Skin Partition

[Conduit]

A while back, it was posted that a spell was added to NifSkope that was supposed to autopopulate a skin partition, but no one could get it to work.  I highly doubt I'd be able to get it to work, but I'd still like to try it out.  So, how, exactly, do you use it?  I can't find it in the spells menu.

[tommyboy]

A while back, it was posted that a spell was added to NifSkope that was supposed to autopopulate a skin partition, but no one could get it to work.  I highly doubt I'd be able to get it to work, but I'd still like to try it out.  So, how, exactly, do you use it?  I can't find it in the spells menu.

Open the character in nifskope.
click on the body (male_basic or whatever it is) to highlight it.
In the left panel go to "niskininstance" under ninode/nitrishape.
Then at the top go to "spells"/"block"/"insert" go down to the "s" part of the list, and about fifth one down is "niskinpartition".
Good luck getting it to work.

[Conduit]

Yes, that inserts a skinpartition, but it's blank.  I specifically remember reading something about autopopulating a skin partition, which I took to mean automatically adding all the necessary data to it.  Was that a misinterpretation?

[tommyboy]

Yes, that inserts a skinpartition, but it's blank.  I specifically remember reading something about autopopulating a skin partition, which I took to mean automatically adding all the necessary data to it.  Was that a misinterpretation?

"autopopulate" is not quite what it does.
You have to choose or enter certain values (like num skin partition blocks, num bones) and it then works out others for you.
I posted all this in detail before, but frankly can't be bothered to work it all out and post all the instructions again, sorry.
Maybe you could check out the nifskope forums for detains.

[Symon]

(from a post by me, long ago elsewhere)
OK, this is what I think you do:
Left click on the NiTriShape (usually editable mesh), then right click and take mesh, make skin partition. A menu pops up giving options for number of bones per vertex and number of bones per partition. Example are given for some games, so we'll have to inspect FF3R meshes to derive suitable values or just experiment.

Poof, valid looking (to me) NiSkinParition. I still don't have enough offending meshes (to be honest I've yet to experience the issue) to test it yet.

I haven't had time to play with this and felt discouraged after Tommyboy tried and bounced.

Good luck!

[Conduit]

Thanks.

[Conduit]

Okay, I've checked this out, and I think there may be a programming bug in the Make Skin Partition spell.  Every skin partition block has number of strips set to 1 (it's set to 0 in every working skin partition I've looked at), and the triangles array has a length of 0, regardless of the number of triangles.  You can right click on the triangle array and click update array and it will be set to the correct length, but it'll be completely blank.  Also, the bone indices are completely different from the bone indices in working skin partitions.  Whereas every bone index in a working skin partition is [0, 1, 2, 3], in a partition made from the Make Skin Partition spell, the indices are [0, 0, 0, 0], [1, 0, 0, 0], or [2, 0, 0, 0].

All of this happens regardless of the settings used.

[Lunarman]

there might be something useful here

http://frworld.wikispaces.com/Fixing+a+Ghosting+Mesh

[tommyboy]

Okay, I've checked this out, and I think there may be a programming bug in the Make Skin Partition spell.  Every skin partition block has number of strips set to 1 (it's set to 0 in every working skin partition I've looked at), and the triangles array has a length of 0, regardless of the number of triangles.  You can right click on the triangle array and click update array and it will be set to the correct length, but it'll be completely blank.  Also, the bone indices are completely different from the bone indices in working skin partitions.  Whereas every bone index in a working skin partition is [0, 1, 2, 3], in a partition made from the Make Skin Partition spell, the indices are [0, 0, 0, 0], [1, 0, 0, 0], or [2, 0, 0, 0].

All of this happens regardless of the settings used.

I have seen the number of strips set to 0 in a "spelled" block (just alter the number of blocks then update the array, the lower ones are set to 0. They still wont work, but are set to 0).
Bear in mind that this was not made with FF meshes specifically in mind. You'll note that when you run the spell it mentions oblivion and Civ4, NOT FF. So I guess it defaults to the type of values used by them.
Also bear in mind that I did say it's not a completely automatic process. You have to enter some values yourself, and probably tweak others.(You can manually enter values for the bone indices, for instance)
However, if you can identify all that needs setting up manually, and which values vary from working FF ones, its possible that the html could perhaps be edited to default to settings closer to FF ones, I think (though I admit I have no direct evidence to support this theory).
The problem is, it's a LOT of work for something that only about four people have ever shown much interest in (probably because it doesn't work right, but it may just be apathy on the entire subject).
The other problem is that the settings that work for any given mesh wont necessarily work on any other. this is because the number of vertexes varies in most meshes (though a LOT are male_basic or female_basic derived), so the niskinpartitions will need to be redone over and over.
I'm not saying its impossible, nor do I want to discourage you.
I just know that when I stop to think about manually adjusting maybe 20-500 FF meshes, my head aches.
But maybe the spell can be fixed and gotten closer to automation.
That would be kewl.

[Symon]

Okay, I've checked this out, and I think there may be a programming bug in the Make Skin Partition spell.  Every skin partition block has number of strips set to 1 (it's set to 0 in every working skin partition I've looked at), and the triangles array has a length of 0, regardless of the number of triangles.

Interesting. I think 'no of strips' is to do with 'NiTriStrips' which is as alternative to 'all triangles' that Oblivion uses but FF3R does not. I'd guess that No of Strips should always be 0 for an FF3R mesh.

[Conduit]

Okay, I've checked this out, and I think there may be a programming bug in the Make Skin Partition spell.  Every skin partition block has number of strips set to 1 (it's set to 0 in every working skin partition I've looked at), and the triangles array has a length of 0, regardless of the number of triangles.

Interesting. I think 'no of strips' is to do with 'NiTriStrips' which is as alternative to 'all triangles' that Oblivion uses but FF3R does not. I'd guess that No of Strips should always be 0 for an FF3R mesh.

Yes, even for a mesh that has NiTriStrips instead of NiTriShape (the nazi tank, for instance), No of Strips is set to 0 in the skin partition.  The No of Strips problem is easy to fix manually, but there are hundreds of triangles and bone indices, so the problems with them are the real obstacles.  Would it be possible to pass this information on to the Niftools developers?

[Symon]

FF3R tanks uses NiTriStrips? Learn something every day! Interesting.

I'll do so once my evil, shattering, IRL schedule (office move that seems to be taking for ever) is over.