Print

Please login or register.
1 Hour 1 Day 1 Week 1 Month Forever
Login with username, password and session length

[WxTech]

[Edit: fixed typo in file name in thread title]

Perhaps I'm far behind the times. I very recently installed an early version of microdem, probably as used by the early map modders back in the 4.08/4.09 days (which is where I'm still at!). It includes a conversion file for us Il-2 map makers.

Out of curiosity, with a spreadsheet I opened the .dbf file (as named in the thread title) that's used to apply the grey scaling for map_h.tga. I wanted to graph the relationship between elevation and the RGB. A range of elevations is assigned to a particular RGB value; of zero width up to 64m, then increasing in range with the increasing elevation.

For example, elevations between 1009m and 1024m are assigned RGB 192.

In the graph I noticed one really big departure from the trend, and two much subtler departures. Here are the errors, which you can change with any basic text editor:

The bad values are found in the first of the 5 columns, identifying the MIN elevation in the range. (The other columns are; MAX elevation in the range, the Red value, the Green value and the Blue value---all color values equal for grey.)

#1 For the range 424 to 424, the first should be 417.

#2 For the range 555 to 560, the first should be 545.

#3 For the range 2347 to 3520, the first should be 3457 (huge error!)

In your text editor, simply search for the bad number via Find (CTRL-F). Only in the case of the "424" will the number appear twice; replace the first.

All this presupposes the same file is still in use...

[Uzin]

In the file ELEV_COLORS.dbf of 11.4.2002 I cannot find the errors you are speaking about, see the table:

Code: [Select]

MIN MAX RED GREEN BLUE
-30460 0 115 115 115
0 76 113 153 89
76 152 117 170 101
152 305 149 190 113
305 457 178 214 117
457 609 202 226 149
609 762 222 238 161
762 914 242 238 161
914 1066 238 222 153
1066 1218 242 206 133
1218 1371 234 182 129
1371 1523 218 157 121
1523 1675 194 141 125
1675 1828 214 157 145
1828 1980 226 174 165
1980 2132 222 186 182
2132 2285 238 198 210
2285 2437 255 206 226
2437 2589 250 218 234
2589 2741 255 222 230
2741 2894 255 230 242
2894 3046 255 242 255
3046 4569 255 255 255
4569 6092 219 240 255
6092 30460 255 255 255

The file may be d/l here:
https://www.mediafire.com/file/4im1i9cqn1v6heh/ELEV_COLORS.dbf/file

[WxTech]

Uzin,
You're highlighting the *color* file.

The problematic one is the greyscale one (all RGB values the same on each line) used to make map_h.tga.

[Uzin]

Sorry, you are right,
 in the file ELEV_COLORS Il-2 Sturmovik.dbf of 31.10.2009 there the  mistakes you pointed out are present, indeed.
What to do?
 What is the effect of repairing the mistakes?
Any clue?
Are all the former maps sacrificed by these mistakes?

[WxTech]

Uzin,
Just before reading your latest, it dawned upon my feeble brain that I made a typo in the file name (but *that* I'll blame on my possibly dumber phone ).

As to the consequences of the file's incorrect values for the MIN elevation of a range... The first two are minor and the third is a biggie. But that third one applies only to pretty high elevations, and likely would not factor into the majority of our maps. And the first two errors would probably introduce visually inconspicuous artifacts.

I wonder if the software has some 'smarts' built in order to account for any such discrepancies as typos in conversion files?

One way to check on the effects would be to generate an artificial landscape DEM of an exactly smooth, flat, tilted plane, then see how this overlay conversion might introduce thin 'stripes' at the relevant elevations.

[SAS~Ghost129er]

Quick question, is that why I have these really weird red spots on the map in MicroDEM..?

[WxTech]

Ghost,
Could be. If you were to pick a couple or few pixels immediately adjacent to these spots, and find at least some are not far from the RGB grey values noted above, that would be a good clue.

[SAS~Ghost129er]

Hi Wx,

Yeah, that's what I had done - at first I thought it was missing data, but clearly not, because missing data is just transparent/a white square. So I did just that, thanks for the heads up though.