Well, that stuff looks great, but it's an apples to oranges comparison. 95% of that stuff uses full 3D deformations, but heightmaps in spring only support deformations in one dimension (Z, assuming Z-up). It might be possible to do some of that stuff in spring, but some of it is impossible and most of it is pretty much impractical from a performance standpoint. I'm not aware of any spring map with that level of detail or special 3D features like rock arches either.enetheru wrote:sigh. Have a look at this thread and tell me again that 3d applications like z-brush/mudbox and to a lesser degree blender is not good for terrain...
http://www.polycount.com/forum/showthread.php?t=125857
Meh, well there are several reasons why that might be. You might/probably have better hardware than I do, you might be using a different blender version, or perhaps better settings, but I haven't been able to get it to perform well given any of the settings I've managed to tinker with so far. I wasn't able to paint sea of dunes in blender hardly at all, even at 1/4 the vertex count and given the relatively small contrast in the terrain height.enetheru wrote:This has never been my experience, it handles cases like this easily and paint response is fast. also painting large terrains by hand is a PITA, use procedural materials.
I dislike procedural texturing, it tends to create maps that are ugly and/or generic. There are a few exceptions where you might be able to get away with it, but I consider it to be lazy at the very least. I agree about the PITA part though, but super huge maps aren't very playable/don't get played anyway. Another exception would be
EDIT: Well I read through the old thread on cattle and loveplay and now I understand that WM does that because smoth made it do so and for no other reason. Also idk, WM may have a free version, don't know how usable it is though.
There are probably lots of ways to create that effect. I'm not sure that extruding anything in blender would be my first choice though. Editing a height map in SME or something along those lines is more like painting, whereas modeling in blender is more like working a complicated machine. Specifically though, a machine designed to manipulate things in 3 dimensions whereas we only need one here.enetheru wrote:The setup is very simple, and does render fast because it doesn't require bells and whistles. has more control because you an use more than a base grid mesh to create the z-depth pass. for instance rocky mountains need a road? simply extrude a path down a spline and you get a smooth base
My point was that a tool for making spring maps could be better for making spring maps by specializing for the limitations and possibilities that they present, compared to a tool made for more general work, much of which is extraneous for map making. Not to mention the extra skills required for using programs like blender; just figuring out how to make texture paint work for maps was an incredible chore and required learning most of the interface. Importing maps is still a chore, for that matter, and can be a pain if you happen to change parts of it in other applications.enetheru wrote:The premise to build a toolchain with plugable components is more valuable than shoehorning everything into one tool and doing an average job of it.enetheru wrote:loading a heightmap and info maps from an image on demand
Find out what the industry vets are doing and do something similar.
Avoid NIH
There are certainly instances where external tools are better or more appropriate, but at least for terrain and texture you can't really get much better or more WYSIWYG than an in-game editor based on the actual implementation the game uses.
Be careful what you wish for.enetheru wrote:And so many sighs... I wish I could simply mind meld and get all this back and forth over with.
