Celestia Forums

Hi All,

I made a normalmap uing the windows verion of nm16 and was happy with the results
except for one thing. I'm not sure if it's the data or nm16 (window version), but a mark
is put on the vertical edges that show up as either lighter or darker depending on light direction.
Has anyone experienced this? Is there a way to correct it?
The data was downloaded from " http://www.ngdc.noaa.gov/mgg/topo/gltiles.html ".
This thumb shows the mark and a couple of other thumbs that show the output from nm16.









cartrite

Cartrite,

I find your normal maps generally very crisp (which is not trivial ro get) . But I don't understand what mark you are talking about. I do remember that the Windows port of nm16 (done by someone, not Chris) had a bug. I think I spotted that bug and may have written about it somewhere in the forum a long time ago. ==> Search engine...

Cheers,
Fridger

Fridger,

The top thumb shows a black line. While I'm in Celestia over this area, the line will turn black (as shown). Then as the light direction changes to overhead the line dissappears. Then it will get lighter as the light direction changes further. This line corresponds to the vertical edges of tiles C1 and D1 where the world is divided into 8 tiles A1 thru D2. The original tiles that I ran NM16 on where further divided into 16 sections. The windows version of NM16 outputs PPM's and I assembled these into the 8 tiles A1 thru D2 that were VT'd. The "black line" appears were the vertical edges of those tiles are.

cartrite

I've seen these posts. That is were I got the link for the windows version.
It was posted by pixel I think or guest. The original version of NM16 doesn't change the missing data to zero. That' why I tried the windows version.

hi the win nm16 is moer than likley my port
and there is a bug
on thr left side and along the top there is a row of single pxiles that are not remaped corectly
my "temp fix" is to clone the
# for the top
clone the second row over the first

# for the left side
clone the sceond collom over the first collom

this is by far not a great fix the code was old , the first realace of the code and i have not done a more reacent mingw build.

if someone else wants to do a mingw build that would be nice
i have swithced to linux

Thanks John,

I'll try this.

cartrite

Some closing notes on this thread,

The problem is on the right side of the output from windows nm16. That is column 8191 (or the last column on the right) on my 8K textures. All tiles. A1 thru D2. I copied column 8190 to 8191 and it looks good so far. It will take a while to VT them again so I'm not sure. If there is a problem I'll let Ya all know. This NOAA data source seems to be the best source I came accross on the In NET for normalmaps. It has 18 different sources.

Later, cartrite