Mavic 2 pro flight log data ( firmware v01.00.04.00 ) - height / altitude info

[new2mavic]

I have spent some time on comparing various height / altitude related data fields in the .TXT and .DAT flight log data files of my M2P. As far as I can see from this flight, the use of sensor fusion algorithm to derive the height / altitude info is not that obvious. It seems that the height displayed on DJI GO 4 is derived from barometer data and the VPS value from the sensors at the belly. The original .TXT and .DAT files are also attached.

This is the screen capture of the flight:

To view this content we will need your consent to set third party cookies.
For more detailed information, see our cookies page.

Accept third party cookies

Comments are more than welcome .

 

[sar104]

The IMU height is computed from the inertial sensors and corrected for drift and bias by the barometric data. The VPS height is not used in that computation at all for obvious reasons - that it is not measuring height above the home point.

 

[BudWalker]

I have spent some time on comparing various height / altitude related data fields in the .TXT and .DAT flight log data files of my M2P. As far as I can see from this flight, the use of sensor fusion algorithm to derive the height / altitude info is not that obvious. It seems that the height displayed on DJI GO 4 is derived from barometer data and the VPS value from the sensors at the belly. The original .TXT and .DAT files are also attached.

This is the screen capture of the flight:

To view this content we will need your consent to set third party cookies.
For more detailed information, see our cookies page.

Accept third party cookies

Comments are more than welcome .
View attachment 91034
View attachment 91035
View attachment 91036
View attachment 91044

The .DAT file has a record type that makes it way into the .txt file pretty much intact. This record type contains the height data so it's not surprising that General:relativeHeight and OSD:height match. They may not match exactly due to data being sampled at slightly different times.

I had always thought that height data was just determined by the barometric data. But, it appears that the IMU:velZ data is being used to correct the height data. This can be seen from this post
A Legitimate Near-Flyaway with My Fourth M2Z

EDIT: @sar104 and I were posting the same thing but he hit the post button before me. Seems to be a regular occurrence.

 

[sar104]

The .DAT file has a record type that makes it way into the .txt file pretty much intact. This record type contains the height data so it's not surprising that General:relativeHeight and OSD:height match. They may not match exactly due to data being sampled at slightly different times.

I had always thought that height data was just determined by the barometric data. But, it appears that the IMU:velZ data is being used to correct the height data. This can be seen from this post
A Legitimate Near-Flyaway with My Fourth M2Z

EDIT: @sar104 and I were posting the same thing but he hit the post button before me. Seems to be a regular occurrence.

I've been taking speed typing classes for just this reason.

 

[new2mavic]

....I had always thought that height data was just determined by the barometric data. But, it appears that the IMU:velZ data is being used to correct the height data....

OK, I am bored today so I decided to do a little experiment. I put my M2P in a trash bag and connect the mouth to my vacuum cleaner . When the vac cleaner is turned on, the pressure inside the bag will be reduced making the drone believe that it is gaining height. The actual vertical speed will stay zero because the drone is resting on the table throughout the process and I wonder what figures will show up in the flight records.



Here is the screen capture of DJI GO 4 :

To view this content we will need your consent to set third party cookies.
For more detailed information, see our cookies page.

Accept third party cookies

It can be seen that the barometer data has a rather big contribution to the vertical velocity shown on DJI GO 4 and that in the .TXT file. Just wondering how much contribution the accelerometer data has. May be = z vel - differential of height wrt time ?




The .DAT and .TXT files are attached

 

[BudWalker]

OK, I am bored today so I decided to do a little experiment. I put my M2P in a trash bag and connect the mouth to my vacuum cleaner . When the vac cleaner is turned on, the pressure inside the bag will be reduced making the drone believe that it is gaining height. The actual vertical speed will stay zero because the drone is resting on the table throughout the process and I wonder what figures will show up in the flight records.

View attachment 91146

Here is the screen capture of DJI GO 4 :

To view this content we will need your consent to set third party cookies.
For more detailed information, see our cookies page.

Accept third party cookies

It can be seen that the barometer data has a rather big contribution to the vertical velocity shown on DJI GO 4 and that in the .TXT file. Just wondering how much contribution the accelerometer data has. May be = z vel - differential of height wrt time ?

View attachment 91147


The .DAT and .TXT files are attached

I did an experiment like this - probably more than 2 years ago. I used a large ziploc bag. I'll try to find the post.