Symbols in Area
mike robel
🖼️ 15 images Surveyor
I am playing with Symbols in Area along to model different tree densities in different size polygons. I'd like to be able to establish a density so that regardless of the polygon size, the density is consistent.
If all areas are the same size, it is easy-peasy just change the number to get the effect you want. But different size polygons require different amounts to fill the polygon.
One work around is to use a set number, say 5000 symbols and reapply it until it looks like I want.
I tried a little more controlled experiment.
Test 1 shows rectangles in which the density increases from top to bottom and from left to right the rectangles are different sizes. As you can see, each rectangle gets a different number of trees drawn in it and the density in a particular row looks fairly consistent. Interestingly enough, the number of trees is never equal to the number I specify and is different in each rectangle. For this test, it was 40, 400, 1000, 2000, and 4000 trees from top to bottom. In general, I think it looks okay.
Test 2 shows polygons with the same parameters. The numbers are different than the first test, which is to be expected since the area is different, but still not putting in the numbers I put in.
Test 3 shows where I set the number of trees to 40 and applied it to each rectangle. Note all are different (which with randomness should be expected.
Test 4 shows a different map, much larger than the three tests, and more complex polygons. Each fill was the same and I specified 1000 trees. The other settings were the same as for the previous tests. Note the number of drawn trees is much, much less than 1000, the trees are larger than the other tests.
The last shot shows my settings.
I would have thought that the scale might be different because it is a different, larger map. But why are so few trees drawn? Some of the large polgyons overlap with each other instead of being one huge polygon.
If all areas are the same size, it is easy-peasy just change the number to get the effect you want. But different size polygons require different amounts to fill the polygon.
One work around is to use a set number, say 5000 symbols and reapply it until it looks like I want.
I tried a little more controlled experiment.
Test 1 shows rectangles in which the density increases from top to bottom and from left to right the rectangles are different sizes. As you can see, each rectangle gets a different number of trees drawn in it and the density in a particular row looks fairly consistent. Interestingly enough, the number of trees is never equal to the number I specify and is different in each rectangle. For this test, it was 40, 400, 1000, 2000, and 4000 trees from top to bottom. In general, I think it looks okay.
Test 2 shows polygons with the same parameters. The numbers are different than the first test, which is to be expected since the area is different, but still not putting in the numbers I put in.
Test 3 shows where I set the number of trees to 40 and applied it to each rectangle. Note all are different (which with randomness should be expected.
Test 4 shows a different map, much larger than the three tests, and more complex polygons. Each fill was the same and I specified 1000 trees. The other settings were the same as for the previous tests. Note the number of drawn trees is much, much less than 1000, the trees are larger than the other tests.
The last shot shows my settings.
I would have thought that the scale might be different because it is a different, larger map. But why are so few trees drawn? Some of the large polgyons overlap with each other instead of being one huge polygon.
Comments
Any thoughts on why test 4 is so different?
I wonder if it possible to specify a density per some area (in^2 or cm^2) rather than specifying the number of symbols?
Thanks
The "Random" placement type follows exactly the same logic, except that the candidate point is positioned randomly within the grid area instead of exactly on the grid point. After the candidate position is chosen, the position is tested for validity and discarded if not valid. What that means is that you may have specified "1000" max symbols and a "random" layout pattern, but you may get far less than that number of symbols if your polygon has a complex shape with multiple holes. This behavior is pure programmer laziness, allowing the same loop to operate for all cases and not requiring a special loop for random placement.
It's not possible to directly specify a density of symbols. The grid distance should get you that sort of thing, though. You can get something similar to that idea by using a pattern where the random HV distance is similar to grid HV distance. This type of layout will ensure that the overall density is roughly that of your grid, but the randomness will hide the actual grid value.
If your heart is truly set on having random placement and you want a minimum distance between symbols (that is, something like "no closer than 5 meters", consider using the Minimum Distance feature of fill with symbols. During the symbol placement phase of the operation, the command makes a check to ensure that no already-placed symbols are closer than the specified distance before it drops a symbol (if there isn't a good place, it will discard the candidate symbol). It only checks against already-placed symbols rather than all symbols on the map, unfortunately.
On the large map, setting HV/DV to be the same as the random distances had a small effect, but not enough for what I wanted to do. Changing the scaling factor to 5, setting minimum distance to 0 and smoothing from 0.5 to 0.05 increased the number of trees drawn, as seen below.
That being said, I guess for my smaller maps which take up a space of <=10 sq km, the way I did it will be good enough and I can just fiddle with it. Certainly it is better than placing every tree individually or perhaps making a group of them and copy/pasting them individually.
For larger maps, passed on scales of 1:50000 to 1:250000 maps individual trees are not particularly needed and I can indicated density with color or transparency.
Here's the steps.
1. Select Draw|Fill with Symbols.
2. Select Browse. within the Profantasy Directory it goes to Profantasy/CC3Plus/System/Fillers and a list of FSCs is displayed.
3. Navigate to Profantasy/CC3Plus/Symbols/Castles/CA149 Vegetation.fsc and push open.
4. Program returns to Fill with Symbols and the FSC is displayed at top, but nothing has changed in the display of symbols.
5. Select Load. Program returns to Profantasy/CC3Plus/System/Fillers and displays a list of symbols ending in FIL. There is another directory named annual and more are shown, but not from CA149.
6. Navigate to Profantasy/CC3Plus/Symbols/Castles and note only directories are displayed. selecting other directories also shows nothing.
The symbols that are displayed in Symbols in Area all appear to be *.PNG while the symbols loaded with Fill with Symbols all end in *.FIL.
I took screen shots, but I think the steps are clear.
It's not just that catalog. The only symbols that appear to show, at least in my experience, are those in the System folder. If you look in the browse input box, you see the address is preceded by an "@" sign.
I have also noticed that not all symbols in a catalog show up when you search for a *.png. For example, the symbols in Profantasy/CC3Plus/Symbols/Modern/Overland/Overland.FSC are not displayed during a search. They look like this:
Now I am retired and work on my own boardgames.
- The SYMFILL command which Joe talks about is the command for Symbols in Area, NOT Fill with symbol (That's the FOREST command).
- *.FIL files aren't symbols, they're just saved settings for the Fill with Symbols command. The Load button in the Fill with Symbols dialog is not how you select a symbol catalog, that is done via the browse button at the top. These .fil files only exists for a few predefined settings, and does not cover every symbol catalog out there. You can make your own .fil files if you need to, just hit save in the dialog after setting things up as you want them.
- For Fill with symbols, browsing to a new catalog isn't supposed to change the list under symbol name. Those are not the symbols in the catalog, they are the symbols that have been picked for use with the command. Click the ... button to pick a symbol from the catalog you selected with browse for any given spot in the list.
- Sue's right regarding the "missing" symbols, they are indeed vector symbols.
- Sue, there's nothing wrong with your catalog (as far as I can tell)
The FOREST command is for filling things with a hierarchy of symbols that is designed to minimize the number of symbols, while SYMFILL is a generalization of the ESC command to two dimensions and fills things with symbols from a symbol catalog without regards to hierarchy (but allowing random selection, variable alignment, scaling, and so on). FOREST is the original filler command, while SYMFILL is a relatively recent addition. I normally don't encounter FOREST directly (it's usually part of a drawing tool done by someone else), while I'm much more likely to see the SYMFILL GUI.
Each of the commands has a very distinct niche and they can get significantly different results.
Drawing the contours by hand would get you more organic shapes.
ESCLOAD @System/Fillers/CC3_Orchard.symfill
SELSAVE
SELBYP
SYMFILLM
SELREST
You might also want to load the fill setting CC3_Orchard.symfill (Draw > Symbols in Area) and check that it has the correct symbol selected ("Decid Tree fruit n").