[Proj] Differences in Output between proj and cs2cs

[Frank Warmerdam]

On Tue, 01 Mar 2005 15:12:52 -0500, Gerald Evenden
<gerald.evenden at verizon.net> wrote:
> If they haven't messed with "proj" and the way I programmed it, the
> "datum=WGS84" parameter on the proj runline is ignored.  Only the
> +ellps=WGS84 has any meaning to proj.  It converts the geographic
> coordinates to planar coordinates based upon a spherical earth
> having a radius the same as the semi-major axis of the WGS84 ellipsoid.
> That is, the Equidistant Cylindrical only functions in spherical mode.
> If you execute proj with the -V option, proj will list +parameters that
> are not used by the projection as well as a moderately complete
> diagnostic of the projected numbers.
> 
> In the case of the cs2cs program a datum change is taking place because
> of the two +datum arguments.  Thus the geographic coordinates are
> probably changed before they get to the projection program/procedure.

Martin,

I would confirm Geralds analysis.  It would appear that the 
proj=eqc definition is internally altering the ellipsoid to be a
sphere, while the proj=latlong case does not do this.  So the
cs2cs case is doing a conversion from the sphere to the
WGS84 ellipsoid since all it sees internally is the adjusted
ellipsoid definition, not the one you original requested.

I sometimes wish it was more obvious when only spherical 
ellipsoids were allowed with PROJ.4 (and libproj4). 

Best regards,
-- 
---------------------------------------+--------------------------------------
I set the clouds in motion - turn up   | Frank Warmerdam, warmerdam at pobox.com
light and sound - activate the windows | http://pobox.com/~warmerdam
and watch the world go round - Rush    | Geospatial Programmer for Rent