35 réponses à ce sujet

[Wallack]

No, thank you for helping

 

I ported 2 more helments:

 

This one still needs work with the .PLT, but this will be done this afternoon

 

 

This one needs to be bigger:

 

 

Just want to try this afternoon to follow the tutorial blindly

[OldMansBeard]

I looked at converting Skyrim robes into NWN a while back. Didn't really succeed but I can offer some thoughts. The basic problem to be addressed is that the skeletons are different. There are a number of possible approaches:

1. Discard the Skyrim skeleton and the vertex weights, add a NWN skeleton and paint the new weights manually
2. Ditto, but compute the new weights automatically from the old ones in some principled way
3. Keep the skeleton and the weights and construct new animation supermodels for the skyrim skeleton in NWN

Approach (1) is the only feasible option without writing special software. I tried adapting CM3 to do (2) but wasn't satisfied with the results, particularly for robes with knee-length kilts or skirts for which there are no bones in NWN. Approach (3) would be the best in the long-term because it would only need to be done once and then the resulting supermodels could be used for any number of robes. I had some success with adding code for approach (3) to CM3 but didn't develop it far enough to be widely useful as a package. I can say that software to generate supermodels for an enhanced skeleton from the vanilla supermodels, by adding bones and interpolating keys for them mathematically, is definitely possible but it's not simple and when I tried, after a few months my brain got tired and I shelved it.

[MerricksDad]

what toolset are you using to export the mdls you used for testing?

NwMax 0.8 b60

 

As you can see I also do not compile these models.

[Wallack]

After I finish all the .mdls for my haks, should I compile them?

[MerricksDad]

After I finish all the .mdls for my haks, should I compile them?

 Depends on your personal needs. Heavy models benefit from initial load time speed increase. I generally don't, so that I can easier debug stuff later without having to first decompile. It also lets me process stuff with a text editor by entire directory when they are uncompiled.

[Pstemarie]

I compiled a lot of models when I updated Q to v2.0. I honestly didn't see much difference in load time. What I did get instead were lots of broken models and several robes where the geometry got completely excised from the model (must be the model had errors). Overall, in my reformed opinion, compiling just doesn't seem worth it anymore unless you know the model is free of bugs and you won't be editing it anytime in the future.

[Rolo Kipp]

<being a feather...>

 

Those feathers are crying out to be danglymesh.

Just sayin ;-)

 

(and yeah, I like hawaian shirts...) <nobody's gonna get the reference>

Hah!

 

<...on the wind>

[Wallack]

<being a feather...>

 

Those feathers are crying out to be danglymesh.

Just sayin ;-)

 

(and yeah, I like hawaian shirts...) <nobody's gonna get the reference>

Hah!

 

<...on the wind>

 

How can I do that?

[Rolo Kipp]

<penciling in...>

 

I have two prior commits for the weekend (CCC and Elven's burning body), but I'll work on a little tutorial if I get the time.

 

Short version: Color the vertexes shades of grey (black is loosey goosey, white is stiff as a board - and yeah, only the red channel is supposed to count, but just do the shades of grey). Add an Aurora flexmesh modifier (winging the name as I can't fire up tools right now) and adjust the parameters.

 

So make the helm part all white vertexes (stiff) and paint the feather vertexes increasingly darker shades (with the tip ending up around 128 or so - more would be too wild)

 

<...a quick sketch or two>

[MannyJabrielle]

Wallak, those helmets and robes are looking hot!  I particularly like that last helmet.

[Wallack]

The robe won't look hot, don't know how to convert it into a real robe, so I'll keep with the helmets

 

Made a little bit bigger the last one to fit better the scale