Maybe this is the wrong category to put this in, I apologize in advance.
I’ve noticed that using the level of detail option in blender allows for you to use a separate lower poly mesh as a replacement for a higher poly version at far distances. I’m curious if something similar is possible, but instead of the mesh being substituted based on the distance, but the angle at which it is being viewed so that at each angle there’s a different mesh.
Here is an example how it would look similar to:
Not using the inbuild LOD system, but this would be easy enough to do in python.
I figured there likely wouldn’t be a function built in for this. but that’s good news it’s possible and easy[depends haha…]
As far as I know using python is writing a script for this then. Is this section of the forum appropriate to ask how I’d go about doing the angle changing script in python?
Look at that:
@Monster,
That’s really cool. Probably works awesome for effects.
but blender seems to really hate transparency for the most part. I can’t even get the hair I intend to use this for to display properly in the game only in viewport. So I’ll likely have to do figure out how to do a python script that works very similarly to the LOD option.
But thanks anyway.
Semi-transparency is always an issue especially when they intersect each other.
This is as this kind of pixels break the assumption that the render order does not matter (as on opaque pixels). This requires sorting. Sorting is slow … really slow. To make it even possible it is assumed the faces can be sorted, which is not necessarily true (e.g. intersecting faces). Something like hair tend to be used intersecting. This means you need all of your artistic skills to create the look you want to get.
If blender can show hairs and other transparency using items perfectly fine in viewport glsl, I’d think in the game engine it wouldn’t have any issues. Not to mention the fact that even in viewport the layers of the hair ALREADY have to be arranged to show properly in the first place. It doesn’t make much sense to me. Transparency is extremely useful to avoid using much higher poly objects in many cases. Given that I’d assume it’s most efficient to have lower poly models, I guess transparency somehow is even slower than that…
But I think I’ll just use viewport rendering, and make use of a camera along with shapekeys to achieve an angle changing effect.