rpmurph83 Posted at 2016-8-23 20:39
The other thing I should mention that in reality is a more serious problems is all 4 brand new Phantom 4's would not hold their height usually when under 10 feet off the ground. If you take off a few feet and then let go over the controls, all four drones would hover in position, but would slowly drop to several inches off the ground.

Maybe you live in Colorado and I'm off base, but I'm pretty sure my GPS reads in feet above the GEOID (or orthometric height, this is a survey term, and this is what GPS uses as the "reference" elevation).  This is not, but can be compared to, mean sea level. I'd say you're just fine, I'm actually surprised you returned three of these for that. At least, this is my thought as to why your reading says 22 feet when you're only 1' off the ground (as indicated by your downward sensor).

Note that GPS elevations are notoriously inaccurate.  I am a geomatics professional and can say that even on our survey grade equipment the vertical accuracy is about 3 times that of our horizontal (usually in the 1cm range, mapping grade is usually 10m±).  The hardware we have in these drones is not capable of even coming close to that level of accuracy.  I'm currently working on a test project for a thesis paper and am using RTN-GPS ground control (around 1cm precision) with my Phantom 4 aerial photography to produce orthomosiacs.  I'm using Photoscan and my raw horizontal solutions are in the 10-feet range until I add the ground control to the solution.  The vertical is close to three times that.  This is to be expected.  The barometer is accurate "relative" to the ground, but may not give an exact elevation but unless you have some serious weather situation that dramatically alters the barometric pressure in a few minutes you're fine.  But I'd imagine a weather event that would do that would be dangerous to fly in.

Note that only at exactly elevation 0.0 (near mean sea level) will you get a 0.0 when on the ground.  Your distance sensors will display that, but GPS displays your orthometric height above the geoid.

I agree that the height measurement of the P4 seems to be inaccurate. This week ik flew a litchi waypoint mission at 5 m (~2 feet). It drifted slowly towards the ground to less than 2 m. I had to restart the drone a few times to remain the proper height. I blamed it on the weather. A shower was coming.

It was inaccurate on all 3 of my p3's. my p4 and now my p4pro.
I have to fly a certain amount of feet off the ground for doing surveys and by the time I am done with 1 section the altitude ranges from 6ft to 15ft when on the ground.
It does help to land power-off the motors and then take off again, doing this several times give me some idea that I am at the altitude I want to be at.
This the only way I found to do it.
Yes it is a problem especially if you are counting on way points and height.
Yes on an airplane there is also this disparity in height but I don't recall it being anywhere close to what I am seeing on Phantoms.
when we are moving into more and more work that is measurement based instead of photography I am concerned.
Ill see what happens when my Inspire 2 comes in.not expecting any difference.

Not sure if this post/thread is still being checked in on, but I've got a new scenario to throw in.  I've got a P4 that has been giving me "NA" readings for VPS height; either that or it will say "0.3ft" or "1ft".  The error occurred whenever I went into a hover, no matter the height (2 feet, 20 feet, 200 feet, all would say "NA" or something less than 1 foot).  I tried recalibrating the downward facing vision cameras using DJI's Assistant program, and noticed that my camera gimbal was showing up on the downward facing cameras.  While holding the P4 off the ground, I could get it it to give me accurate height readings if I pushed the camera gimbal forward a quarter of an inch.  I thought that this must be the problem, but after fiddling with the gimbal position and the angle of the VPS camera sensors inside the body of the P4, no luck.  So then I tried covering the ultrasonic sensors with a piece of tape, and lo and behold I had VPS accuracy again, anywhere from roughly 4 feet off the ground and higher (I haven't flown it around yet, but from holding and lifting the P4 off the ground at various heights, the VPS would accurately display my "hover" height).  I now believe the VPS cameras were working fine all along, but it was my ultrasonic sensors that were getting confused by the camera gimbal in front of them.  What doesn't make any sense to me is that they had been working correctly before, and I've had no crashes or hard landings, and yet something changed in positioning to cause this error.  

Has anyone else experienced issues of getting either a VPS reading of "NA" or something lower than 1 foot, no matter the actual height of the drone?  Does anyone have any ideas on how to correct a malfunction involving just the ultrasonic sensors?  I've looked carefully at them and cannot see any obstruction or debris around the ultrasonic sensors, so I'm unsure as to why it suddenly began giving consistently false height information.

Hansogren 9 Posted at 2017-7-25 15:45
Not sure if this post/thread is still being checked in on, but I've got a new scenario to throw in.  I've got a P4 that has been giving me "NA" readings for VPS height; either that or it will say "0.3ft" or "1ft".  The error occurred whenever I went into a hover, no matter the height (2 feet, 20 feet, 200 feet, all would say "NA" or something less than 1 foot).  I tried recalibrating the downward facing vision cameras using DJI's Assistant program, and noticed that my camera gimbal was showing up on the downward facing cameras.  While holding the P4 off the ground, I could get it it to give me accurate height readings if I pushed the camera gimbal forward a quarter of an inch.  I thought that this must be the problem, but after fiddling with the gimbal position and the angle of the VPS camera sensors inside the body of the P4, no luck.  So then I tried covering the ultrasonic sensors with a piece of tape, and lo and behold I had VPS accuracy again, anywhere from roughly 4 feet off the ground and higher (I haven't flown it around yet, but from holding and lifting the P4 off the ground at various heights, the VPS would accurately display my "hover" height).  I now believe the VPS cameras were working fine all along, but it was my ultrasonic sensors that were getting confused by the camera gimbal in front of them.  What doesn't make any sense to me is that they had been working correctly before, and I've had no crashes or hard landings, and yet something changed in positioning to cause this error.  

Has anyone else experienced issues of getting either a VPS reading of "NA" or something lower than 1 foot, no matter the actual height of the drone?  Does anyone have any ideas on how to correct a malfunction involving just the ultrasonic sensors?  I've looked carefully at them and cannot see any obstruction or debris around the ultrasonic sensors, so I'm unsure as to why it suddenly began giving consistently false height information.

Have you tried calibrating the VPS in DJI Assistant 2 ?

If not then that would likely sort things out.

I tried that multiple times without any change to the VPS "NA" or height error message happening.  It was only once I covered the ultrasonic sensors over that I was able to get correct height readings.  I took it out yesterday for some test flights with the ultrasonic sensors covered up, and the VPS downward facing cameras worked correctly (reading my height anywhere from roughly 2 feet to 35 feet off the ground).  I've got a new problem today though, as my compass 2 has supposedly disconnected (so says the app).  I'm not sure I believe it, but haven't been able to figure that problem out (I've checked the wiring connection points for it and everything looks fine, no crimps or broken wiring).  I've just tried deleting the app and reinstalling it, wondering if it was a software bug, but no luck yet.

I recently crashed my P4P while running a Dronedeploy mission. The mission altitude was set to 60' but twice in the mission the flight log shows 43' resulting in a crash into a tree.

fans99a8dcf2 Posted at 2017-7-29 15:19
I recently crashed my P4P while running a Dronedeploy mission. The mission altitude was set to 60' but twice in the mission the flight log shows 43' resulting in a crash into a tree.

"The mission altitude was set to 60' but twice in the mission the flight log shows 43' resulting in a crash into a tree. "

Does that not then suggest a problem with Dronedeploy or its setting ?

If the sensors were logging 43 feet then that would seem right given your description.

Hans Ogren Posted at 2017-7-26 09:45
I tried that multiple times without any change to the VPS "NA" or height error message happening.  It was only once I covered the ultrasonic sensors over that I was able to get correct height readings.  I took it out yesterday for some test flights with the ultrasonic sensors covered up, and the VPS downward facing cameras worked correctly (reading my height anywhere from roughly 2 feet to 35 feet off the ground).  I've got a new problem today though, as my compass 2 has supposedly disconnected (so says the app).  I'm not sure I believe it, but haven't been able to figure that problem out (I've checked the wiring connection points for it and everything looks fine, no crimps or broken wiring).  I've just tried deleting the app and reinstalling it, wondering if it was a software bug, but no luck yet.

Have you tried a refresh and factory reset of aircraft using DJI Assistant 2 ? There may be a glitch in the firmware. Only times I've read of compass disconnects is where they have been faulty. Has it had any hard landings or crashes ?

DJI-Paladin Posted at 2016-8-23 20:17
Hello rpmurph83 , barometer performance can be influenced by air pressure and air flow. Also, the altitude number will change when the reference changes.But you don't need to worry about it too much if your drone can hold its attitude and altitude well in the air. In low altitude above the ground, the performance of ultrasonic unit would be better in detecting the height.

You cannot say this. The data acquired for Land mapping and other surveys gets waste due to an incorrect altitude calculated by Compass.

I have asked the similar question under my account - https://forum.dji.com/thread-182836-1-1.html.

hardrone Posted at 3-5 01:51
You cannot say this. The data acquired for Land mapping and other surveys gets waste due to an incorrect altitude calculated by Compass.

I have asked the similar question under my account - https://forum.dji.com/thread-182836-1-1.html.

'The data acquired for Land mapping and other surveys gets waste due to an incorrect altitude calculated by Compass. '

The altitude is not calculated by the compass, the compass has nothing to do with altitude at all.

Geebax Posted at 3-5 18:36
'The data acquired for Land mapping and other surveys gets waste due to an incorrect altitude calculated by Compass. '

The altitude is not calculated by the compass, the compass has nothing to do with altitude at all.

Under any condition, for what I have said this -- It does not make much difference if your drone holds the altitude. For this point, I have opposed and said that.

What I know is, along side barometers, even GPS is used for altitude calculations.

Geebax Posted at 3-5 18:36
'The data acquired for Land mapping and other surveys gets waste due to an incorrect altitude calculated by Compass. '

The altitude is not calculated by the compass, the compass has nothing to do with altitude at all.

I am more interested in getting a solution in the response rather than having an objection to my quick words.

My question, if I have not shared, is - https://forum.dji.com/thread-182836-1-1.html

Hope you will come back with a solution

hardrone Posted at 3-5 21:35
Under any condition, for what I have said this -- It does not make much difference if your drone holds the altitude. For this point, I have opposed and said that.

What I know is, along side barometers, even GPS is used for altitude calculations.

No, GPS is not used for altitude calulations either, it is way too innacurate. The ultrasonic transducers on the underneath of the aircraft are used to measure altitude when within the range of 0 to 3 Metres, above that, only the barometic transducer is used to measure altitude. The simple fact is that there is no economic method of altitude measurement that is able to be used. A radar measurement system would be more accurate up to a point, but is too expensive, too heavy and uses too much electrical power to be practical.

What is wrong here is not the measurement system, but people's expectations about the accuracy of height measurement. It is not an exact science, and accurate height measuring systems are outside the scope of consumer devices like the DJI drones. If the aircraft is not accurate enough for the surveying tasks you are speaking of, then you are using the wrong tool for the job.

Geebax Posted at 3-6 00:03
No, GPS is not used for altitude calulations either, it is way too innacurate. The ultrasonic transducers on the underneath of the aircraft are used to measure altitude when within the range of 0 to 3 Metres, above that, only the barometic transducer is used to measure altitude. The simple fact is that there is no economic method of altitude measurement that is able to be used. A radar measurement system would be more accurate up to a point, but is too expensive, too heavy and uses too much electrical power to be practical.

What is wrong here is not the measurement system, but people's expectations about the accuracy of height measurement. It is not an exact science, and accurate height measuring systems are outside the scope of consumer devices like the DJI drones. If the aircraft is not accurate enough for the surveying tasks you are speaking of, then you are using the wrong tool for the job.

Now, here the issue is, almost all the drones use Barometric transducer system for altitude measurement.

Coming to the Survey thing, the drone used to give precise altitude since purchase date. Its just from the last 2 to 3 missions  have been observing that the Drone altitude in set altitude calculation and the exif I see, differs. It used to have a very less difference in terms of cm.

Geebax Posted at 3-6 00:03
No, GPS is not used for altitude calulations either, it is way too innacurate. The ultrasonic transducers on the underneath of the aircraft are used to measure altitude when within the range of 0 to 3 Metres, above that, only the barometic transducer is used to measure altitude. The simple fact is that there is no economic method of altitude measurement that is able to be used. A radar measurement system would be more accurate up to a point, but is too expensive, too heavy and uses too much electrical power to be practical.

What is wrong here is not the measurement system, but people's expectations about the accuracy of height measurement. It is not an exact science, and accurate height measuring systems are outside the scope of consumer devices like the DJI drones. If the aircraft is not accurate enough for the surveying tasks you are speaking of, then you are using the wrong tool for the job.

I do not think that if such issues being faced by the people, there won't be any solution to this.
I am doing surveying missions now. whereas people are doing this things for the last 3-4 years, at least, on DJI devices. So, something must be there.

hardrone Posted at 3-6 01:43
Now, here the issue is, almost all the drones use Barometric transducer system for altitude measurement.

Coming to the Survey thing, the drone used to give precise altitude since purchase date. Its just from the last 2 to 3 missions  have been observing that the Drone altitude in set altitude calculation and the exif I see, differs. It used to have a very less difference in terms of cm.

"the Drone altitude in set altitude calculation and the exif I see, differs. It used to have a very less difference in terms of cm. "

There have been many references to the difference between the barometric (feet above take off point) & GPS Exif readings (unit of measurement in metres above sea level).

There may be some information useful to you:-

https://forum.dji.com/thread-169225-1-1.html
https://forum.dji.com/thread-142010-1-1.html
https://forum.dji.com/thread-154027-1-1.html

Aardvark Posted at 3-6 02:30
"the Drone altitude in set altitude calculation and the exif I see, differs. It used to have a very less difference in terms of cm. "

There have been many references to the difference between the barometric (feet above take off point) & GPS Exif readings (unit of measurement in metres above sea level).

Thank you so much for spending out the time in sharing the links.

Though, I did not see much solvable to my issues.

Now, somewhere it talks about the unit which is set for altitude value. That is not an issue for me.

Another thing is, people have doubt about the EXIF calculation as it contains ASL data. That idea I already had so it is not the part of the doubt for me.

So, mainly my vision system is not calibrating after trying a lot and another thing is, I am not getting the correct EXIF calculation of what altitude I am practically flying at (along with adding msl of area to it)

I just hit a tree full throttle with my P4 screen read 41 metres height tree was no more than 25   WTF