Dead or Alive series formats and tools

[b0ny]

doa5Uarchivarius(v17b) - for unpacking "Dead or Alive 5 Ultimate" archives. you may need to install python 3 to run this

updated. now it's version 1.7b (redownload)
new in version 1.7:
-added one more key, to try, if failing at decryption. this solves the keygeneration fail for "GENFU_COS_001.TMCL"
-added support for inflated chunks in the deflated stream. this solves the incomplete files problem, like "HAYABUSA_DLCU_004.TMCL"
-added more names for files that come with patches and are not mentioned in the original file list("Halloween","TROPICAL?","Yaiba")

new in version 1.7b:
-fixed a few files in filelist
-a little bit of a speed improvement

mikulover39 did a great job restoring the filenames for marie's costumes, though he overlooked two costumes that where created by team ninja with much love and only using the <ctrl+c> -<ctrl+v> keys.

also i couldn't reverse the right names for leifang's hair from the halloween set

[edit]

just checked the files integrity, for all tmc/tmcl files from original and dlc lnk, upacked with the version 1.7 - everything seems ok, "CIRCUS.TMCL" is even one byte longer then it's caption says.
these files are for the ps3 version:
ALPHA152_COS_001_A.TMC
EIN_COS_TEST.TMC
LEON_COS_TEST.TMC
MOMIJI_COS_TEST.TMC
RACHEL_COS_TEST.TMC

[mikulover39]

Thanks b0ny, but I seem to be getting "IndexError: list index out of range" whenever I try to extract a TMCL file

[Darko]

The new unpacker is not showing any name, lol.

Lol, never mind, I renamed the file correctly and it Worked. Stupid "U"

[logansan25]

When i tryed open files .TMC from pack DLC Dead or Alive 5 - Zack Island Stage, appear this:

http://prntscr.com/3c3o6w

[b0ny]

so, guys. have you tested the archivarius yet?
any errors to report?

i've found at least one. will fix it later.

logansan25
have you tried to unpack this file with the latest archivarius. the problem could be that the file is incomplete(fixed in latest version). i'll have a look...

[Protocol X27]

Awesome stuff, b0ny. No wonder you've been quiet lately,

The new archivarius seems to work fine for the basics. I'll start using that version and report back if I see anything.

On the note of archivarius, has anyone run across the MPM files for Private Paradise animations. I'm curious if those would be easy swaps based on the characters.


Also, b0ny, since you're in between projects, any chance you could help explain the nature of the coordinate transformation for the few models that have odd rotations. Since you fixed the Noesis script, I was wondering if you understand the nature of the hex to make those changes. I don't think we'll be seeing an update to the 3dsmax script anytime soon, so I'm thinking it may be possible to manipulate the hex data so the script will import properly. I'm struggling to get the right questions answered on coregrafx.

[moge1975]

Hello, b0ny

I'm using your [doa5Uarchivarius(17b)] tool.

9114F6154E035B4366604265E28D9048251AC1504B
4250ACFD1B15ED82E17B54580EE3AC43F76232854B

this DLCs all　Character not work

this is bug?

extract is OK
reimport is not work on 360
not Freezing, but. no loading....Character

other DLC is Perfect works..

[Darko]

Extracted the new dlc (nurse costumes) with the new ectractor and all is working fine.

[b0ny]

[moge1975]
this is doa5 dlc1 and dlc2, and it unpacks them both perfectly for me(could be that your archives are corrupted?)
but if you mean the file swap (injection), seems like there is a problems with injection in some costumes in doa5u in general.

i need you to give me more details on your problem...

[Protocol X27]
noesis is computing the transformation matrices on a high level of abstraction. i'm interested in low level(manual) computing for transformation matrices(and also weights) too. i'll need that soon, and if i won't find myself how to do this i think i'll open a thread asking people smarter than me for help.

i intend to do a doa5 to doao converter, and for that i need to set the model to the t-pose, for that i need to apply all the non-applyed matrices, then move the bones in the t-pose(its stored in the tmc file).

[Darko]

[moge1975]
this is doa5 dlc1 and dlc2, and it unpacks them both perfectly for me(could be that your archives are corrupted?)
but if you mean the file swap (injection), seems like there is a problems with injection in some costumes in doa5u in general.

i need you to give me more details on your problem...

[Protocol X27]
noesis is computing the transformation matrices on a high level of abstraction. i'm interested in low level(manual) computing for transformation matrices(and also weights) too. i'll need that soon, and if i won't find myself how to do this i think i'll open a thread asking people smarter than me for help.

i intend to do a doa5 to doao converter, and for that i need to set the model to the t-pose, for that i need to apply all the non-applyed matrices, then move the bones in the t-pose(its stored in the tmc file).

Regarding that, How is the DOA4 format going??

[b0ny]

How is the DOA4 format going??

the million-dollar question is "where are the joints(weights)". the mesh has weight data layers, which are empty by some reason. and the join's block in the cat file is deprecated for doa4. i have no idea what team ninja was really thinking...

also i have converted the mot files. and i think it's possible to convert 'bin' animation files. but i already failed in a way converting doa3 animations to doao, this is even more complex(but possible)

here you can try zack's outfit. you have to use the loader provided with the "DoAX->DoAO" converter(t's patched to support doa4 files/physics)

here is how doa4 models look in doao now: (why xentax doesn't support hiding spoilers?)

[Protocol X27]

[Protocol X27]
noesis is computing the transformation matrices on a high level of abstraction. i'm interested in low level(manual) computing for transformation matrices(and also weights) too. i'll need that soon, and if i won't find myself how to do this i think i'll open a thread asking people smarter than me for help.

i intend to do a doa5 to doao converter, and for that i need to set the model to the t-pose, for that i need to apply all the non-applyed matrices, then move the bones in the t-pose(its stored in the tmc file).

Cool, I was thinking that might be helpful to get someone more advanced involved as well. Thought I'd at least pick your brain first. Penguin has posted that the process somehow involved multiplying something times the inverse of the matrix. The problem for me is that I don't understand if vertex location values get multiplied for if the Offset matrix just gets replaced by the inverse or something. The language barrier has definitely been in the way. At least if there's more than one person interested maybe that'll improve the chance of getting assistance. I don't mind writing a script to perform calculations as long as I know which data to adjust.

[b0ny]

import_TMC-GMD.py - doa5(u)/ng3 import plugin for bleder

doa5ng3_360.py - doa5(u)/ng3 plugin for noesis

now you can import/preview the models that were "corrupted" before: GENFU_DLCU_002 and the faces for AKIRA/BRAD/JACKY/LEON/RIG/RTM

the dramatic story behind the fix:
i have found that the index for the "vertices_buffer" and the index for the "indices_buffer" is not the same thing. before, the index for the "vertices_buffer" were used for both, but some meshes have one "indices buffer" and multiple "vertices buffers", that was breaking the order and a wrong "vertices buffer" buffer were picked, this resulted in a corrupt mesh. now this is fixed, i have found the right variable for the "vertices_buffer" index.

[Protocol X27]

Holy @#)$*(@#)($*

How do I like spam on the thank you button more than once!?!?!! That even surpasses the alpha feature as far as usefulness! The ability to extract those other models is huge for XNALara stuff.

[Darko]

Lol This means: Brad Wong, Rig, Jacky and Leon complete models. Thanks.