Community Forum

Cross-post P4RTK for PPK

(Michael Lambert) #1

If you PPK gurus have time this could be a big post…

(Christian Grüner) #2

I’m not really sure what it is about, just see a lot of trouble extracting/using the RTK functionality?

(Michael Lambert) #3

The main problem we are facing right now is trying to compile the correct logs to do PPK. DJI has formatted the majority of their files in BIN format and even when you can figure out how to get the data extracted it’s not working in RTKLIB. With the P4RTK you have to turn off rtk and run in PPK mode.

As you can see we have been at it for a couple of months. I don’t have a p4rtk myself, but I am trying to lend a hand and if it doesn’t work out it will solidify my reasoning for not going to that drone. On the other hand, my feeling is that we will figure it out and that we will be able to use bases other than the DJI in a PPK scenario. I know that there are a lot of folks on this forum that are much more experienced that PPK than I am. There are a few clues that popped up late last night so I’m going to give it another run and will report back to that thread.

(Matt Bodnar) #4

Do we know what OEM GNSS board is in the “DJI” base receiver?

Trimble?, Novatel? Hemisphere/Unistrong? Tersus?

This could lend a clue. I dont see why DJI would transform the raw messages coming from the gnss board just to block people. But i could be wrong.

(Michael Lambert) #5

That is the exact question I asked of the southwest regional sales manager and of course he did not have an immediate answer. He would report to tech support. I’m almost certain that it is a trimble device from the data I have been able to extract, but once I find something that will decompress the BIN without it being recognized as a disc img file I will find the code and hopefully confirm that.


Could you put up a sample dataset? I know you don’t have one but I had a hard time following thread. I downloaded the few on there but want to start from the very beginning.

(Michael Lambert) #7

This is the first one I saw from Gyokap.



I am confused on the formatting of the Rinex. There appears to be two position columns embedded. I am assuming one is raw and the other be the RTK position?


It looks like it was figured out and had nothing to do with the DJI format but rather incorrectly downloading the CORS base data?

(Michael Lambert) #10

There is now a new thread with an experienced P4RTK user…


Nice work GregO!

When does that .bin file get used? Not in RTKPost, right?

(GregO) #12

Correct, the .bin isn’t used in RTKPost. It just uses the the the text files.

(Michael Lambert) #13

Wow, I wasted all that time… :joy:

(GregO) #14

From: https://dronelife.com/2018/10/15/phantom-4-rtk-launched-globally-today-djis-ultimate-mapping-solution/

In areas with lack of RTK coverage, the DJI Phantom 4 RTK allows for the use of Post Processed Kinematics (PPK). The DJI Phantom 4 RTK captures original satellite observation data as well as the ephemeris data and stores it in a PPKRAW.bin file in RTCM 3.2. format. Additionally, the Phantom 4 RTK converts the satellite data on the fly to the RINEX format (Receiver Independent Exchange Format) and writes the data into a RINEX.obs file. Thanks to the internal synchronisation of the GNSS position and camera, the Timestamp.MRK file provides a precise recording of the highly accurate image position. All operation-relevant data is stored on a Micro-SD card in a unique folder for each mission. The consistency of the Phantom 4 RTK’s data reduces the time needed to validate the data and results in an efficient post-processing workflow with minimal to no manual adjustments.

From: https://www.reddit.com/r/UAVmapping/comments/9w47ki/workflow_postprocessing_dji_p4_rtk_data/

PPKRAW has the satellite observations in RTCM3.2 MSM5 format.
EVENTLOG has the time log in binary
Rinex has the rinex observations
Timestamp contains the following, in ASCI

  1. Photo number
  2. Exposure time of GPS seconds of the week
  3. Exposure time of GPS week
  4. Phase compensation between north antennae to CMOS centre (mm)
  5. Phase compensation between east antenna to CMOS centre (mm)
  6. Phase compensation between up/down antenna (downward is positive) to centre of CMOS (mm)

Here is a sample from the timestamp file:

1	275295.301059	[2024]	   -26,N	    19,E	   191,V	-36.41144502