A New Package of Dynamic Tools

[ShawnStovall]

I would like it if NewTek would dedicate a group of programmers to developing a set of integrated dynamic tools that are all interactive with each other. When I say this people might think that I'm just a newbie who can't find the dynamic objects in Layout, I'M NOT! I would like to see more intuitive cloth, as well as the addition of water and (NewTek developed) hair. For me, a poor hobbyist, this would be the biggest improvement for simulation in LightWave.

[3D Kiwi]

Ill back that up, if they dont make and effort with that soon, I will be one less lightwaver, The only thing keeping lightwave on my desktop is free render nodes.

[erikals]

Yup, but first the Dynamic simulation has to be improved. LW 9's improved dynamics aren't really very improved, AFAICS.

[Ztreem]

I agree, The dynamics in LW is not good enough and they don't work together in a effective way.

[OFF]

since v.5 we had strong positions on modeling, bones deformations, rendering, but still doesn't correspont to mordern requirements with dynamics(CLOTH!!!?) heh

[Mylenium]

Yupp. LW needs definitely some improvements on all Dynamics fronts. I don't think liquid/ fluid sim is a must have (way overrated because it's very specialized and not many average users will use it that often), but particles need to be massively sped up and have much better collision detection; rigid bodies must be motions, not displacement effects; everything must work with everything and so on... In the final analysis, maybe licensing PhysX or Havoc technology might not be that bad an idea even if it slightly raises the price.

Mylenium

[erikals]

i Havok

... and that is realtime?

[evenflcw]

You do calculations on high poly? Calculations on low poly will always be faster and if the plugin doesn't remesh internally then you should do it manually (and link high poly to low poly standin). What you end up rendering with will probably always be more detailed than what you actually need for convincing dynamics. Either way, there is no doubt in my mind Havok is faster, more accurate and more realistic than hard_FX. I say that never having used it though.

[Mylenium]

As far as I have seen, Havok is completely unuseable for a serious animation package. It's designed to handle extremely low poly stuff and the precision is awful. Good enough for realtime games, but I wouldn't want to simulate any physics for rendered animations with Havok.

Even so, it still beats LW's sim on some level. The thing is, I could live with fast, less accurate physics if it gives me the option to interactively adjust it and bake the motions easily. The time saved in the setup of the sim can be spent on tweaking and cleanups (which even with good solvers are almost always required).

At the moment LW does not give me any of that - all sim must be explicitly calculated, calculations take forever, collisions are very imprecise, one type of dynaimc cancels out others when applied to the same object and since it's displacement based, it's a pain to adjust later on using the quirky edit path tools. On top of that, all settings use very abstract values that in no way relate to realworld units, making it hard to find good settings in the first place. Lastly, even baked motions are awkward to handle as there is no way to time-stretch them, revert them or even re-calculate interpolation based on existing data from selected frames. They also use different data formats for each item type, so it's not possible to transfer e.g. particle data to softbodies, which is common practice in other programs.

If at least two of those criteria were improved - in my case precision and editing of motions - all the other shortcomings would be much more bearable.

Mylenium