Please be aware that MapInfo has released a patch for MapInfo 7.8 which should improve the datum conversions, you are discussing. But I wouldn't know how much better they have made these transformations, so I guess you'll have to test yourself - but of course this patch requires a MapInfo 7.8 Peter Horsb�ll M�ller GIS Developer Geographical Information & IT COWI A/S Odensevej 95 5260 Odense S. Denmark Tel + 45 6311 4900 Dir +45 6311 4908 Mob +45 5156 1045 Fax + 45 6311 4949 E-mail [EMAIL PROTECTED] http://www.cowi.dk <http://www.cowi.dk/>
________________________________ From: Steve McArthur [mailto:[EMAIL PROTECTED] Sent: Tue 22-Feb-05 17:01 To: Susan Yu Cc: MapInfo-L (E-mail) Subject: RE: MI-L NAD27 to NAD83 coordinate conversion On the limitations that Gilbert Hach� wrote on, I don't think these are specific to NAD27 to NAD83 datums or Canada. MapInfo uses quite a generalised formula for converting between datums, or more precisely, spheroids. The reason for this is mainly for speed - that's why you can quickly add tables of differing projections into the one Map window. When I was working back in Oz we encountered this problem when dealing with conversions between AMG84 and MGA94 which are based on completely differing mathematical models of the earth (the spheroid). The only real way to get a precise conversion is to use distortion grids (ie NTV2 grids), but this would be horrendously slow and cumbersome to implement in MapInfo (in my belief). These grids are based on measured 'distortions' between the two differing spheroids rather than mathematical conversion formulae, which is why I think MapInfo would steer away from some general implementation. It does, however, make for an ideal localised add on. Hope I've got the right end of the stick with the NAD/Canadian datums Regards Steve McArthur -----Original Message----- From: Susan Yu [mailto:[EMAIL PROTECTED] Sent: 22 February 2005 15:34 To: Gilbert Hach� Cc: [email protected] Subject: Re: MI-L NAD27 to NAD83 coordinate conversion Wow! Thanks all, for all the quick replies! I wish you were all my co-workers, because then I'd be more productive at my work, and my boss would be happier with me! I've received many suggestions. I started with the easiest fix first and it seemed to work: (1) Jason Adam wrote: "Susan, I think you are doing it right, but the column isn't big enough to hold the data. Make sure the X and Y fields are either Float or Decimal (10,5) and try the Update again." I did try this and all my -99.99999 disappeared with numbers that look much better. I feel like a knob for not checking this sooner! (2) Steven Smith suggested: "I normally Export the file to a MIF then change the header. Normally done in the MIF file on the line with Coordsys Earth Projection. There are 4 coords which bind the area of the maps. CoordSys Earth Projection 8,104,"m",57.0000000,0.0000000,0.9996000,500000.000,0.000 " I would have definitely tried this if Jason's suggestion didn't work out since I'm all for editing MIFs. I haven't tried it, so I don't know if it would have worked for me. Andrew Canfield suggested a similar method, so I believe that this is commonly done with much success. (3) My last concern is what Gilbert Hache wrote below. I have read in the list group archive concerning the Canada NAD27 limitations in MapInfo. I'm using MapInfo 7.5. Do you think at this version, MapInfo have resolved this issue? I have checked out the NRC website (Geodetic Survey Division) and tried their online NTv2. http://www.geod.nrcan.gc.ca/index_e/online_apps_e/appNtv2_e/appntv2_geo_e.html This was where I tried to verify that my MapInfo method of converting NAD27 to NAD83 worked. (It seemed to be for a few points.) But since I have over 15000 points, I rather not check each one but was hoping to see if anyone was aware if there actually still was a NAD27 Canada MapInfo problem in existance...! Thanks again everyone, for your help! Regards, Susan On Tue, 22 Feb 2005 09:58:52 -0500, Gilbert Hach� <[EMAIL PROTECTED]> wrote: > You have encounter "the Mapinfo canadian problem" > > Mapinfo has a problem converting data from NAD27 to NAD83 and vice > versa IN CANADA. It's a known Mapinfo limitation. Personnaly, I've > been asking to solve since version 5,5 or so... > > The NAD83 conversion used in Mapinfo is made using equations for > AMERICAN standards. In Canada, we use a set of reference points called > NTV2 and equations made to use to reference point to respect Canadian > standards. > > The results of using the Mapinfo internal transformation that are not > Canadian based will create errors that can be a few meters to 20-30 > meters off! (and maybe more...) > > You have to use an external tool to do so like the FME from SAFE. This > tool allows the usage of a transformation grid. You could then set it > to use the NTV2 grid file or other more local grid file (some province > have local, more dense grid files) > > Like I said, Mapinfo is aware of this situation but never supported > the Canadian standards. I asked before every upgrade of Mapinfo Pro > without any results... > > It's a real pain but if you want to have precise data to transform > them, you have to use an external software at this point. FME from > safe (www.safe.com) can be downloaded and active for a few days as > evaluation. I'm not sure if the NTV2 file is provided with it. You may > have to get it from Natural Resources Canada... > > Hope this help > > GH > > -----Original Message----- > From: Susan Yu [mailto:[EMAIL PROTECTED] > Sent: Tuesday, February 22, 2005 09:32 > To: [email protected] > Subject: MI-L NAD27 to NAD83 coordinate conversion > > Hello, > > I have a bunch of coordinates in NAD27. I would like to convert them > to NAD83. So what I've done is create points, setting the projection > at Longitude/Latitude (NAD27 for Canada). Then I saved the table as > Longitude/Latitude NAD83. To retrieve the new NAD83 coordinates, I > used update column's "centroidx" and "centroidy" functions. This > seemed to work okay except for coordinates with longitude less than > -100.00000. In fact, anything less than -100.00000, the new NAD83 > defaults to -99.9999... like it can't calculate past that boundary. > > All of my NAD27 coordinates are everywhere in Canada from BC to the > Atlantic provinces. > > My knowledge of datums and projections is not the strongest, so I was > hoping someone could shed some light on why this occurs. > > Thanks, > Susan > > --------------------------------------------------------------------- > List hosting provided by Directions Magazine | www.directionsmag.com | > To unsubscribe, e-mail: [EMAIL PROTECTED] > For additional commands, e-mail: > [EMAIL PROTECTED] > Message number: 15382 > > -- Susan Toronto, Ontario --------------------------------------------------------------------- List hosting provided by Directions Magazine | www.directionsmag.com | To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED] Message number: 15385 London Borough of Hackney may exercise its right to intercept any communication on its networks - for more information see http://www.hackney.gov.uk/email_disclaimer.html --------------------------------------------------------------------- List hosting provided by Directions Magazine | www.directionsmag.com | To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED] Message number: 15387
