Agisoft Metashape

Agisoft Metashape => Feature Requests => Topic started by: nachogi73 on November 15, 2020, 10:48:22 PM

Title: Spanish geoid rednap08
Post by: nachogi73 on November 15, 2020, 10:48:22 PM
Hi everybody.
I have seen the list of geoids available on the web.  In Europe I even see one created for Portugal.  Would it be possible to have one with the official data of Spain?
I have tried to create it myself but it doesn't work properly.
It would be great to have it created directly by the creators of the software.
Title: Re: Spanish geoid rednap08
Post by: nachogi73 on January 27, 2021, 12:59:58 PM
Hi again:

I found this geoid model available for download fron the oficcial web site of geodetic resources in Spain (

I place it in the folder GEOIDS into MetashapePro propper installation folder, but restarting Metasahe there is a error message: "Can't load geoid: C:/Program Files/Agisoft/Metashape Pro/geoids/es_ign_egm08-rednap.tif".

I want to create the conpound coordinate system whit this data following this link

There are some incompatibility with this file?

I cant attach the file because the size.

Please, some hepl !!!!!
Title: Re: Spanish geoid rednap08
Post by: Paulo on January 27, 2021, 02:26:56 PM
Hi nachogui73,

I did successfully create a tif file corresponding to Geoid es_IGN EGM08 REDNAP. Please follow this link ....

just copy it into the Metashape geoids folder and then you can create compound CS using Vertical Datum EGM2008 REDNAP IGN

Hope this can be helpful,

Updated link to new geoid file referenced to ETRS89
Title: Re: Spanish geoid rednap08
Post by: nachogi73 on January 27, 2021, 08:50:05 PM

Thanks for your reply.

I´m trying  to create the conpound CS based in UTM ETRS89-30 (EPSG 25830). This is the official in Spain in my area.

I have a problem in creation with the Tiff you transform. The system says that it is not compatible with taht CS and chage it to WGS84.

So my projetc coodinates are no UTM form.

Do you know what is the problem?

Thanks in advise.
Title: Re: Spanish geoid rednap08
Post by: nachogi73 on January 27, 2021, 08:59:21 PM

I had not seen the correction you have made on etrs89!!!!!

Thanks again...It seems working now.

Best regards.
Title: Re: Spanish geoid rednap08
Post by: nachogi73 on January 30, 2021, 01:11:48 PM
Hi again Paulo:

Thanks for your contribution, the CCS works correctly.

I have read in the forums a lot about the problems in elevation with the phantom 4 rtk.

I have a survey profile and have been working with drones for surveying for a long time.

In my projects with PH4RTK I always measure some control points with survey GPS equipment  to check the accuracy of the model height.

As you already know, there is always a difference. In flights of a large area I have verified that the difference with respect to the GCPs is the same for the same chunk, so I can assume an offset in Z.

The ideal would be to eliminate this difference.

I have read in the forum that a reference work could be done with a plenty of GCPs measured with GPS and add them to the calculation of the PH4 RTK to obtain some fixed camera parameters that would be used in the following works.

Do you know if this really works? Would you have to calculate different camera parameters for different flight heights?

Thanks in advance.
Title: Re: Spanish geoid rednap08
Post by: Paulo on January 30, 2021, 02:18:10 PM
Hi Nachogui73,

como estas?

The problem with flying P4RTK without control is that a small error in estimated focal length will automatically reflect in error in Z for mainly nadir looking projects. So I would fly a control field with lots of GCPs. To decorrelate the various calibration parameters, I would fly the control field with cross grid pattern at 2 different heights (maybe 50 m for one grid and 70 m or other) and also add a few obliques looking in difefferent direction to mainly nadir cross grids. Process the images and CPs in Metashape, optimize and use resulting calibration parameters as fixed in your regular projects. Should give you good results supposing you are mainly doing nadir work at around 50 to 70 m height over terrain.
Title: Re: Spanish geoid rednap08
Post by: flyzk on January 30, 2021, 09:07:59 PM
Hi Paulo,
This approach is good if you only fly in NADIR?
I use to fly a false Nadir (80-85) in order to get more detail.
What should I use to do?

Do the same trick you did 2 times in order to get a calibration parametter reusable for next flights. Then, separate each flight in 2 camera calibration group (one for 85 and the other for 80) and then apply the matching camera calibration parametter?
And every time I change the pitch angle I should previoulsy have determinated the calibration parametter for this angle?
Title: Re: Spanish geoid rednap08
Post by: Paulo on January 30, 2021, 10:23:08 PM
No I would just fly the control field with  a cross grid at 2 heights close to what you usually fly at and with false nadir 80-85. The idea is to fly the control at conditions close to usual but with some different altitudes and always adding a few obliques ( 40-50 inclination) to decorrelate the parameters. Once this calibration flight is done, the optimized calibration should hold as long as you dont radically vary the fligh parameters (height, inclination, etc...).

Title: Re: Spanish geoid rednap08
Post by: nachogi73 on January 31, 2021, 10:51:53 AM
Hi again
Thanks for the reply.
I will try several control flies in diferent elevations following your instructions.
Then I will share the results. We have a lot of wokrs to try and compare......
Lets see next week if wheather is fine....

See you
Title: Re: Spanish geoid rednap08
Post by: nachogi73 on February 07, 2021, 11:55:31 AM
HI again Paulo:

Just want to share our last experience related with this post.

As I already mentioned, we have many projects with P4RTK in which we have precise GPCS with survey GPS .
I have recalculated several of them with the new GEOID REDNAP and the height "offsets" that I had before have virtually disappeared.
I mean, difference in Z axis from 2-3 meters to 2-5 cm.

I made a compound CS last year not with the GEOTIFF, but with one Timble data (the only one I found that Metashape allow), and now I can see that something was wrong with that CS.
I thought, reading several forums posts, that the reason was a bug related with RTK camera or something.

But now I tried with big surface projects with reasonable elevation difference, linear projects measured in 3-4 flights with about 4 kilometers lenght and little projects with only 2 flights: In All of them the elevation accuracy with th enew Compund CS (made with the geotiff you convert for me) remains in a mean of 2-3 centimeters.

Taking account the accuracy of the GCP used for checking XYZ (1cm in XY/ 2 cm in Z), this is a GOAL for us in our survey services.

Thanks again for your helpful contribution.

Title: Re: Spanish geoid rednap08
Post by: Paulo on February 07, 2021, 04:00:27 PM
Great to see that the geoid tiff file was helpful!

Do nor hesitate to share any other concerns or insights on your work,