[gdal-dev] Current state of Winkel Tripel projection support

[Even Rouault]

Dominique,

A few remarks:
- There's no EPSG:54030
- There's a ESRI:54030
- ESRI:54030 is World_Robinson, so no Winkel Tripel
- The PROJ implementation of Winkel Tripel supports since 2015 the inverse method, using 
an iterative approach. But there might potential convergence issues. I see it uses a very small 
epsilon in longitude/latitude (1e-12 radian)
- Robinson uses also an inverse method using an iterative approach

The best would be to identify the points that fail to reproject and see if it is expected or not.

Even

> Hi all,
> 
> I'm trying to reproject a raster from WGS84 to Winkel Tripel (EPSG 54030)
> but unfortunately I encounter some error messages :
> 
> non-convergent inverse meridional dist
> latitude or longitude exceeded limits
> 
> I've found some old messages about PROJ not supporting Winkel Tripel
> projection because of a missing inverse transformation function. But I've
> the feeling that this projection is better supported now. So I would like
> to know if there are still limitations with it and hopefully some
> explainations about these errors and a possible workaround. Currently I've
> tried to reduce the target extent but with no luck.
> 
> Regards


-- 
Spatialys - Geospatial professional services
http://www.spatialys.com
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.osgeo.org/pipermail/gdal-dev/attachments/20200602/f15ed183/attachment.html>