On Tue, 28 Dec 2004 10:48:51 -0600, Curtis wrote in message <[EMAIL PROTECTED]>:
> Frank Olaf wrote: > > > Curtis L. Olson wrote: > > > > > Chris Metzler wrote: > > > > > > > On Tue, 28 Dec 2004 15:55:41 +0100 > > > > Roy Vegard Ovesen wrote: > > > > > > > > > On Tuesday 28 December 2004 14:35, Frank Olaf wrote: > > > > > > > > > > > To where should I direct concerns about the scenery? > > > > > > > > > > > > There is quite a large portion of scenery near my home which > > > > > > is terribly incorrect. From N60.50 to N61.00, E11 there is ..are the southern half of all N60 thru N60.50 tiles valid? These tiles would have "known canted elevations"? The northern half of S60 would have the same problem, or do we have a lucky tile corner naming convention? > > > > > > a large strip running east and west with apparently no > > > > > > elevation data. > > > > > > > > > > This is true for the entire globe. The reason for this is that > > > > > FlightGear gets terrain elevation data from the SRTM: > > > > > > > > > > http://www2.jpl.nasa.gov/srtm/ > > > > > > > > > > They only mapped between N60 and S60. That leaves out most of > > > > > Norway flat. :-( > > > > > > > > If that's the problem, then why does the elevation data, and > > > > thus the terrain, suddenly return as you fly north of 61N? > > > > > > For areas that have absolutely no SRTM data, we fall back to the > > > 1-km globe data set. ..the 0.9.5 scenery is all 1km, we try to do this fallback in the 0.9.7 scenery? > > > This is easy to do in the current code on a tile-by-tile basis, > > > but hard to do if particular datapoints within a tile are bad. > > > > Since the break is more or less where the srtm data ends, I would > > suspect that the problem only is that the point from which to use > > alternative data sources is set too far north. If such is the case, > > it should be easily fixable, no? > > The problem is that SRTM provides data for those areas in question. > Coming up with an automatic algorithm to decide how and when to blend > data source is not necessarily straight forward. If someone wants to > take on the project of fixing SRTM data problems and blending with > other data sources, you are welcome to do it. ..I still haven't wiped my 0.9.5 scenery, so I can play tile mixing games with those 2 sceneries, if that can be of any use. ..on the 0.9.7 scenery, we used SRTM data, was the 1km data too used, or was it discarded? .."one way" is weigh those 2 elevation data sets gradually against each other across the N60|S60 borders, from say 59.45 to say 60.15, to make a 30nm wide "transition tile band" to smooth out the errors, we oughtta be able to make it look about as credible as the old scenery. -- ..med vennlig hilsen = with Kind Regards from Arnt... ;-) ...with a number of polar bear hunters in his ancestry... Scenarios always come in sets of three: best case, worst case, and just in case. _______________________________________________ Flightgear-users mailing list Flightgear-users@flightgear.org http://mail.flightgear.org/mailman/listinfo/flightgear-users 2f585eeea02e2c79d7b1d8c4963bae2d
