GCS missing in 3SM produced in ContextCapture 20 and uploaded to PW ContextShare

Hello Mariusz, 

Could you please let me know which coordinate system you are using and which geoid ? (EPSG codes)? 

Are you downloading the 3SM from ContextShare and trying to attach it in ORD or are you streaming the mesh to ORD ? 

In Update 20 we made some modifications on how the GCS are communicated to the 3SM, to avoid this kind of issues, so I am interrested in knowing more about this specific case. 

Have you logged a service request already ? 

Thanks,

Sylvain

Hi Sylvain,

Thanks for your response.

Our GCS applied in CC is:

IRENET95 / Irish Transverse Mercator (EPSG:2157) + Malin Head height (EPSG:5731).

The geoid entry in vertcs.override.csv is as follows:

We wanted to attach in ORD from ContextShare which means streaming. This works with older meshes i.e. produced in CC19 with the same GCS. settings).

We did notice a warning in CC when setting up production at Spatial reference system:

However, our Block and Reconstruction use the same SRS as above.

We haven't logged SR yet.

The implementation of geoids in Contextcapture is strange approach. It requires full geoid model only to show the model(which covers only small part of it) otherwise alarming errors. Why could not simply embed the part of geoid in 3sm or just use some average offset and to avoid need to have geoid for publishing. The option to be able to use geoids is advantage comparing to other SFM software but disadvantage at the same time as not reasonably implemented.

What you could try is to export this coordinate system definition in Contextcapture as custom WKT format and then reimport back. This way it won't be in unsupported EPSG:2157+ESPG:5731 but as custom coordinate system. Actually it would be good if could have option to drop geoid on production but keep height offset and avoid need for custom GCS.

The implementation of geoids in Contextcapture is strange approach. It requires full geoid model only to show the model(which covers only small part of it) otherwise alarming errors. Why could not simply embed the part of geoid in 3sm or just use some average offset and to avoid need to have geoid for publishing. The option to be able to use geoids is advantage comparing to other SFM software but disadvantage at the same time as not reasonably implemented.

What you could try is to export this coordinate system definition in Contextcapture as custom WKT format and then reimport back. This way it won't be in unsupported EPSG:2157+ESPG:5731 but as custom coordinate system. Actually it would be good if could have option to drop geoid on production but keep height offset and avoid need for custom GCS.

Thanks Oto for your suggestions. I guess we will see how this investigation develops further.