Editing a world map

[Deadshade]

Try Tools>>Global Smooth>>Land Offset with an amount 2 instead of an amount 100 and see if that does anything. .

No it doesn't. In FT applied 10 times global smooth with amount 2 on the part selected between -4000 and -300 (the cliff happens approximately around -500 falling to - 3 000). It didn't blurr the smallest bit and the cliff is still there. Perhaps it doesn't work Under the sea - I can't think of anything else.
Will try to do something with mounds in Wilbur but if I can't find something, I'll simply GIMP it away. I know this is brutish but I hate discontinuities.

[waldronate]

I'm not sure what's going on. Unless, of course, you're using a binary input file and not using Burn Into Surface, of course. In that case, nothing will have any effect on your surface because binary data can't be edited in FT (FT can only edit data that's stored in one of the editing channels).

I whipped up a quick world with the same basic characteristics in FT, including that cliff.


I selected everything below -1000, did a border on that selection, feathered the selection and finally did a global smooth on it:


EDIT: looking again at your previous image, I see that there isn't any overshoot on the cliff, making it likely that you're using the binary input data.

[Deadshade]

I'm not sure what's going on. Unless, of course, you're using a binary input file and not using Burn Into Surface, of course. In that case, nothing will have any effect on your surface because binary data can't be edited in FT (FT can only edit data that's stored in one of the editing channels).



EDIT: looking again at your previous image, I see that there isn't any overshoot on the cliff, making it likely that you're using the binary input data.

Well yes. I imported the map from Wilbur . That goes only when starting with new->binary etc. So I guess it was (is ?) binary.
After that I edited it because I went to FT for the rivers so obviously the river creation worked (if this counts like "effect on the surface").
Then I saved the result in FT format (the map I joined above is the FT format).
And after that, in the multiple attempts, I always loaded the file saved in FT format.

So no, I didn't do Burn into surface at any step because as I understood the manual, it was only necessary if the original binary file (this would be the .mdr Wilbur file) was moved or deleted. It wasn't.

In the case I misunderstood the manual, I will test the smooth after having burned.

[waldronate]

A world created from binary data only uses the data from the binary file and it ignores the offset, prescale offset, and roughness channels. Saving an ftw file from a binary world just saves the name of the binary file in the ftw, not the data. Operations like raise, lower, smooth, and so on will have no effect on the binary data, just on the channels that are not being used. Operations that are based on information derived from altitudes (river finding, temperature computations, coloring, etc.) will use the binary data.

The Burn Into Surface operation samples the binary file into the offset channel at the resolution of the offset channel. The binary file is no longer used after the Burn Into Surface operation.

[Deadshade]

OK after having burn into surface I got 2 things.
- the smooth worked more or less. Not as well as in your example but there was an effect. It didn't cover the whole space between the selection boundaries like in your case - in some places it did in some other it didn't.
- It destroyed the hills, mountains and rivers. Mountains became these smooth worm like bumps. Basically it looked like if the height field lost most of its fractality and the hills are just hemispherical mounds.

So I guess it was from Charybda to Scylla.

[waldronate]

Burn In To Surface downsamples your surface to the currently-set editing resolution. If you don't change it, that means that you get the a 256-wide surface, which will look really ugly. The maximum allowed editing resolution in the 32-bit version of FT is 8190 samples wide (Map>>World Settings, Editing page).

[Deadshade]

Burn In To Surface downsamples your surface to the currently-set editing resolution. If you don't change it, that means that you get the a 256-wide surface, which will look really ugly. The maximum allowed editing resolution in the 32-bit version of FT is 8190 samples wide (Map>>World Settings, Editing page).

I came back to the map after a longer break. YEs, I missed the resolution bit (it is not mentionned in the FT manual in the short comment about Burn into surface).
As I said, I don't get the same result as you did. The cliffs still more or less stay on some places.
Anyway I simply GIMPed it over so that the visual problem is solved even if the heigght field one isn't.

A small question about precipitation field in FT. How is it computed ? Just a random function with some latitude dependence ? Or did it use a partial pressure function at saturation ?

[waldronate]

A small question about precipitation field in FT. How is it computed ? Just a random function with some latitude dependence ? Or did it use a partial pressure function at saturation ?

It uses a base value with a temperature (and thus altitude) dependendence as well as a random amount. See attached for the FT V1 manual section on the elements used. It leaves out the prescale offset things because they weren't in the system at that point.