No. Source uses VTF-format textures, and there doesn't seem to be a Blender plugin that imports VTFs into Blender. And if it could, Blender would have to find the VMT material files of the model (which are typically very far from the QC/DMX/SMD files), look for $BaseTexture in the VMTs, and load the VTFs specified by $BaseTexture (which are sometimes not even in the same directory as the VMTs). Even then, what about VMTs that use special options, like the Refract shader or completely overriding $BaseTexture with $Detail?

So no, even with a VTF import plugin, it's simply not possible to automatically do this with the Cycles renderer, Eevee renderer, nor "Blender" renderer, unless you're okay with Blender Source Tools looking through every single folder on your computer until it finds what it's looking for, making it possibly take hours to import a model.

引用自 Medceplor：

- from Insurgency - Here is the .mdl and the .qc -Links removed from quote-

Insurgency costs money, so releasing those files yourself to the public might be considered semi-piracy (yes, even if you don't release all of the game's files).

Even then, the MDL file is almost completely worthless without at least the VVD file and the DX90.VTX (or in some cases just VTX) file, and the QC file is almost completely worthless without all of the DMX/SMD/VTA/et cetera files that it references.


Anyway, in Crowbar, did you have the option to de-compile sequences/animations enabled when de-compiling the model? And what happens if you try to just import a de-compiled DMX/SMD file in Blender instead of the de-compiled QC file?

引用自 Rupert：

- I also have a _mask texture as well so I can have reflective textures alongside the matt ones. Is there anything I need to know about that? -Images-

That the same things apply. You should make sure that the masking VTF also has "Point Sample" enabled and doesn't have any mip-maps.

引用自 Rupert：

Point Sample flag is unchecked and it says in the info that the VTF has 10 Mipmaps -Image-

Then I suggest converting the image to VTF again, this time without generating any mip-maps, followed by enabling the "Point Sample" flag.

Mip-maps do more harm than good to "palette" textures (they stretch the image to a smaller resolution to allow texture quality to be lowered from distances or with low-quality settings, which is fine for "textured" textures but really bad for "palette" textures), and the "Point Sample" texture flag disables texture filtering (for example, with "Point Sample" disabled, a red pixel will smoothly fade to a blue pixel beside it, but with "Point Sample" enabled, the entire pixel of the texture will be fully red without any sort of fading into the blue pixel beside it).

Do you have the "Point Sample" flag of the VTF enabled, and have you made sure that the VTF does not have any mip-maps?

Immediately above the shape key amount value, there are 3 buttons near the right. The left-most of those (to the right of the "Relative" tick-box) should have a tool-tip saying "Always show the current Shape for this Object". Disable that option.

If it's not against Steam's rules (nudity and/or a model from a paid-for game (piracy)) and isn't illegal (a model from a paid-for game (piracy)), it might be a good idea to link to the SMD file here for us to examine ourselves.

引用自 Jax Bline：

it looks normal in sfm just spikey in blender, when I port it back to SFM will it look normal again?

No. Again, if you export something from Blender and it looks different in Source Filmmaker after being exported from Blender, that's a bug. So again, if it looks "spiky" in Blender, it will/should still look "spiky" once you export it and compile it for Source Filmmaker. This is regardless of whether or not it looked "spiky" before de-compiling it.

If it looks different in Source Filmmaker than in Blender (especially when using DMX flexes rather than VTA flexes), that's a bug. So you should expect it to look "spiky" in Source Filmmaker as well if it looks spiky in Blender. Yes, VTA flex importing using the Blender Source Tools is a little wonky sometimes.

引用自 Artfunkel：

You can select the bones you want, switch your pivot point to individual origins, then scale them. -

Won't scaling be based on the current distance (such that something like 1.2 units away more or less can't be (properly) matched to something that's around 1.0 units away, and I won't know the exact size of in order to apply a correct-ish scaling amount)?

引用自 Artfunkel：

- That's much better than having to remember which setting to use when first importing the file.

Depending on what you mean...
- If you're thinking about people not knowing how it works and/or what to do with the default option, I'd of course still want the default to be based on the model's overall size.
- If you're thinking about me, as an individual, remembering what size to use, I'd probably always have the bone size at 0.25 or 0.5 units, both of which I find pretty easy to remember (and can see if I import the wrong one of so I can re-import while using the other option).