Caves of Castleton - 3d Survey Project
- Thread starter Moose
- Start date
footleg
New member
It is worth also pointing out that the Therion viewer (Loch) supports importing additional Survex .3d files into an existing model (at least in the beta downloads since I asked for this feature a while back and it got added the next day!).
So if you build a .lox model file containing just the surface overlay, then people can download the latest survex .3d file and import it into the Loch viewer which already has the surface model loaded from the .lox file. The latest cave can then be viewed under the transparent surface overlay without anyone needing to build a new .lox model every time. You just need to make sure the surface model has the same magnitude coordinates as the survex file so they are correctly aligned.
So if you build a .lox model file containing just the surface overlay, then people can download the latest survex .3d file and import it into the Loch viewer which already has the surface model loaded from the .lox file. The latest cave can then be viewed under the transparent surface overlay without anyone needing to build a new .lox model every time. You just need to make sure the surface model has the same magnitude coordinates as the survex file so they are correctly aligned.
Lampwick
Member
I'm not into surveying (yet) but I have noticed that Ordnance Survey (c)(tm) have released a shed-load of data, including DTM stuff on a 50m grid and contour data at 10m vertical interval. Is this stuff any use in the 3d model for surface features?
More info on the OS(c)(tm) website, look for "OS OpenData".
More info on the OS(c)(tm) website, look for "OS OpenData".
carabeener
New member
Is the .3d file still around? The link doesn't seem to be working...? 
Rob
Well-known member
This link still doesn't work for me:
http://caving.hucklow.org/castleton_survey/
http://caving.hucklow.org/castleton_survey/
underground
Active member
Looks to me like the hucklow.org domain has expired - Rob, might be worth mentioning to Nick in case he's not aware?
graham
New member
The Castleton 3D Survey model can now be viewed at the following link: http://www.peakdistrictcaving.info/caves-of-castleton-survex-project.htm
Very nice work. I would second graham's suggestion that you look seriously at the cave survey archive site. It's an easy place to store files, of whatever type you like, and have people collaborate on the work of updating stuff, and helps preserve data in the long term. Is there a reason why you don't want to use it?
And I note the point about detailed DEMs and survex. There are moves afoot to improve survex's terrain-handling, now that this sort of data is quite widely available, as are fancy opengl-capable graphics cards. Someone has to get round to writing some code, but I think it's reasonably likely to happen in the not too distant (I do have a really hacky patch that draws a proper surface over the cave, but it needs Z-depth, transparency, lighting, real-world co-ordinates, and reading in of suitable file formats before it's any real use.
Can you give some details of the data format used?
And I note the point about detailed DEMs and survex. There are moves afoot to improve survex's terrain-handling, now that this sort of data is quite widely available, as are fancy opengl-capable graphics cards. Someone has to get round to writing some code, but I think it's reasonably likely to happen in the not too distant (I do have a really hacky patch that draws a proper surface over the cave, but it needs Z-depth, transparency, lighting, real-world co-ordinates, and reading in of suitable file formats before it's any real use.
Can you give some details of the data format used?
Moose
New member
I am happy to store any data with the cave survey archive site, as long as there's some control over adding/editing any of the data to avoid confusion. AA has been on my case, I've just not got round to it.
The bluesky data came to me in dxf format (X,Y,Z) which then took some serious data shuffling to create a series of .svx files with the surface depicted as fixed points. I haven't done the whole 7.5km x 4km area as it's just too much for survex to handle. The whole picture was entered into arcview which created a 1m contour with the addition of a georeferenced dxf output from survex to show the underground.
http://www.peakdistrictcaving.info/castleton_survey/arcvew_print_overall_091005.pdf
The bluesky data came to me in dxf format (X,Y,Z) which then took some serious data shuffling to create a series of .svx files with the surface depicted as fixed points. I haven't done the whole 7.5km x 4km area as it's just too much for survex to handle. The whole picture was entered into arcview which created a 1m contour with the addition of a georeferenced dxf output from survex to show the underground.
http://www.peakdistrictcaving.info/castleton_survey/arcvew_print_overall_091005.pdf
graham
New member
Moose said:
There is, it can be set so that only you, or specific people nominated by you can be given that sort of access. For the Cheddar dataset, for example, there are only three or four people with editing access. In practice, though, only Andrew or I actually do that.
I wonder if there's a decent subversion application that runs on a mac ... :-\
genehammond
New member
Wookey - Ive been helping Moose with various elements of the Castleton work, especially the terrain data and GIS mapping. It would be useful if Survex could take other simple data foramts such as ASCII x,y,z data (lattice points as a simple list), or else ESRI ASCII Grid format. These formats are pretty efficient and easy to use and would make getting high resolution terrain data into Survex a doddle!
Cheers
Gene hammond
Cheers
Gene hammond
Yep. Agreed. The catch is that someone has to do some coding. Olly is a very busy man these days and so am I, so time for Survex coding is thin on the ground. If anyone else gets enthused it would be great. There is a great long list of 'would be nice' as is obvious to anyone who uses the program much.
Looks like shapefile reading always has a handy floss library in the form of shapelib (http://shapelib.maptools.org/)
Adding simple XYZ data should be simple enough.
And there is libdime for DXF data, and libredwg for DWG files.
The harder part of this is getting survex to deal with real-world co-ordinate systems rather than it's own cartesian space. Libproj exists to help with conversions (and we recently added some code for exporting entrance locations in arbitrary co-ordinate systems), but to do this properly needs some thought so any data can be supplied in any specified co-ordinate system and survex will read it in or write it out, as required. If we get that right then whatever crazy stuff people need to do can work.
Looks like shapefile reading always has a handy floss library in the form of shapelib (http://shapelib.maptools.org/)
Adding simple XYZ data should be simple enough.
And there is libdime for DXF data, and libredwg for DWG files.
The harder part of this is getting survex to deal with real-world co-ordinate systems rather than it's own cartesian space. Libproj exists to help with conversions (and we recently added some code for exporting entrance locations in arbitrary co-ordinate systems), but to do this properly needs some thought so any data can be supplied in any specified co-ordinate system and survex will read it in or write it out, as required. If we get that right then whatever crazy stuff people need to do can work.
jarvist
New member
Gene - I found it really quite efficient to push DEM data into Survex as *fix commands, with a linking mesh.
I wrote some (horrible!) python code back in 2009 to do this in a more repeatable manner than the 'one off' I'd done back in 2005. We received some TIFF file of high quality DEM data which we managed to get open in a GIS software package, and then export to a massive .xyz ascii file of tuples. The data was then resliced with these simple python programs, and left as different .svx files to include with differing meshes / presentations etc.
https://github.com/jarvist/migovecsurveydata/tree/master/migovecsurveydata/DEM
I wrote some (horrible!) python code back in 2009 to do this in a more repeatable manner than the 'one off' I'd done back in 2005. We received some TIFF file of high quality DEM data which we managed to get open in a GIS software package, and then export to a massive .xyz ascii file of tuples. The data was then resliced with these simple python programs, and left as different .svx files to include with differing meshes / presentations etc.
https://github.com/jarvist/migovecsurveydata/tree/master/migovecsurveydata/DEM
