Dead or Alive 6 - Modding Thread and Discussion

[perfectdark023]

On 4/20/2020 at 3:25 PM, DarkMissLady said:

Hello there and hello once again, I don't care for these mods until some starts doing FunnyBunny666 Mods on DOA 6 will come back until then, This topic is not really fun anymore too me same art context this why still play DOA 5 and DOA 6 mods are dull and boring here a link will show you on what FunnyBunny666 does with the clothing. https://www.deviantart.com/funnybunny666/art/Momiji-DOAXVV-Dark-Prison-827694837 if not doing this kinda work please just not good enough my validations are facts not here to give anyone the rude or the finger, Just FunnyBunny does this for fun that is it. And also if going to give me crap about this, Then just don't reply at all, everyday go too this guy getting those mods and there going to show up in doa 6 ever.  

I mod for fun too, but its hard trying to have fun with mods when people think you are trying to compete with them....when you're not. Plus I don't only do nude/skimpy mods, I work on mods that I want to do. Not do mods that gives me more attention, that leads to frustration when you are trying rush a mod out or an idea out before someone else.

 

 

"DOA 6 mods are dull and boring" because most modders who know how to port outfits flawlessly which leads into you having to know how to UVmap either don't tell you how to do it, they charge you money, they don't know themselves (they either tell you or don't tell you that they don't know) or they are super vague and you have to figure out what they mean yourself. There's only a few of us that still help, the others that used to help a lot are either dead or not modding DoA6 anymore because of the game itself  or no one helps them on what they need to know (they know how to do this one thing, but no one tells them how to do this other thing they need to know. So they leave and take their knowledge of modding DoA6 with them.)

 

 

I have permission to port FunnyBunny666 mods.....but I need free time and need to build up my skills for porting and I have a lot of mods on the back burner that's more important to me or more time consuming despite me knowing how to do it.

If you want to port other mods, you practically have to know how to do everything. I just learned how to UVmap after months of asking around. And Based on what I do, I have to know how to do everything. If I'm not doing something that's going to be a normal thing with my mods, I pass it to a different modder who normally does it and then combined our work together.

Spoiler

 

 

 

[perfectdark023]

On 4/21/2020 at 7:42 PM, fgh1t6 said:

Somehow I think maybe we have done too much mods that finally kill the game now.   

Yeah that's what I felt for awhile now too. Stuff like Breakables slowly stopped after we started to figure out how to make better looking destructibles and add color changes to other outfits, as if that was Koei and/or Team Ninja's responds to us. Even the DLC characters don't have breakables and Tamaki doesn't have any cool changing outfits too.

[vagonumero13]

2 hours ago, perfectdark023 said:

Yeah that's what I felt for awhile now too. Stuff like Breakables slowly stopped after we started to figure out how to make better looking destructibles and add color changes to other outfits, as if that was Koei and/or Team Ninja's responds to us. Even the DLC characters don't have breakables and Tamaki doesn't have any cool changing outfits too.

Personally, I don't think we are relevant enough for KT/TN as to influence any of their decisions. Their main market are consoles, and in concrete, the PS4.

And there, you can't even dump the game files (apparently this could only be done on older games), let alone, modding.

[用户名623]

34 minutes ago, vagonumero13 said:

Personally, I don't think we are relevant enough for KT/TN as to influence any of their decisions. Their main market are consoles, and in concrete, the PS4.

And there, you can't even dump the game files (apparently this could only be done on older games), let alone, modding.

Yes i don't think we can influence the game company's decision

[fgh1t6]

On 4/22/2020 at 3:58 PM, vagonumero13 said:

Yeah, I know, although once you do it once, it is easier, but maybe I will just add a specific tool to add meshes or something.

Founding may help for the tool developing:

In the <MeshesSection> section if   "Count2 value" is not 0, adding block to the end may not work.

In the <SubmeshesSection> section, each submesh block has a "<U_00 value="XX" >".  information, U_00 value="53" is the common type, counted by the  "Count1 value" in <MeshesSection> blocks.   U_00 value="61"  is special type, counted by the "Count2 value". In the  <MeshesSection>, all U_00 value="53" type mesh block should go first. That is the "Count1 value" counting. then followed by "Count2 value" blocks  for the U_00 value="61" type submeshes. If inserting U_00 value="53" type mesh, the corresponding <MeshesSection> blocks must insert before blocks for U_00 value="61" type submesh. For example if

            <Count1 value="1" />
            <Count2 value="1" />

There would be blocks: <Mesh idx="0">,<Mesh idx="1">, and if we are inserting a submesh with <U_00 value="53"/>: the two counting would be:

            <Count1 value="2" />
            <Count2 value="1" />

then the newly created block should be <Mesh idx="1"> not <Mesh idx="2">.

 

 

 

[vagonumero13]

15 minutes ago, fgh1t6 said:

Founding may help for the tool developing:

In the <MeshesSection> section if   "Count2 value" is not 0, adding block to the end may not work.

In the <SubmeshesSection> section, each submesh block has a "<U_00 value="XX" >".  information, U_00 value="53" is the common type, counted by the  "Count1 value" in <MeshesSection> blocks.   U_00 value="61"  is special type, counted by the "Count2 value". In the  <MeshesSection>, all U_00 value="53" type mesh block should go first. That is the "Count1 value" counting. then followed by "Count2 value" blocks  for the U_00 value="61" type submeshes. If inserting U_00 value="53" type mesh, the corresponding <MeshesSection> blocks must insert before blocks for U_00 value="61" type submesh. For example if

            <Count1 value="1" />
            <Count2 value="1" />

There would be blocks: <Mesh idx="0">,<Mesh idx="1">, and if we are inserting a submesh with <U_00 value="53"/>: the two counting would be:

            <Count1 value="2" />
            <Count2 value="1" />

then the newly created block should be <Mesh idx="1"> not <Mesh idx="2">.

 

 

 

 

That's a good find.

53 = 0x35 = 0x20 | 0x10 | 0x4 | 0x1

61 = 0x3D = 0x20 | 0x10 | 0x8 | 0x4 | 0x1

(I'll make that value be written in hex in next version)

 

So, it is the flag 0x8 that checks if it counts towards Count1 or Count2. If i see the info is consistent in the 100% of g1m of CE, I could make the tool reorder the meshes in the MeshesSection if needed, as anyway <Meshes> don't seem to be referenced by other entities (but I'll have to confirm that), aswell as omitting Count1 and Count2 and have the tool calculate it.

[XUTERO]

On 4/13/2020 at 10:26 AM, kollector said:

 

 

Thanks!

 

So, about the data transferring ,

 

I believer you also transfer the bevel weight, don't you ?

 

Because i only transfer the vertex groups.

 

 

 

 

 

I'm having a skinning problem even after transferring weights (vertex groups) as in this video; https://youtu.be/zWE0xP4MgR8?t=2546

Should I be using a complete naked body like the "White Lace Body" to transfer weights? If so where can I find it? (I'm transferring from a half upper body only)
Also, I transfer weights (vertex groups) from imported Raw Buffers instead of from a Frame Analysis. Not sure if it plays a role? 

Perhaps some skinning parameters in blender are wrong? I'm not familiar with this particular 3D software. 
 

Otherwise all the rest (Mesh replacement, UVs transfer, Texture replacement) have been working like a charm. 
I'm almost there. 

[fgh1t6]

1 hour ago, XUTERO said:

 

 

I'm having a skinning problem even after transferring weights (vertex groups) as in this video; https://youtu.be/zWE0xP4MgR8?t=2546

Should I be using a complete naked body like the "White Lace Body" to transfer weights? If so where can I find it? (I'm transferring from a half upper body only)
Also, I transfer weights (vertex groups) from imported Raw Buffers instead of from a Frame Analysis. Not sure if it plays a role? 

Perhaps some skinning parameters in blender are wrong? I'm not familiar with this particular 3D software. 
 

Otherwise all the rest (Mesh replacement, UVs transfer, Texture replacement) have been working like a charm. 
I'm almost there. 

Looks like a bone mis-matching to me. Did you extract VB/IB with g1m_export_with_vgmap ?

[XUTERO]

2 hours ago, fgh1t6 said:

Looks like a bone mis-matching to me. Did you extract VB/IB with g1m_export_with_vgmap ?

No, I extracted VB and IB with the normal "g1m_export.exe". I am unsure what the "g1m_export_with_vgmap.bat" is supposed to do? 
What are VG maps, what and how are they used? 

[gokuri]

I've got linkdata export making empty folders sometimes, scifi AYAHAIR036 for all the kidsalb.g1t that aren't the actual hair, and AYACOS011_c01_kidsalb which I assume is where the gold trim bits of the pirate outfit are.

[vagonumero13]

39 minutes ago, gokuri said:

I've got linkdata export making empty folders sometimes, scifi AYAHAIR036 for all the kidsalb.g1t that aren't the actual hair, and AYACOS011_c01_kidsalb which I assume is where the gold trim bits of the pirate outfit are.

 

Look at the file sizes of those .g1t. If they appear as 1 KB (it is actually less), then, they don't contain actual image data, and LINKDATA can't extract them.

These .g1t contain just color information, in RGBA float format. To edit these, you are going to need an hex editor and a calculator to convert from typical color hex format to float format (although there may be websites for that)

 

Edit: i looked at the pirate .g1t, and it is indeed like that. It has the color data for pure black with full opacity: R = 0.0, G = 0.0, B = 0.0 and A = 1.0

Here is a website I found that can convert between html colors and this format: https://corecoding.com/utilities/rgb-or-hex-to-float.php

It saves you the trouble of using a calculator.

 

[fgh1t6]

1 hour ago, XUTERO said:

No, I extracted VB and IB with the normal "g1m_export.exe". I am unsure what the "g1m_export_with_vgmap.bat" is supposed to do? 
What are VG maps, what and how are they used? 

g1m_export_with_vgmap.bat unify the bone name with 3DMigoto blender plugin. You should always use it to extract VB/IB . After working on a mesh, Clear up all non-bone vertex groups. (Use the  "remove unused vertex group" plug in may help). push space bar type in "assign 3DMigoto vertex groups" to launch the vgmap repair routine. set the "Vertex group step" to 3. Then you can safely export as VB/IB and use g1m_import.exe to import back to g1m file.

[gokuri]

3 hours ago, vagonumero13 said:

 

 

 

Edit: i looked at the pirate .g1t, and it is indeed like that. It has the color data for pure black with full opacity: R = 0.0, G = 0.0, B = 0.0 and A = 1.0

 

 

ah, thanks. full opacity black is not what I expected. the gold must live somewhere else 

 

[fgh1t6]

21 minutes ago, gokuri said:

ah, thanks. full opacity black is not what I expected. the gold must live somewhere else 

 

Extract AYA_HAIR_036.mtl from CharacterEditor and use fid_utility.exe in teh g1m_tool package to convert it to text file. In,sdie it tells you which file set is used.

[gokuri]

3 hours ago, fgh1t6 said:

Extract AYA_HAIR_036.mtl from CharacterEditor and use fid_utility.exe in teh g1m_tool package to convert it to text file. In,sdie it tells you which file set is used.

two 1kb files and the normal looking hair file, no sign of the ears. I extracted all the g1t that weren't 1kb and ¯\_(ツ)_/¯

I'm content to give up if the mystery isn't striking anyone else's curiosity. 

00000_000000-BC4.dds 00000_000000-BC7.dds 00000_000000-DXT1.dds 00000_000000-BC4.dds 00000_000000-DXT1.dds 00000_000000-BC7.dds

[Guest]

9 hours ago, XUTERO said:

No, I extracted VB and IB with the normal "g1m_export.exe". I am unsure what the "g1m_export_with_vgmap.bat" is supposed to do? 
What are VG maps, what and how are they used? 

 

 

My Leifang breakable mod  used  the normal "g1m_export.exe", which was enough for weight transfer.

 

The problem you had should be that the shape  between the orginal mesh and you modded mesh had too much difference;

 

or the number of vertex groups was not right.

 

 

[blackdragon99]

39 minutes ago, gokuri said:

two 1kb files and the normal looking hair file, no sign of the ears. I extracted all the g1t that weren't 1kb and ¯\_(ツ)_/¯

I'm content to give up if the mystery isn't striking anyone else's curiosity. 

There's no texture for it, just the color information, like vago said. The color/reflectiveness etc. of the metallics is determined by the color channels in the kidsrfr.g1t texture i'm pretty sure.

[Joschuka]

On 4/24/2020 at 1:38 PM, fgh1t6 said:

Founding may help for the tool developing:

In the <MeshesSection> section if   "Count2 value" is not 0, adding block to the end may not work.

In the <SubmeshesSection> section, each submesh block has a "<U_00 value="XX" >".  information, U_00 value="53" is the common type, counted by the  "Count1 value" in <MeshesSection> blocks.   U_00 value="61"  is special type, counted by the "Count2 value". In the  <MeshesSection>, all U_00 value="53" type mesh block should go first. That is the "Count1 value" counting. then followed by "Count2 value" blocks  for the U_00 value="61" type submeshes. If inserting U_00 value="53" type mesh, the corresponding <MeshesSection> blocks must insert before blocks for U_00 value="61" type submesh. For example if

            <Count1 value="1" />
            <Count2 value="1" />

There would be blocks: <Mesh idx="0">,<Mesh idx="1">, and if we are inserting a submesh with <U_00 value="53"/>: the two counting would be:

            <Count1 value="2" />
            <Count2 value="1" />

then the newly created block should be <Mesh idx="1"> not <Mesh idx="2">.

 

 

 

Great catch ! I'll check if that info is consistent too across the different games. Been wondering for a while what that value was used for

[perfectdark023]

 

[fgh1t6]

Damn! the animated UV hash code dose not support transparent! It is useless! The only idea I can have is to make a advertisement board!

 

[gokuri]

WIP/ messing around

 

unintentional but neat shimmering glow effect, maybe somebody can do something on purpose with it. note the reflection on her chin/jaw and palm.

 

i should probably hided the buttons, huh?

[vagonumero13]

Did anyone of you succesfully modified kidsmm1 textures? (the transparency textures to show bra in school uniforms and some other costumes).

I remember that a while ago some of you said that you weren't able to see the modifications in game.

 

I'm currently coding a g1t parser for qrdbtool (to show previews), and I noticed these textures have some extra flag header, and I think they actually have more data that what Linkdatatool extracts/replaces. In fact, I think linkdatatool only extracts half of the data...

 

I guess that I will have to reuse my g1t parser to create alternative g1t tools.

 

 

[blackdragon99]

1 hour ago, vagonumero13 said:

Did anyone of you succesfully modified kidsmm1 textures? (the transparency textures to show bra in school uniforms and some other costumes).

I remember that a while ago some of you said that you weren't able to see the modifications in game.

 

I'm currently coding a g1t parser for qrdbtool (to show previews), and I noticed these textures have some extra flag header, and I think they actually have more data that what Linkdatatool extracts/replaces. In fact, I think linkdatatool only extracts half of the data...

 

I guess that I will have to reuse my g1t parser to create alternative g1t tools.

 

 

I could never get edited kidsmm1 files to work, but someone else on here found that renaming a kidsalb file to kidsmm1 and using that instead makes it work in game.

[fgh1t6]

Spoiler

2 hours ago, vagonumero13 said:

Did anyone of you succesfully modified kidsmm1 textures? (the transparency textures to show bra in school uniforms and some other costumes).

I remember that a while ago some of you said that you weren't able to see the modifications in game.

 

I'm currently coding a g1t parser for qrdbtool (to show previews), and I noticed these textures have some extra flag header, and I think they actually have more data that what Linkdatatool extracts/replaces. In fact, I think linkdatatool only extracts half of the data...

 

I guess that I will have to reuse my g1t parser to create alternative g1t tools.

 

 

 

36 minutes ago, blackdragon99 said:

I could never get edited kidsmm1 files to work, but someone else on here found that renaming a kidsalb file to kidsmm1 and using that instead makes it work in game.

 

No luck here neither. I have change all sets but it looks like there is another copy of the original stored somewhere. For these sweating-make-transparent texture that use kidsmm1, there are three special entries in the <MaterialAttributes> :

                <Attribute name="vMMPrms">
                <Attribute name="nMMLocalTCI">
                <Attribute name="vMMBld">
Do know how to use them yet. 

 

By the way, @vagonumero13, if it is not too much to modify, could you make an experimental version of g1m_xml for me? I found that in the <AttributesSection>/<MaterialAttributes> blocks, the <DATA value="..."> is decided by the the <U_0C value /> . It seems a Data type: 1- 4 is an array contains 1-4 float32 numbers. 5 is single INT32. <U_0E value/> is the count of the Type.  For example, <U_0C value="4" /> ,<U_0E value="3" /> would means the data is 3 Arrays each array has 4 float32 data. In the current version, there would be 48 "0x**" in the data string. It would be great if it could be changed to 12 numbers like " 0, 0.3, 1.0, 1.0...".

 

[vagonumero13]

1 hour ago, blackdragon99 said:

I could never get edited kidsmm1 files to work, but someone else on here found that renaming a kidsalb file to kidsmm1 and using that instead makes it work in game.

That makes sense. I guess that's because LINKDATA doesn't change that part of the header I'm talking about, and also because it may keep the other half of the data (the original one) there. I guess I will have to look further into it when I make my g1t tools.