Please correct me if im wrong but i believe Key-frames are faster than armatures in general correct? If there are any exceptions or if im completely wrong please tell me.
Thanks ahead of time.
By key-frames do you mean shape keys? You need to use key-frames for armature animation too.
I’m not sure what the speed difference is, but usually armature animation takes less memory.
Yes, yes i mean shape keys. I get mixed up with some of the similar sounding terminology.
And what i mean in terms of speed is for example if i were to make an animated plant or building piece, or just about anything else, which would be the best over all best way to animate many moving things onscreen at once? I see many people using shape keys for things like simple plant animations, but I’ve manage to fully animate a character with shape keys alone. I’ve done the opposite too with simple plant animations being animated with armatures. I don’t care how long it would take in the end, having the fastest frames per second is what i care about the most.
once again thanks ahead of time.
It would depend on the number of vertices vs. the number of bones. It also depends on how well each one is coded in Blender (for instance, with armature animation it is possible to have a lot of the work done on the graphics card).
The only way to be sure is to do a test with hundreds of objects, but I would guess that in most cases shape key animation would be quicker.
Thank you, as you know there many moving objects in an RTS or MMO and efficiency is important. i would like to know though what you mean with number of vertices vs number of bones. Some of the models in my game range any where from 20 to around a thousand with the later appearing very little. Do armatures increasingly get less efficient as the amount of vertices goes up per bone?
I haven’t done any tests to see if Blender is unusual… I just use bones for things that make sense like moving limbs; and shape keys for things like changing the overall shape of something.
The benefit of bones is that you can move many vertices in your mesh by just moving that one bone. That’s why they’re more efficient memory-wise as well. But if you don’t have many vertices compared to bones then you don’t get that memory savings. An extreme example is having a bone for every vertex–that would be terrible!
The other issue is multiple bones influencing the same vertices. To get good deformation you will often have areas of your mesh that are influenced by several bones. That can hurt performance.
I did some tests and the performance is really similar for what I tested ~200 vertices ~20 bones. I would suggest just using whichever is easier.