Fractal Terrains 3 header format with ETOPO1

elkrem

Hello all,

I'm attempting to use the ETOPO1 data in FT3, but I must be missing or incorrectly defining some parameters. Is there some documentation on the .hdr format, or has somebody successfully been able to import this data? I've gotten ETOPO2 to work (and ETOPO5 that comes with FT3, obviously), but I can't seem to transfer that information. The best I get is just a giant 0 height map.

Here's what I'm using so far:
BYTEORDER I
LAYOUT BIL
NROWS 10800
NCOLS 21600
NBANDS 1
NBITS 16
BANDROWBYTES 43200
TOTALROWBYTES 43200
BANDGAPBYTES 0
NODATA -32768
ULXMAP -180.000000
ULYMAP 90.000000
XDIM 0.01666666667
YDIM 0.01666666667

jslayton

I use the bedrock images for ETOPO1 rather than the global relief images, but the data should be the same. I also use the floating-point surface rather than the integer one. It's a little more data, but I'm sure there won't be any oddities in the data translation ( http://www.ngdc.noaa.gov/mgg/global/relief/ETOPO1/data/bedrock/grid_registered/binary/etopo1_bed_g_f4.zip ). For this file, the attached settings work.

The HDR format in FT is something that came with the original ETOPO5 data files from the USGS, I think. I can't recall where it came from, though. As you have no doubt noticed, it's not quite the same format as the HDR files that ship with the ETOPO1 data.

Your HDR file looks good, with the exception that BANDROWBYTES and TOTALROWBYTES should be 43202 and NCOLS should be 21601 (possibly NROWS should be 10801 as well, but I didn't look at it that closely). When FT asks you to load the HDR file, you'll need to set the right and bottom values to +180 and -90, respectively (the slight overage from the roundoff error for XDIM and YDIM causes all manner of havoc internally). Using these settings I was able to get the 16-bit integer bedrock ETOPO1 to load.

elkrem

Thanks so much, that worked perfectly! You were right in that I was letting the XDIM and YDIM multiplications just calculate the width, and that floating point rounding error was hosing the whole thing. I was using 43200 and 21600 because the binary format has one fewer row and column, but I just switched to the 16-bit and used your settings. Wow, sure is beautiful!

Thanks again!