[gdal-dev] GDAL 1.11.0 and Cosmo-skymed L1A imagery

[Even Rouault]

Victor,
> 
> Recently, I updated GDAL to the version 1.11.0 and found the problem when
> read CSK L1A imagery :
> gdalinfo HDF5:"CSK_SCS_TEST_FILE.h5"://S01/SBI shows that image has 2 bands
> encoded as Int16
> 
> While 1,5 year ago, in the ticket http://trac.osgeo.org/gdal/ticket/4227
> there was a patch to interpret CSK L1A imagery as complex data. Similar
> example : gdalinfo HDF5:"CSK_SCS_TEST_FILE.h5"://S01/SBI would output 1
> band encoded as CInt16.

No, the patch applied only improved handling of dimensions of CSK-L1A datasets 
that didn't work before. The HDF5 driver has never had code to handle complex 
data with the related GDAL appropriates data types (CInt16, CInt32, CFloat32 
or CFloat64). This would be a further enhancement.

> 
> So, my question is whether it is inteded to come back to complex
> interpretation of CSK L1A imagery and for which version ?

I'm not aware of any pending work on this. As usual, this depends on if people 
contribute patches to improve that or support a developer to do it for them.

Even

-- 
Spatialys - Geospatial professional services
http://www.spatialys.com