Thanks for the many responses I received on this. I made some incorrect
assumptions about Mapinfo. First of all, I didn't realize that something I
thought was so complicated could be handled so easily - I thought that it
handled the projection and the datum separately, as Arcview does. In Arcview,
just changing the projection does not also change the datum. It's a multistep
operation. In Mapinfo, the File-Save Copy As command will change both the datum
and the projection where appropriate. (And it does it exceptionally fast
compared to Arcview, I must add.)
Another error I made was in not specifying the proper projection of the file to
be exported. It was a file which had been previously imported from Arcview.
This, of course, is critical if one is to reproject the data!
A couple of insightful responses follow:
********************************************
Berk,
There's one other point I forgot to mention in my reply of a few minutes
ago. Are you aware that Mapinfo "reprojects" (misuse of "projection" again)
files of different projections/datums/coordsys on the fly that are in the
same map window? Because of this, it may not be necessary to convert them
yourself- MapInfo should handle the conversion automatically when you open
them into the same map window. This is something that Arcview can't do, so
you may not have been aware of it.
There was a thread on MI-L recently about this feature, regarding the
accuracy of the "on the fly" calculation. Some precision is sacrificed for
speed of calculation, which may be causing the offset you're getting. The
Save Copy As method makes no precision tradeoffs though, so that method
should *definitely* work out!
Dennis Waardenburg
*********************************************
Converting Coordinates from One Datum to Another
When converting coordinates from one datum to another, MapInfo has used the
Molodensky (3-parameter) and Bursa-Wolfe (7-parameter) methods. These are
general-purpose methods that can convert coordinates from any datum to any
other datum.
After the NAD 83 datum was introduced, NOAA developed a program called
NADCON, which stands for North American Datum CONversion. This is a very
specialized program that converts coordinates only from NAD 27 to NAD 83
and vice versa. For this specialized task, it's much more accurate than
the Molodensky general-purpose method; NADCON is accurate to about 0.1
meter, and Molodensky is accurate to only 10-30 meters. Most U.S.
government agencies, including the Census Bureau, have standardized on
NADCON for converting between NAD 27 and NAD 83.
The NADCON algorithm is used to convert coordinates between NAD 27 and NAD
83 if those coordinates lie within the areas covered by NADCON (United
States, Puerto Rico, and the Virgin Islands). If the coordinates lie
outside those areas, or if they use datums other than NAD 27 or NAD 83,
MapInfo uses the Molodensky or Bursa-Wolfe conversion methods.
Due to the file access required, the NADCON conversion method can be
slightly slower than the Molodensky method. If you want to turn off the
NADCON conversion, add a �NADCON� entry to the registry. The registry entry
should have this path:
HKEY_LOCAL_MACHINE\Software\MapInfo\MapInfo\Common\NADCON
If this entry is set to zero, then the Molodensky conversion method will be
used instead of NADCON.
If using the 16-bit version of MapInfo, add this line to the [MAPINFO
COMMON] section in MAPINFO.INI to turn off the NADCON conversion:
NADCON=0
,,,
/'^'\
( o o )
---------------------------oOOO--(_)--OOOo------
Michael Naughton
********************************************
If it is merely a change of datum that you want, e.g., from UTM NAD27 to UTM
NAD83, then File>Save Copy As _is_ the correct procedure.
>From your description of what happens when you do that, however, I wonder if
the datum was incorrectly specified when the layer was first imported? If
the new projection specifies NAD83 but the polygons still don't overlay the
Tiger data, that is probably the case. I'll have to think about how to
resolve that one.
Best,
Drew Phillips
Quaternary Geology Section
Illinois State Geological Survey
****************************************
----------------------------------------------------------------------
To unsubscribe from this list, send e-mail to [EMAIL PROTECTED] and put
"unsubscribe MAPINFO-L" in the message body, or contact [EMAIL PROTECTED]
