Yep, I haven't seen a modifier yet that whose deformations couldn't be captured by animesh. Lowering the number of vertices will (of course) lower the amount of data. Same thing with lowering the number of samples but NWMax's MaxScript is very touchy- I don't think 1 is an acceptable value so always make sure you put in something 2 or greater. After that the spinner works nicely.
While it can generate a lot of data at with a higher number of samples you can tune the sample number
down until it looks right in-game then call it good. You'll know when you've gone too low for a particular animation because it simply won't look the same in-game as it does in Max. No point in having more samples than you actually need, anyway. It just generates needless data that the tweening will approximate just as well- at least if you didn't tune it down too much.
If you get a sanity check error from NWMax about too many verts, try your model anyway. It'll probably work or if not try compiling it first. Compiling your model in advance also usually reduces the size.
This one seems to prune out static vertices and just keep important data. I whipped up this model and it started out at 1.45 megs but after compiling was only 445k:
At least that's my assumption about why the size dropped so much. (?) I didn't tune the samples down and the original capsule had way more verts than necessary. If you've given a bit of thought to a model you want to use animesh on, you can get the size down pretty damned low.
If there are any GMax users out there, you can do this stuff with GMax just as easily though GMax doesn't have as many modifiers as full blown Max. Sen started a
thread on the topic in the Max user's forum which has useful information for those just starting out with it.
Modifié par OldTimeRadio, 03 octobre 2011 - 06:02 .