I have found a bug in the Xtherion command "\help\Calibrate bitmap".

The command does not work when the longitude is between Greenwich and 1 degree west (example -0.2333). I do not think it should work for the latitudes between the equator and 1 degree south.

I think the problem comes from the "ll2ozi" function of the Xtherion "mapcalc.tc" file. It takes the integer value of the longitude and latitude (int [$ lat]). When the longitude is between 0 and 1 of west corresponds a negative value of 0, which the integer function considers a error.

The problem is that I just have two karstic zones that are located in those longitude.

As long as that error is not corrected I have developed in following trick.

Process twice the pdf file with the command "\help\Calibrate bitmap".
With the first execution I create the png file. Later it changed the name to the pdf, and I open it in an editor, and search "thCalibrate". I eliminate the minus signs of the coordinates (L, F).
Execute again "\help\Calibrate bitmap" with this new file, generating the file ".map". I open and edit the .map file changing the E by W in the coordinates.
Finally I rename the ".map" file to match the first created png file.

It works but it is a bit cumbersome.

Regards,

Evaristo.

El 26/03/2018 a las 13:25, Evaristo Quiroga via Therion escribió:

The reverse engineering is not with the PDF. Martin do a good job here putting in all the therion output pdf a comment with the reference. You only have to open the file and search "thCalibrate". You found the next text:

/thCalibrate0 (X=-22.16017pt, Y=260.12132pt, L=-0.3641990570, F=42.2992656650)
/thCalibrate1 (X=1921.38345pt, Y=199.8752pt, L=-0.3608862211, F=42.2991894208)
/thCalibrate2 (X=3864.92708pt, Y=139.62904pt, L=-0.3575733981, F=42.2991130810)
/thCalibrate3 (X=-0.2183pt, Y=967.96619pt, L=-0.3641616744, F=42.3001613659)
/thCalibrate4 (X=3886.86893pt, Y=847.47388pt, L=-0.3575359218, F=42.3000087771)
/thCalibrate5 (X=21.72351pt, Y=1675.81102pt, L=-0.3641242900, F=42.3010570667)
/thCalibrate6 (X=1965.26714pt, Y=1615.56487pt, L=-0.3608113605, F=42.3009808177)
/thCalibrate7 (X=3908.81078pt, Y=1555.31871pt, L=-0.3574984437, F=42.3009044731)
/thCalibrate8 (X=1943.32529pt, Y=907.72003pt, L=-0.3608487917, F=42.3000851193)
/thCalibrate (HS=3913.97253pt, VS=1573.34683pt, HD=546.7244489428)

All the information is here: pdf coordinates (X,Y), real coordinates (L,F) and the pdf width (HS) and height (VS).

You only need know the width and height from the generate PGN, and you have all the data to do the calculation.

The reverse engineering is to understand the ".map" format and found in the tcl file (mapcalc.tc) the equations to do the conversion.

Regards,

Evaristo.

El 26/03/2018 a las 12:57, Nikita Kozlov via Therion escribió:
You can export 2 PNGs one with grid to put GCPs on, and second PNG without grid, but exactly same size (in pixels).

And use GCPs put on first grid-equipped PNG to georeference second "grid-free" PNG.

Grid is used only to get pixel coordinates.

Dealing with pdf and 'reverse engineering' for me seems a bit complicated stuff,

since in addition to 6 georeference parameters, you need to deal with scale and so on,

so I'd rather used quick and dirty solution with grid and rasters.
_______________________________________________
Therion mailing list
Therion@speleo.sk
https://mailman.speleo.sk/listinfo/therion
_______________________________________________
Therion mailing list
Therion@speleo.sk
https://mailman.speleo.sk/listinfo/therion