[gdal-dev] GDAL 3.x and AutoCAD GeoTIFF georeferencing

[Kirk Waters - NOAA Federal]

Hi all,
I'm finding a change in georeferencing between GDAL 2.4 and 3.x that
results in AutoCAD not being able to figure out where on earth a geotiff
should be. I don't have AutoCAD, so I'm reliant on what my downstream users
are telling me. These are elevation Tiffs, so they do have a vertical
component to the SRS and that may be where the difference lies. I know it
isn't up to GDAL to solve AutoCAD's lack of standards adoption, but I'm
wondering if anyone has come across this and if there's a way to force the
outputs that AutoCAD will read.

Two small files are included in a zip file at
ftp://ftp.coast.noaa.gov/tmp/kwaters/autocad_issues.zip. One
(Job640647_ches_coned.tif, made with GDAL 2.4) has georeferencing that
works in AutoCAD, the other was made with GDAL 3.3.1 and has georeferencing
that fails. Both were made with gdal_translate -a_srs EPSG:3582+6360 with
the difference being the version of GDAL. They are not the exact same
location on the ground.

If I look at the files with listgeo, the older version has more tags and
includes the horizontal and vertical units. With gdalsrsinfo, the newer one
shows the more complete COMPOUNDCRS that includes the vertical component.

I'm hoping I don't have to go back to GDAL 2.4 for AutoCAD users. If anyone
has come across a solution before, I'd love to hear it.

Thanks,

Kirk Waters, PhD                     | NOAA Office for Coastal Management
Applied Sciences Program      | 2234 South Hobson Ave
                                                 | Charleston, SC 29405
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.osgeo.org/pipermail/gdal-dev/attachments/20211007/5893ccb0/attachment.html>