only 8 zero bytes when unset, this is only set when there are bone(s)? SEE s_t2fsDataModelInfo
array count is the same as amount of poses/orientations, to is mostly absent.
model_1_0: same child count as model_1_2, model_1_4, model_1_5
only 57/2430 models have this
same child count as model_1_0, model_1_4, model_1_5
model_1_2_x: child count might be different per sibling
model_1_2_x_x
model_1_2_x_x_0: first u32 is texture-id
Read as 16 bit values, what to do depends on whether what flag is set. When bit number 16 is flagged, the next 15 bits are 3 indices, 5 bits each, so max 32 verts per indice block, thus we have to keep a vert offset. When bit number 15 is flagged, bit numbers 1 to 8 are the bone id. When bit number 14 is flagged it indicates a new indice block, bit numbers 1 to 5 is the vert count to add to a vert offset after this block has been processed, bit numbers 6 to 10 indicate some operation that is currently not fully understood yet, the seen set values are either 0, 11 or 22, it has to do with using vert indexes outside of the current inside block, when set to 0 there seems to be no issue.
some indexes?
model_1_3_x_1
node/bone
same child count as model_1_0, model_1_2, model_1_5
possible hotpoints, only 116 entries on all models but looks interesting
same child count as model_1_0, model_1_2, model_1_4
only 607 entries on all models, all set to 0000 0000 0800 0000
model_2: variable count equals pose count
model_2_x
almost always has count, but most times all values are just 0x0000, counts differ
count is the same as there are bones, these appear to be two 16bit values, possible signed. Maybe rotation count and position count? These are either set to 0xFFFF or seem to count up starting from 0.
model_2_x_2
array count equals bone count
Mostly unset
Mostly starts with 0x00000000, lots of same values for all models, some variation
based on data sizes and some repeating values this is likely bone rotation data, however I fail to make sense of the data, it looks compressed or maybe some sort of other bit stream. all start with 0x71c35b87 or 0x71c35b07, these bytes reoccur in large files, sizes between them are weird (some are even others are uneven).
based on data sizes and some repeating values this is likely bone position data, however I fail to make sense of the data, it looks compressed or maybe some sort of other bit stream.
0x1C, mostly unset