Re: Tree replace Speedtree model
Posted: Mon Feb 25, 2019 8:01 am
Great idea, I really hope we can eventually replace everything with 3D models. Mostly because I want to create a third person version of Daggerfall.
Tools for Daggerfall
https://forums.dfworkshop.net/
My frame rates totally bogged when I ran the mod. I have a super old Geforce card even though it's SLI. I like to run the distant terrain mod to get that huge world feel. Unfortunately, DFU became unusable at that point. It looks pretty. I had been wondering when someone was going to do something with speedtree.Midknightprince wrote: ↑Sat Feb 23, 2019 8:47 am I had to go from 4 terrain distance to 2 (it went from about 50+ fps to 5 or 6), i am getting 30+ with real grass maxed, and all mods/texture packs.
The trees show up in winter, but only in locations, took the modded tree textures out too (thanks for providing the #'s in descript), spring/summer/fall seems to work fine.
So I'm thinking if it lags this much with a handful, all of them may be super demanding, they look good though (shadows, they fit in, and they even blow in the wind).
This.jayhova wrote: ↑Wed Feb 27, 2019 4:24 am I'm not a big fan of the current way tree motion is done in unity and similar environments. Tree motion IRL is pretty subtle. Mostly small branches move while to rest of the tree is still. the same thing goes for grass 90% of the time grass is quite still with only a very small amount of motion perceptible.
This is a very general criticism of the motion technology. In simulation, the tree branches tend to be monolithic clumps like cotton balls. In nature the parts that move the most are the leaves out on fine twigs. From a distance this motion is virtually invisible. Most grass and tree simulations have large parts of the foliage moving. Most days the wind is quite gentle and the motions are very small. You can think of this as a representation of how the plant resists the energy potential of the wind. As the wind presses against the plant the first parts to absorb the energy of the wind are the smallest and lightest parts on the outside. If for instance you were sitting in the tree you would feel far less wind than if you were on the outside of the tree. This is because the outer parts of the tree have already absorbed a good deal of the energy. The outer areas of the tree have also converted the motion from unidirectional to turbulent due to the nature of leaves and their tendency to rudder in the breeze. Even when the motion in the air is sufficent to move large areas of the tree, the most extreme motion is on the outer portions of the tree.pango wrote: ↑Wed Feb 27, 2019 7:22 am But currently movements are not tied to local wind yet, so I assume that's why it's too strong. I still give the benefit of the doubt and hope movements will be more subtle when they're tied to wind, but we'll have to see how movements "scale" with lower winds, will they look more like your video...
Also, it means that most of the time (that is, unless it's noticeably windy), perfectly still trees would still be a very good approximation of real world, if that helps with performance...