altitude, they exif data shows a similar altitude for all of the images.
But GPS altitude is very, very inaccurate and swings a lot so you can get +100 feet to -100 feet and more over a 20 minute flight which makes the data 100% useless to anyone.
The reason why DJI fly with barometer altitude data is that it is  accurate to a foot ot so and stable while GPS altitude can be a couple  of hundred feet out and varying a lot over just a few minutes, making vertical position holding impossible.

Here's what Garmin have to say about GPS altitude accuracy.  Note the last two sentences.

        How accurate is the GPS elevation reading?

GPS heights are based on an ellipsoid (a mathematical representation of the earth's shape), while USGS map elevations are based on a vertical datum tied to the geoid (or what is commonly called mean sea level). Basically, these are two different systems, although they have a relationship that has been modeled.

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.

I guess I forgot about one thing for the GPS stuff close to the ground is the fact most systems have a radar altimeter aiding it.

And I totally agree that baro altitude won't change much (normally) over a 20 min period.

labroides@yahoo Posted at 2016-7-6 06:48
Ken ... there is no good reason for DJI to have changed this.
As kirby says "Looking at embedded GP ...

It's been brought up before and that's the way they want it.

DerikA Posted at 2016-7-6 08:29
Just a note on Baro altitude. It varies with the environment. So if the temperature changes the baro ...

Very good description, thanks.

labroides@yahoo Posted at 2016-7-6 08:45
Barometric altitude doesn't change much over the length  of a 20 minute Phantom flight and if you s ...

I too would like JPEG with BARO, but it is what it is.
Sorry about that.

DJI-Ken Posted at 2016-7-6 11:55
I too would like JPEG with BARO, but it is what it is.
Sorry about that.

But it isn't what it was.

There can't be any possible reason for the change.
It has to be a mistake.
The GPS data is completely useless to everyone.
It's hard to see why they crippled a perfectly good photogrammetry machine for no reason.

I am wondering if there is something in the flight log data that indicates when a photo was taken? If so, you would be able to find the corresponding altitude reading.  Not all that straightforward, but if it does record it, all the info would be contained in a single csv file that can be opened as a spreadsheet.  If it is not recorded in the flight data, you could still refer the time the photo was taken to the spreadsheet to discover the altitude at that time.

This is the flight data in Healthy Drones, with photo data extracted from the data.  Not sure where the recorded altitude comes from, but the csv contains data to 7 decimal places and the values seem to update often.  It does seem to show the home point elevation being relative to sea level.



Looking at the csv data, it does have a column marked isPhoto.  There is a 1 recorded when a photo is taken, that row also contains the ASL altitude data, but you will have to subtract the (home) starting value to get AGL altitude.



Using Phantom Help, the csv contains altitude data to 1 decimal place, but home point starts at 0.  Unfortunately neither have a column marked BAROMETRIC, so use your best guess as to where the data comes from.

endotherm Posted at 2016-7-6 05:55
I am wondering if there is something in the flight log data that indicates when a photo was taken? I ...

Thanks everyone.  The shooting in raw is worth a look.  I am fairly familiar with EXIF formats.  My Nikon and Canons embed GPS data and loads of other data in photos.  I will try to cross breed some of that editing to the DJI

endotherm Posted at 2016-7-6 19:55
I am wondering if there is something in the flight log data that indicates when a photo was taken? I ...

' There is a 1 recorded when a photo is taken, that row also contains the ASL altitude data, but you will have to subtract the (home) starting value to get AGL altitude."
But because of the always changing inaccuracies of GPS altitude data and their magnitude, none will be of any use.

labroides@yahoo Posted at 2016-7-7 13:50
' There is a 1 recorded when a photo is taken, that row also contains the ASL altitude data, but y ...

I thought one of the log files (FLYXXX.DAT perhaps?) recorded every sensor and fart and twitch that the aircraft made.   Surely the BARO data is in there somewhere?  I'm not sure what they can extract from the .txt file as the resultant .csv from Healthy Drones and Phantom Help are different -- I guess they just extract and display what they think is relevant in their own format.  They are probably only provided GPS ALT data in the .txt file but that's only an assumption, is there any way of actually confirming that it is from the GPS and not the barometer?  The point of the exercise for me was to try and locate the barometric data hidden away in the logs.

edit:
Just for completeness, I examined one of the .dat files contained in the aircraft and decoded it using the DJI log converter at https://www.djilogs.com.  This gives you a preview of a few lines from the file, or else pay $1 for each record.

It did contain both the GPS altitude and Barometric altitude.  For info and comparison I post them here:

  

DATA.heightGps	DATA.heightBaro
0	0
0	0
0	0
0	0
0	0
0	0
0	0
0	0
0	0
0	0
0	0
0	0
0	0
0	0
0	0
0	0
0	0
0	0
0	0
0	0
0	0
0	0
23.11718369	0
22.5156517	0
21.96697807	0
22.8444252	0
21.95349503	0
22.68781853	0
23.00413895	0
22.6452961	0
22.83197975	0
22.07899857	0
21.44627762	0
22.75212097	0
22.9668045	0
22.71270943	0
23.82652092	0
22.68263245	0
23.42415428	0
23.60667419	0
22.74589729	0
22.86516762	0
22.83197975	0
22.41608429	0
22.88798332	0
23.82029915	0
23.28207779	0
22.28540039	0
23.0902195	0
22.28021431	0
23.62222862	24.17038918
23.3982296	24.18822098
23.57141495	24.19463158
23.62119293	24.18015862
23.34533882	24.1717701
23.07673645	24.16424751
23.86592674	24.18070602
23.93229485	24.17997169
22.670187	24.18310547
22.88279915	24.18522072
23.0871067	24.17451859
22.63285065	24.1725769
23.10473824	24.17926407
23.33289337	24.17469025
23.60874748	24.17081261
23.41171074	24.1961441
24.08473015	24.18724442
23.93436813	24.18028831
23.35674667	24.16545868
23.21466827	24.16228104
23.75082016	24.17941284
23.67096901	24.20496941
22.75108337	24.20908546
23.33185768	24.20987701
24.27034569	24.21059799
23.50504494	24.19935417
23.831707	24.19470024
23.14622116	24.20038033
23.31422615	24.1972599
22.92324638	24.19983292
23.4666748	24.19483566
22.93257904	24.18430138
23.00828743	24.18361282
23.72696877	24.18205452
23.42415428	24.19314575
23.15970421	24.1989193
23.72178268	24.18865967
23.9416275	24.1901207
23.28311539	24.19322205
23.5828228	24.19007301
23.38785744	24.19470215
23.37126541	24.19930077
23.82652092	24.2029686
23.43141365	24.20367622
23.26133728	24.2042923
23.4666748	24.19971275
23.17940712	24.19816589
23.21777916	24.1859417
23.89703751	24.17720795
23.50815582	24.17300034
22.71063614	24.72836304
23.31630135	24.7280426
23.90325928	24.7239399
23.22503853	24.73405266
23.34741211	24.739254
23.34015274	24.76031876
22.93983841	24.76620102
23.5361557	24.80589294
23.03006744	24.83077621
22.37874603	24.84324455
23.59008217	24.8859024
23.91051865	24.91109467
23.89599991	24.91462517
22.62973785	24.91393089
23.48534012	24.91386795
23.59008217	24.91909599
23.32044983	24.9224205
23.29452324	24.92528343
23.15659142	24.9237957
23.02177048	24.92115974
22.72308159	24.9192791
23.67822838	24.91665268
23.16385078	24.91689873
23.3733387	24.91738319
23.7414856	24.91590309
24.59905243	24.91670799
22.53950691	24.92029381
23.64193344	24.91719055
23.51230431	24.91479874
23.09333038	24.91352844
23.68341446	24.9143219
23.0632534	24.91337585
23.1617775	24.91376114
23.49363709	24.91344643
23.45837784	24.91096687
23.28104019	24.91199493
22.47831345	24.91278648
22.85375977	24.91167259
23.82029915	24.91269302
23.43452644	24.91530418
23.29659653	24.90999031
23.02280808	24.90893555
24.0079937	24.90733528
22.73760033	24.90922928
23.07880974	24.9175911
23.40548897	24.92869949
22.98547173	24.94246674
23.71037674	24.9536705
23.56934166	24.95451927
22.53950691	24.95412827

Now I just have to find a corresponding entry in the txt file and determine whether the csv contains the baro data or the gps data.