Format for CCS .dat files?

[scott967 0]

I'm brand new to the soaring game, and found this excellent site. I d/led and installed CCS2004 and used it to scan FS9 mesh, but I'm wondering what the format is of the .dat file so I could build it directly from elevation source data? I see the header and it appears to be 2 bytes per sample, but I can't figure out what the data represent.

TIA

scott s.

.

[Peter Lürkens 30]

The parameters in readable text are quite self-explanatory.

The coded part later on contains in a flat sequence, starting from the southern-western-most position information of slope patches, in a 2-byte structure. CCS samples the world in fractions of 3 arc seconds. The very first element of a dat-file is mapped to the southern-western-most coordinate in the header, the second element 3 arc second eastward, until a row is full. Then it advances 3 arcsecond northward to the next row and so on.

The 2byte structure contains orientation and tilt of a slope patch in hex-notation (0-F)

However, there is some confusion in the mapping of existing slope files, depending from their origin.

Originally an entry in a slope file created by CCS itself is mapped to the real world on a nearest sample or centered mapping scheme. This means that a sample in the dat-file identifies a square area of +/- 1.5 arc seconds around its associated location.

In difference to that, at least the majority of slope-files produced by external tools, usually from SRTM-data, have an integer-truncation or lower left corner mapping scheme. This means that a sample defines the east- and north-ward 3 arcseconds from its associated coordinate.

The first byte of each element indentifies the orientation of the slope patch, starting from N(0)-NNE(1)-NE(2)-ENE(3) ... NNW(F), i.e. in 16 clockwise directions.

The second byte identifies the tilt, again in 16 steps. Again, there is some confusion about it. It turns out, that slope-files created by CCS originally appear to calculate the tilt angle as something like s = int(15*atan(3y/x)/90°), being y the elevation difference and x the base difference of the corners of the slope patch. Thus the parameter s is not the geometric tilt angle of the slope. Instead, for small tilt angles the slope lift effect is roughly a factor 3 higher than could be expected in reality.

In difference to that, the independent slope-data files usually contain the slope tilt data as something like s = in(15*atan(y/x)/90°)) which is the geometric tilt angle and results in a realistic strength of the slope lift.

In simple terms, a slope-data file from CCS makes much stronger lift for flattisch slopes as the independent files.

If you plan to create a tool for making dat-files from geo-data, I propose to use the independent type, or to offer a selection between the two.

regards, Peter

[scott967 0]

Thanks Peter. Now that I know what I am looking at will will start to play around

scott s.

.