Jump to content

Costume Customizer Mod (v2.5.2) - Added Yukino Support, plus skin/shadow fixes and additional menu.ini fixes


Recommended Posts

Posted
8 hours ago, KuroKaze78 said:

Rejoice! For I finally got around and fixed the issue with the Common body butt. Once again more people brought this up in other topics so even though its been this way since the beginning and I was hoping for someone to kinda implement the Hi-Res mesh UV into the clean common body mesh I didn't really ever see it happen (unless I missed it). Anyhow, I used a vertex-group corrected mesh provided by @IceTree5 (that was actually given to me over a year ago) and spliced it with the nipple and pubic regions of the HI-RES mesh and then proceed to fix all the seams from the splicing and to properly blend with any existing mod-pack body mesh. As a result, you get v1.9.3 to hold you over while I continue working on v2.0. The original issue is primarily from the weight paint as I did all the body mesh splicing and everything then transferred the old weights and it still had the original issue, but since the clean mesh looked nicer in the long-run i decided to finish using the new mesh.

 

From my testing it looks like it should be fine and there aren't any glaring mistakes (keyword glaring), I did notice i have a small impression on the inner thigh but no where near as bad was what it used to, I have some noticeable cut-over from where the HI-RES pubic region meets the original clean mesh, but that was always there on the original mesh so maybe I'm just using outdated skin textures that didn't blend it very well.

 

Please realize that this only fixed the default body mesh if a mod-pack didn't provide their own body mesh. If they did, then they will need to correct it on their side. I am still in the process of generating an updated blender file with the new body mesh and weight paint/vertex groups. In the mean-time (hopefully no longer than a week), you can always transfer weights from the included Common.vb/ib in the shared folder rather than using BodyWeights_Common_HighRes in the v2 blender file.

 

 

 

On the topic of the v2.0 update, I'm in the process of switching how I develop the INI files to use Jinja2 templates. This largely means nothing to everyone else but basically I started driving myself nuts trying to make adjustments in testing since I had to make the same corrections in tons of places either through clever regex replacements or manually. So far I've only converted about 5000 lines of the tenative v2.0's 13000 lines and reduced those 5000 lines to ~250 lines.

 

Butt is improved but this section here is now badly rendered (left leg contacting crotch).

 

You mention,

 

"I did notice i have a small impression on the inner thigh but no where near as bad was what it used to,"

 

but this looks worse than before so I'm not sure if this is what you are referring to.

 

DOAX-VenusVacation_200315_220548.thumb.jpg.0da5529a65b88aa4e491d6e3f8a4bc0d.jpgDOAX-VenusVacation_200315_220630.thumb.jpg.dfcb28a708b112579269fc186cef50f3.jpg

Posted

@Sladesamm Neither of these are what i caught. I noticed a slight difference in the 2nd image but thought that was just the texture, turns out while investigating the 1st image you posted, the 2nd one is actually a hole in the mesh that i need to seal up. Regarding the 1st issue, it's going to be a pain because that's on the seam between the groin/chest mesh and the leg mesh and i have to keep the seam the same as before to be compatible with whats already been released. Therefore, I think my only option to make it not look so out of place is to blend the weight paint more between the new mesh and the old mesh to try and smooth it out some.

 

 

Edit: Can some people please try these out. So far I've only attempted to fix the broken seam/white spot. I've stayed up too late as it is, so I'll try to do something about the outside leg spur tomorrow. To test, just extract the contents in the CostumeCustomizer/shared directory.

 

Edit2: Forgot to remove the extra vertex group in shared.zip for the one person that already downloaded the first one.

shared2.zip 

 

Spoiler

Preview (Please disregard the legs/arms missing)

image.png.c66cc451e17de2aa2a10224686dda19b.png

 

Posted
7 hours ago, KuroKaze78 said:

@Sladesamm Neither of these are what i caught. I noticed a slight difference in the 2nd image but thought that was just the texture, turns out while investigating the 1st image you posted, the 2nd one is actually a hole in the mesh that i need to seal up. Regarding the 1st issue, it's going to be a pain because that's on the seam between the groin/chest mesh and the leg mesh and i have to keep the seam the same as before to be compatible with whats already been released. Therefore, I think my only option to make it not look so out of place is to blend the weight paint more between the new mesh and the old mesh to try and smooth it out some.

 

 

Edit: Can some people please try these out. So far I've only attempted to fix the broken seam/white spot. I've stayed up too late as it is, so I'll try to do something about the outside leg spur tomorrow. To test, just extract the contents in the CostumeCustomizer/shared directory.

 

Edit2: Forgot to remove the extra vertex group in shared.zip for the one person that already downloaded the first one.

shared2.zip 387.92 kB · 4 downloads  

 

  Hide contents

Preview (Please disregard the legs/arms missing)

image.png.c66cc451e17de2aa2a10224686dda19b.png

 

Sorry, for the confusion. Both pictures I posted are the same issue at different angles. I tried to center the issue on the captures. You are addressing my concern so that's good. I think you are seeing the hip sticking out on the 1st picture but that wasn't my issue. That can be worked around with camera angle and/or waiting for different frame of animation. That is more of a limitation of the game so don't lose any sleep over that on my account.

 

Edit: I'm responding to this right before work, so I won't be able to test fix until later. I tested the update and it works great - thanks.

 

On the subject of indents on the bodies. Since you are able to fix the butts, are there plans to do the same with the breasts? There is a similar issue regarding them as well.

Posted

I want to share some new things I found out regarding how the IDandFresnel parameter works in the ps-cb2 constant buffer. From looking at some of the pixel shader hlsl code and some poking around I found these things out:

  • Only x, y, and w are used as found in ps shaders
  • each parameter (x, y, w) are multiples of ~0.0039215 (1/255)
    • Analysis of unique values found in the ps-cb2 buffers I've exported (sample size: ~50,000 ps-cb2 buffers)
      • x appears to be between [0-255] / 255
      • y appears to only appear between [0-1] / 255
      • w appears to only appear between [0-5] / 255
  • x and y together form a sort of key to a dictionary of material properties
    • Outcome using the same key between DMM and Steam produce different results
    • May be tied to the cs-t12 property in the Compute Shader
    • Spoiler

      Resource bind info for gMaterialParameterList
      // {
      //
      //   struct MaterialParameter
      //   {
      //
      //       float4 base_color_;            // Offset:    0
      //       float metalic_;                // Offset:   16
      //       float reflectance_;            // Offset:   20
      //       float specular_scale_;         // Offset:   24
      //       float glossiness_;             // Offset:   28
      //       float4 ibl_diffuse_;           // Offset:   32
      //       float4 ibl_specular_;          // Offset:   48
      //       float4 rim_;                   // Offset:   64
      //       float4 velvety_;               // Offset:   80
      //       float4 velvety_rim_param_;     // Offset:   96
      //       float4 ambient_;               // Offset:  112
      //       float emissive_;               // Offset:  128
      //       float light_scatter_out_scatter_rate_;// Offset:  132
      //       float light_scatter_depth_offset_;// Offset:  136
      //       uint flags_;                   // Offset:  140
      //
      //   } $Element;                        // Offset:    0 Size:   144
      //
      // }

       

      // Resource Bindings:
      //
      // Name                                 Type  Format         Dim Slot Elements
      // ------------------------------ ---------- ------- ----------- ---- --------
      // gMaterialParameterList            texture  struct         r/o   12        1

      • Inspection of the cs-t12 buffer showed it has the space to contain up to 4608 different Material Parameters
        • DMM contains 1086 non-zero material parameters
        • Steam contains 1081 non-zero material parameters
      • Some of the non-zero material parameters seem like faulty values or possible uninitialized memory space
        • DMM contains 438 material parameters before the first "faulty" looking material parameter
        • Steam contains 433 material parameters before the first "faulty" looking material parameter
        • These could correlate to YX key: 0x00 0x00 - 0x01 0xB5 encoded as floats of # / 255 and that could co-incide with what's been witnessed.
      • So far, I've been unsuccessful in finding a way to override the cs-t12 buffer, but I haven't tried much yet.
  • impact of the w value not yet fully understood

 

I created a simple custom shader to increment the valid values of x/y and found that there can be some crazy material types that can have a big impact on the resulting texture. Some of the settings also caused the material to strobe at various rates (not that I can capture that in a screenshot) or glow.


 

Spoiler

 

Normal:

image.png.c5fcb0e82f4f621aa7b903f8ffd088c4.png

 

Different IDandFresnel settings:

image.png.c71d57f8e5cfd3e5e59558a6f1e03e54.png

image.png.072501732853e2ea943a29bb3fb5748b.png

image.png.783530a166f55e740a87455099b59648.png

image.png.3f866832d4e799fe40f38ff8571f1976.png

 

 

Still working on how the x+y value corresponds to what values so that I can try to compensate/account for when the keys change, as I've noticed some that use the ps-cb2 for their reflection/glossy properties change every so often when the game is updated likely because the material parameters shift.

Posted

 

On 3/29/2020 at 12:37 AM, KuroKaze78 said:

I want to share some new things I found out regarding how the IDandFresnel parameter works in the ps-cb2 constant buffer. From looking at some of the pixel shader hlsl code and some poking around I found these things out:

  • Only x, y, and w are used as found in ps shaders
  • each parameter (x, y, w) are multiples of ~0.0039215 (1/255)Analysis of unique values found in the ps-cb2 buffers I've exported (sample size: ~50,000 ps-cb2 buffers)
    • x appears to be between [0-255] / 255
    • y appears to only appear between [0-1] / 255
    • w appears to only appear between [0-5] / 255
  • x and y together form a sort of key to a dictionary of material properties
    • Outcome using the same key between DMM and Steam produce different results
    • May be tied to the cs-t12 property in the Compute Shader
    •   Reveal hidden contents

      Resource bind info for gMaterialParameterList
      // {
      //
      //   struct MaterialParameter
      //   {
      //
      //       float4 base_color_;            // Offset:    0
      //       float metalic_;                // Offset:   16
      //       float reflectance_;            // Offset:   20
      //       float specular_scale_;         // Offset:   24
      //       float glossiness_;             // Offset:   28
      //       float4 ibl_diffuse_;           // Offset:   32
      //       float4 ibl_specular_;          // Offset:   48
      //       float4 rim_;                   // Offset:   64
      //       float4 velvety_;               // Offset:   80
      //       float4 velvety_rim_param_;     // Offset:   96
      //       float4 ambient_;               // Offset:  112
      //       float emissive_;               // Offset:  128
      //       float light_scatter_out_scatter_rate_;// Offset:  132
      //       float light_scatter_depth_offset_;// Offset:  136
      //       uint flags_;                   // Offset:  140
      //
      //   } $Element;                        // Offset:    0 Size:   144
      //
      // }

       

      // Resource Bindings:
      //
      // Name                                 Type  Format         Dim Slot Elements
      // ------------------------------ ---------- ------- ----------- ---- --------
      // gMaterialParameterList            texture  struct         r/o   12        1

      • Inspection of the cs-t12 buffer showed it has the space to contain up to 4608 different Material Parameters
        • DMM contains 1086 non-zero material parameters
        • Steam contains 1081 non-zero material parameters
      • Some of the non-zero material parameters seem like faulty values or possible uninitialized memory space
        • DMM contains 438 material parameters before the first "faulty" looking material parameter
        • Steam contains 433 material parameters before the first "faulty" looking material parameter
        • These could correlate to YX key: 0x00 0x00 - 0x01 0xB5 encoded as floats of # / 255 and that could co-incide with what's been witnessed.
      • So far, I've been unsuccessful in finding a way to override the cs-t12 buffer, but I haven't tried much yet.
  • impact of the w value not yet fully understood

 

I created a simple custom shader to increment the valid values of x/y and found that there can be some crazy material types that can have a big impact on the resulting texture. Some of the settings also caused the material to strobe at various rates (not that I can capture that in a screenshot) or glow.


 

  Reveal hidden contents

 

Normal:

image.png.c5fcb0e82f4f621aa7b903f8ffd088c4.png

 

Different IDandFresnel settings:

image.png.c71d57f8e5cfd3e5e59558a6f1e03e54.png

image.png.072501732853e2ea943a29bb3fb5748b.png

image.png.783530a166f55e740a87455099b59648.png

image.png.3f866832d4e799fe40f38ff8571f1976.png

 

 

Still working on how the x+y value corresponds to what values so that I can try to compensate/account for when the keys change, as I've noticed some that use the ps-cb2 for their reflection/glossy properties change every so often when the game is updated likely because the material parameters shift.

@KuroKaze78

 

the new fix on the butt line is perfect, thanks for the great work

 

but i got an issue with momiji and other non HD hi-metal skin on the breast

 

theres like an angle on top of the breast, anyway can be fixed?

 

thanks man

DOAX-VenusVacation_200329_181945.jpg

Posted

@ayawrxsti - Turns out that has always been there to an extent but it wasn't as noticeable in the old mesh and a bit more noticeable in the newer clean meshes. However, I think i minimized the impact now.

 

image.png.571c65744182c36c290121c8a5e0598d.png

 

I've also packaged up a clean release with the previous patch "shared2.zip" and there was a minor issue pointed out about saving settings when selecting mod-packs 12-15 that's been corrected in version 1.9.4. Hopefully this will be the last of the model corrections for a bit so i can actually prep a new pre-staged blender file.

 

 

Posted
6 hours ago, KuroKaze78 said:

and there was a minor issue pointed out about saving settings when selecting mod-packs 12-15

Oh, so it was a real issue after all. Thanks goodness, because I thought it was something insane on my end when I wasn't able to save any setting in these particular costume slots. Thanks for the fix. 

Posted
7 hours ago, KuroKaze78 said:

@ayawrxsti - Turns out that has always been there to an extent but it wasn't as noticeable in the old mesh and a bit more noticeable in the newer clean meshes. However, I think i minimized the impact now.

 

image.png.571c65744182c36c290121c8a5e0598d.png

 

I've also packaged up a clean release with the previous patch "shared2.zip" and there was a minor issue pointed out about saving settings when selecting mod-packs 12-15 that's been corrected in version 1.9.4. Hopefully this will be the last of the model corrections for a bit so i can actually prep a new pre-staged blender file.

 

 

omg i just cant thank u more of this:)

 

great work man

Posted

Hey everyone, I'm trying to teach myself how to do some modding using CC.  I'm using the tutorial and I'm stuck on the section Fixing seams on Body Meshes

 

As you can see in the attached picture, I'm trying to get rid of the "seam" that separate my modified common groin mesh from the common breast and common legs meshes.  I followed the guide to the letter as far as I can tell, but instead of mending the seams it just obscures the faces of my modified groin mesh.... as I hope you can see in the picture.  Any advice?  Did I miss a step?

 

Just to confirm, this step is telling me how to fix the gap between the groind and leg meshes one can see if the second photo, right?

help.jpg

help.jpg

Posted

@Geradness - There are probably several different ways to address this. It does kind of depend on how similar the vertices are. If the number of vertices and their intended position are pretty much the same and are just slightly off, you just have to make sure the vertices line up, which looks like that's the case.

 

Spoiler

Mis-algined vertices

image.png.e1056b2a71af6307d45e1753ec0fa881.png

 

Go to Edit Mode on the one you need to correct. Select all the vertices for the seam.

image.png.b8a0ed92886ec99830aca6f7ae45a867.png

 

Go to the Properties > Data > Vertex Groups and add a new Vertex Group. Assign the selected vertices to the group by selecting the "Assign" button afterwards

image.png.27bf0cbc92444dec8b01a622d38f54ee.png

 

Go back to Object Mode, make sure the mesh you need to line up is also visible.

image.png.0a17ee8bd8cba24150dde8a3d3913739.png

 

Add a new modifier to the imported mesh: Shrinkwrap, Set the Target to the mesh you need to line up to. Set the Vertex group to the new group just created. Set the  mode to "Nearest Vertex"

image.png.78aaa8c25a20a8049aa93d3cf072cc14.png

 

Make sure it looks good and that you didn't miss a vertex accidentally (if so, just select it and "Assign" it to the vertex group like before). Once everything looks good. You'll need to Apply the modifier and delete the vertex group that was created (since it will mesh with the weight paint if exported with it still created).

 

image.png.98548831a07a2f36a209d33dd0463d2b.png

 

Hopefully this helps. I'll likely add these tips in the modder section later on.

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now
  • Recently Browsing   0 members

    • No registered users viewing this page.
×
×
  • Create New...