Hi AndreI've added to my GTL editor the ability to load the OT's terrain classification. However, I'm a bit confused about the relationship between classification and elevation. Your documentation states that the height value is multiplied by 4 to obtain the 2-bit bit shift, and that FU3 divides the height value by four and subtracts the sea level value when rendering the elevation. Does this mean that the classification bits are part of the overall height value? Or does FU3 also subtract the value of the classification bits along with the sea level value?I know this may sound like a stupid question but I'm asking this because the OT's classification is all over the place: distribution of the three land classes is so random that classification values appear to me to be 'accident' rather than design. In fact, the distribution is so random that I thought there must be a bug in my code, but your GeoTiles Viewer returns the same results. The only land class that is ever rendered uniformly is a deliberately flattened airport area, and this leads me to suspect that classification values are just by-products of the elevation...? Can you help at all?Many thanksJonathan S.

Hi AndreI've added to my GTL editor the ability to load the OT's terrain classification. However, I'm a bit confused about the relationship between classification and elevation. Your documentation states that the height value is multiplied by 4 to obtain the 2-bit bit shift, and that FU3 divides the height value by four and subtracts the sea level value when rendering the elevation. Does this mean that the classification bits are part of the overall height value? Or does FU3 also subtract the value of the classification bits along with the sea level value?I know this may sound like a stupid question but I'm asking this because the OT's classification is all over the place: distribution of the three land classes is so random that classification values appear to me to be 'accident' rather than design. In fact, the distribution is so random that I thought there must be a bug in my code, but your GeoTiles Viewer returns the same results. The only land class that is ever rendered uniformly is a deliberately flattened airport area, and this leads me to suspect that classification values are just by-products of the elevation...? Can you help at all?Many thanksJonathan S.

Hi Jonathan,I never have worked with outer terrain before. My GeosViewer are valid for the elevation file of the high resolution area only.The file format is a .tag file and so you can open parts of outer terrain definitions with it, but the internal file structure is different and the results may be wrong. Classification for the high resolution area is coded as the last 2 bits of the height value and are not the same than the outer terrain classification. For high resolution the 32 bit value ishhhhhhhhhhhhhhcch is the height value cc the classification, 0,1,2 and 3 for land, water, urbanisation and unused by LG but set for flattened airport area by me :-) There are projects to expand the high resolution area to Vancouver and so I need to adapt the outer terrain elevation for Vancouver island, actually not existent. Deception pass with bridgesI will take a look at the outer terrain structures and modify my GeosViewer or build a OuterViewer this will need several weeks.Can you send a copy of your outer terrain editor to me please, this will help the analysis of the outer terrain files :-), thanksAndre

Hi Jonathan,I never have worked with outer terrain before. My GeosViewer are valid for the elevation file of the high resolution area only.The file format is a .tag file and so you can open parts of outer terrain definitions with it, but the internal file structure is different and the results may be wrong. Classification for the high resolution area is coded as the last 2 bits of the height value and are not the same than the outer terrain classification. For high resolution the 32 bit value ishhhhhhhhhhhhhhcch is the height value cc the classification, 0,1,2 and 3 for land, water, urbanisation and unused by LG but set for flattened airport area by me :-) There are projects to expand the high resolution area to Vancouver and so I need to adapt the outer terrain elevation for Vancouver island, actually not existent. eception pass with bridgesI will take a look at the outer terrain structures and modify my GeosViewer or build a OuterViewer this will need several weeks.Andre

Hi AndreThe OT's tag file structure is different only because it includes GTL data for each megatile. Otherwise the format is identical to the geotiles.tag files used by the high-res regions. This means that your GeoTiles Viewer can still generate the OT's classification, although the sequencing of the class bits is different. It's obvious that water is 0, but the remaining three land classes are, as mentioned, distributed so randomly that it's difficult to say which classes the values 1, 2 and 3 actually represent. This is why I asked whether LGS set the classification bits 'deliberately' or whether the classification bits are just 'accidents' produced by the overall height value. This is easier to understand is you compare the high-res classification at level 5 with the OT's classification at the same level. Distribution of the high-res classes is as you would expect: urban and land classes appear exactly where you'd expect to see them. By comparison, distribution of the OT classes is all over the place: an urban area can be made up of a crazy mix of all four classes. My tool produces the same results as your GeoTile Viewer.Theoretically, it should be easy to isolate the 'flattened airport' class, but at the moment I'm not even sure about this! I've loaded the classification of two tiles I know contain flattened airport areas: in one tile the flattened airport area is classed as 2 and in the other as 3! These areas appear to be the only examples of 'deliberate' classification yet LGS used different values in different tiles! Gaaaargh! :( Best guess so far is land = 1, urban = 2, flattened area = 3 but I'm open to suggestions!RegardsJonathan S.

Hi AndreI'll send you the editor itself, but because this is a spreadsheet (it's in OpenOffice format and contains lots of Star BASIC code) I'll include the source code separately so that you don't have to download OpenOffice in order to see what's what. The editor includes a hidden worksheet containing the L0, L2, L5 and GTL file offsets so I'll send this as a separate Excel file just in case it comes in handy.Just promise me you won't laugh at my decidedly second-rate programming skills! :( RegardsJonathan