I have two 2048 x 2048 terrains in .BMP format that I want to combine. When I load them via File Input devices they only occupy a 512 x 512 region of the 2048 world size. The node map I’m using is exceedingly simple, just the two file inputs, a combiner (set to max), and a file output. I set the world size to 2048 as in the original files. The output is a mostly flat (height = 0) area with the small 512 x 512 combined heightfield down in the lower left corner (origins at 0, 0).
Can anyone tell me how to combine these files at their original 2048 x 2048 size?
How do I set them to be 100 percent of their original size? Adjusting the “Specific scale” slider so that they fill the entire 2048 area stretches them out and makes them look squashed, stretched, and unnatural. I don’t understand why the files are not loaded at their original 2048 x 2048 resolution (?).
OK, I’m still confused-- forgive me for being so dense. Here is what I have:
Two bitmaps generated in WM, each with: file size 2048 x 2048, largest world size slider at 31.51 km, detail set to 17.827 km/pixel.
When I load those via file input devices, the closest the Specific Scale slider can be set to the largest world size is either 30.72 km or 34.48 km, but not 31.51 km. Is that what you mean by “the size they were meant to be?” When I set them to one of those sizes they only occupy a portion of the new 2048 x 2048 space, AND the pixel spacing looks wrong, like it’s unnaturally stretched out.
The earlier comment that I have to increase terrain resolution to 4096-- set which to 4096? The original bitmaps? I can’t do that because I run out of memory trying to build them (only 1 Gb available). Do you mean that I have to build the terrains at 4096, then combine them at 2048? Why?
I CAN output the images as .ter files and combine them with a simple MAX function in TG, but I want to be able to manipulate them first in WM. I’m going this route because I don’t have sufficient memory to built both terrains simultaneously and combine them in one pass.
Again, forgive me for my inexperience, but could someone explain the steps for doing this in WM in baby talk? Thanks!
I’d still like to figure this out, but the immediate urgency is removed-- I just discovered how to turn off memory conservation. This allows me to build both terrains and combine them in one pass, albeit a long one. But it works, and keeps everything at the expected resolution.
one thing that might help is the knowledge that WM doesn’t fundamentally work in a resolution-space; it works in world-space. different resolutions simply give more or less detail to the terrain in world space. So when you import a terrain using the File Import device, it will occupy the same space whether it is a 128x128 file or a 1024x1024 file.
With that in mind, you would:
Output the files at any resolution. Record the vertical height and world size in the world sizing dialog.
In the new network, that you wish to read them into, make sure you set the vertical height and the world size the same as in #1.
Create your file input devices, load the terrains, and set the sizing slider in the file input device to be the equal of the world size you set in #2. You may have to pan the loaded terrain as well with the transformation part of the file input dialog, although this is probably not necessary if you just want full-scene terrains.
The terrain produced by the File Input device will then be the exact same terrain in height, slopes, etc as before.
If the terrain looks different, it may well be – you said you were using the .BMP format. BMP is an 8bit format which doesn’t contain enough range to adequately describe the heightfields. If you use .TER, .R32, or another high range format you will achieve identical looking heightfields.
You can’t… You need to know what size they were meant to be, and put that value in the slider… WM doesn’t guess the pixel spacing from an image.