#Pandora's Tower Wii ripping help

32 messages ยท Page 1 of 1 (latest)

clever dawn
clever dawn
#

Anyone?

clever dawn
#

trying again cause i still dont know๐Ÿ˜”

lusty creek
#

I can look at these later

clever dawn
lusty creek
#

The bin file only contains text

#

acc_dyn_pdt.ddb
srsc_dyn_HairFR.ddb
srsc_dyn_HairSR.ddb
srs_dyn_hairB.ddb
srs_dyn_hairBT.ddb
srs_dyn_hairF.ddb
srs_dyn_hairSL.ddb
srs_dyn_hairSR.ddb
srs_dyn_skirt.ddb

lusty creek
#

the atb file could be a sort of Aqualead ALTB Table

#

Dunno

#

This is what Dolphin Texture extractor says about the rest

#

Info: Scan "srs_brcA0.aig" recognized as Aqualead ALIG Image.
Info: Scan "srs.aod" recognized as Aqualead ALOD Object Definition.
Info: Scan "srs.0.ams" recognized as Aqualead ALMS Mesh Collision.

clever dawn
#

Aqualead?๐Ÿค”

lusty creek
#

Seems like a japanese vgm company

clever dawn
lusty creek
#

Could get this out

#

it's the aig file

#

mipmapped texture

clever dawn
#

yeah i knew the aig are the textures

#

but i wish i could extract the models

lusty creek
#

I think the models are the ddb files

#

ams is the mesh

#

the atb a sort of value table

#

Dunno

#

The model itself seems compressed

clever dawn
thorny blaze
#

I was looking at this game recently, and found some meshes in the .aar files. The scales of the raw submeshes are all over the place. I didn't find UVs or normals, but I'll write down what I found so far. Probably have to find position/scale/rotation listed for each mesh, or it's all put together correctly with bones and animations.

#

You can also use Ninja Ripper with Dolphin, but it's posed.

#

In an .aar file you can find the beginning of a mesh by searching for ALMS (41 4C 4D 53), then not far beyond that is a header with some info about the vertex block. It seems to be marked by 00 01 04 03 usually. The blue highlight in my screenshot is the vert count (05 41 = 1345 vertices). The green highlight is where the stride is stored (0C = 12), and right after that the 02 10 is the data offset, from the beginning of the 00 01 04 03 header to where the vertex data actually starts.

#

For faces, the indices start after the vertex data, in this case at 0x7700 (I'm using mct_.aar as an example). You can usually spot it where there's an 80 00, followed by 2 bytes that represent the length of this index section (5C A0) and then 98 00 marks the start of the tri strips.

#

In my screenshot, I'm reading the indices in sets of 3 (I'm guessing UV/normal/face), so for example 00 24, 02 BE, 02 BE is a set, where 02 BE is the face index. Then there's a 00 byte separator, then the next set of indices. A tri strip break is marked by 98 00.

#

I've seen some meshes that store indices as shorts, others as bytes, and also with/without 00 byte separators. Some meshes also seem to have more than 3 indices per set, so I'm guessing multiple UV maps, but not sure.