Install Steam
login
|
language
简体中文 (Simplified Chinese)
繁體中文 (Traditional Chinese)
日本語 (Japanese)
한국어 (Korean)
ไทย (Thai)
Български (Bulgarian)
Čeština (Czech)
Dansk (Danish)
Deutsch (German)
Español - España (Spanish - Spain)
Español - Latinoamérica (Spanish - Latin America)
Ελληνικά (Greek)
Français (French)
Italiano (Italian)
Bahasa Indonesia (Indonesian)
Magyar (Hungarian)
Nederlands (Dutch)
Norsk (Norwegian)
Polski (Polish)
Português (Portuguese - Portugal)
Português - Brasil (Portuguese - Brazil)
Română (Romanian)
Русский (Russian)
Suomi (Finnish)
Svenska (Swedish)
Türkçe (Turkish)
Tiếng Việt (Vietnamese)
Українська (Ukrainian)
Report a translation problem
Do you need the QC and other files as well?
Edit: I'm fixing the legs, so they don't look weird when bending them.
See this
https://i.imgur.com/8ZKXTDV.png
It may help you and rather than add shapekeys to fix mesh for bone movements, it is better to actually fix the weight painting of the mesh to the bones that move them.
Off the topic question: Why don't people make muscle flexing in other softwares then port it to SFM? I never really saw muscle flexing in SFM before.
https://steamcommunity.com/sharedfiles/filedetails/?id=1395889832
https://steamcommunity.com/sharedfiles/filedetails/?id=1503983169
... but my opinion is that there's little point to it other than seeing if it's possible.
The Source engine is an obsolete engine these days, with basic shaders and dated lighting. SFM looks "okay", but its main selling point is being comparatively simple to learn and use - advanced animation like this is hard to set up, and really doesn't add much to the scene other than being an easter-egg for experienced animators to spot. Most of the audience would barely notice.
Good job on that flexing stuff and exactly my opinion on Source engine, it's basic, runs well and can still make a difference. I also think releasing models that have muscle flexing will give experienced SFMers something interesting to work on, other than just posing and doing facial expressions. Who knows how long will SFM last, it's a playground for beginners and kind of an introducution to making stuff.
https://i.imgur.com/5qBKy6U.gif
The "Corrective" flexes in Blender and the "Corrective" flexes referred to in Valve models are kind of not the same and kind of the same. Both use "Drivers" to control what is happening but where they differ is that Blender corrective shapes are used to do things like muscles and detwisting mesh by creating vertex animations based on calculations given to the corrective shape and the blender engine uses these calculations to apply the shape in the animations produced.
The SFM engine can do vertex animations, but the corrective shapes merely limits the amount a shape is allowed to move and writes the calculations into the visible shape. If you move a bone, the bones does not see the shape to apply it. (like a muscle flex bulge.)
However a word of advice here, one thing I noticed in your blend file is that your decompiling and setting your export (work) folder to a USERMOD path. VERY BAD IDEA.
Get away from using USERMOD as a work folder. It will avoid confusioning and breaking SFM because it is trying to read stuff from the work path you're using.
D:\Steam\steamapps\common\SourceFilmmaker\game\workshop\models\spiderman_homecoming\characters
Sorry I was sure it was somewhere into the usermod folder path when I tried the compiling. (may have been workshop as I didn't change the setting until I found that the qc wasn't finding the dmx exports) But none the less, you should avoid using any SFM folder as a work folder. Same reasoning.
Although creating some corrective flexes via the smooth tool may seem easier than creating a good weight paint in the first place, flexes are kind of a one trick pony - they can make only a single shape, whereas bones can be manipulated in any wide variety of ways, so flexes cannot completely compensate for any possible movement of a bone.
While it can be used to correct minor issues, you should really try to correct major issues as well as you can in the weight paint.
Compare your flexes to the flexes I have to do the exact same thing on the model I pictured before.
(Warning: Somewhat NSFW - I've blurred it, but like most of my character models, the base body is nude)
https://imgur.com/yq8ibTP
You can see that the flexes are much less dramatic, because they have less correction to do in the first place, meaning that, should the twist not be exactly the one they were designed to fix, they're still going to be close to the right result.
(Also, Cerys has a major advantage here, as she has twist bones for the top and bottom end of each major limb bone - which is itself unpainted - so the flexes have even less work to do).
Actually, the whole reason she has to be nude there rather than in her underwear ties into this. I've had to redo a lot of the weight paint and several of the corrective flexes because I tried to rely too much on them.
An attempt to fix the posing around the groin area the lazy way resulted in a flex that looked fine when correctly posed, but which created a scrunched up mess when applied to the base pose. The vertex normals from said scrunched up mess were getting compiled into the flex and completely messing up the lighting in that area, kind of defeating the point of improving the pose.
To fix this, I've had to heavily redo the weight paint around that area and then basically junk most of the corrective flexes I created before (because they no longer matched the weight paint), and I've not yet got around to fully transferring and tweaking that new paint to her underwear. Hence why I can't demonstrate the flex with her modesty intact.
Lots of extra work because I was lazy in the first place. If you truly want to create high quality models with advanced features, you really can't afford laziness.
Until I figured out a (partial) workaround fairly recently, I don't think anyone had found an answer to automatically driving flexes based on bone rotation.
$boneflexdriver uses bone translation as their control, and VRD helpers cannot be used to control $boneflexdriver.
Given that a lot of model authors can't be bothered to implement procedural twist bones, the idea of adoption of muscle flexes is... unlikely.