[PROJ] Proj / cs2cs differences

[Even Rouault]

Roberto,

the proj utility only works with a projection operation, where input 
geographic coordinates are expressed in the base CRS of the projected 
CRS (which is EPSG:4162 "Korean 1985" here), and doesn't apply any datum 
transformation. So the equivalent of your proj invokation  with cs2cs 
would be cs2cs +init=epsg:4162 +to +init=epsg:5174

If you use EPSG:4326 as the input CRS, then there will be a datum shift 
from EPSG:4326 to EPSG:4162

Regarding the pipeline output by projinfo -s -t, you can't use it with 
the proj utility, which is a historic one and is aimed only at 
understanding +proj={projection_id} style of PROJ strings. You should 
rather use the newer cct utility.

Even

Le 18/09/2021 à 14:13, Roberto Toldo a écrit :
> I was able to get the actual pipeline transformation used by "cs2cs" 
> by running:
> /projinfo -s EPSG:4326 -t EPSG:5174/
>
> Still, I'm having problems using the output proj string:
> /+proj=pipeline
>   +step +proj=axisswap +order=2,1
>   +step +proj=unitconvert +xy_in=deg +xy_out=rad
>   +step +proj=push +v_3
>   +step +proj=cart +ellps=WGS84
>   +step +inv +proj=molobadekas +x=-145.907 +y=505.034 +z=685.756 
> +rx=-1.162
>         +ry=2.347 +rz=1.592 +s=6.342 +px=-3159521.31 +py=4068151.32
>         +pz=3748113.85 +convention=coordinate_frame
>   +step +inv +proj=cart +ellps=bessel
>   +step +proj=pop +v_3
>   +step +proj=tmerc +lat_0=38 +lon_0=127.002890277778 +k=1 +x_0=200000
>         +y_0=500000 +ellps=bessel
>   +step +proj=axisswap +order=2,1/
>
> The error I get is "can't initialize operations that take non-angular 
> input coordinates". If I remove the "unitconvert" step, I can run proj 
> bin, but I get wrong results (even if I manually convert the input 
> from degree to radians). Any suggestions?
>
> Many thanks,
> Roberto
>
>
> On Fri, Sep 17, 2021 at 9:35 AM Roberto Toldo 
> <roberto.toldo at 3dflow.net <mailto:roberto.toldo at 3dflow.net>> wrote:
>
>     Hello!
>
>     I found some issues/differences from the "proj" bin and "cs2cs"
>     bin utilities. I'm using the last version, directly compiled from
>     Github.
>     Probably it's something due to my inexperience with the library,
>     so please bear with me :)
>
>     Specifically, I was trying to convert coordinates from geodetic to
>     EPSG:5174 ( Korean 1985 / Modified Central Belt).
>
>     I used proj with the parameters extracted from "projinfo EPSG:5174":
>     proj +proj=tmerc +lat_0=38 +lon_0=127.002890277778 +k=1
>     +x_0=200000 +y_0=500000 +ellps=bessel +units=m +no_defs
>
>     and then cs2cs with direct conversion parameters:
>     cs2cs +init=epsg:4326 +to +init=epsg:5174
>
>     However, I obtain quite different results.
>
>     For example, If I use lat_0 and lon_0 as input (127.002890277778,
>     38), I obtain exactly false_easting and false_northing (200000,
>     500000) when using proj4 (as expected), while I obtain something
>     different when using cs2cs (200185.04 499695.19).
>     I would be expecting the same results. I tried with other
>     coordinate systems (e.g., EPSG:32633 UTM zone 33N), and I noticed
>     no differences between proj and cs2cs results.
>     Can anybody help me understand what the reason for the differences
>     is? Are the parameters used internally by cs2cs different for
>     EPSG:5174? I also tried setting the accuracy, but it doesn't seem
>     to work.
>
>     Thanks,
>     Roberto
>
>
>
> -- 
> Roberto Toldo
> Technical Director & Co-founder | 3Dflow srl
> +39 045 4500109 | roberto.toldo at 3dflow.net 
> <mailto:roberto.toldo at 3dflow.net> | http://www.3dflow.net 
> <http://www.3dflow.net>
>
> _______________________________________________
> PROJ mailing list
> PROJ at lists.osgeo.org
> https://lists.osgeo.org/mailman/listinfo/proj

-- 
http://www.spatialys.com
My software is free, but my time generally not.

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.osgeo.org/pipermail/proj/attachments/20210918/22d385bf/attachment.html>