Yep ! https://drive.google.com/file/d/0B0PqQOPaer0bZkt5OVdIRTB5VHc/view?usp=sharing

it depends on the particle system. Usually rotation of the particle is not affected (and that's what I've chosen to do with my rain particle : the 'wind' should keep blowing in the same direction), but position usually follows the parent movement. Overall : yes, it works, but remember to move the slider in and out the shot time window in order for the particle to refresh its position

whichever you want, both work. Don't forget to lock its position to the camera after adding it.

yeah, you're most certainly right, and that's a pity, because the effect is so cool and would be awesome to implement easily into sfm. An easy way to reproduce a physical transformation like becoming a zombie would be having both models at the same spot, and shrinking the first model at every bone (thus not a globally centered shrinking but a bone-centered one) so that the second model ends appearing. Or instead of shrinking making the bones bigger.
But it's true transparency effects like that one are so hard to make in sfm, and easy to reproduce using post-production :(

I bump this thread because it's actually a very interesting question : how did he manage to create a transparency/invisibility gradient ?
Btw, we can see the same effect again aroudn the 3rd minute of the part Two.

https://steamhost.cn/steamcommunity_com/sharedfiles/filedetails/?id=309686213

Done ! :D
Have a look here : https://steamhost.cn/steamcommunity_com/sharedfiles/filedetails/?id=299562235
Tutorial here : https://www.youtube.com/watch?v=tMz_75-nyjk

I'll be the first one giving some very basic tips :

• Getting started
  The scripts are located in
  SourceFilmmaker\game\XXX\scripts\sfm\***
  Where XXX is platform for the default scripts, but can be any game folder listed in usermod\gameinfo.txt like usermod, tf, workshop ...
  and where *** can be
  
  1. the current folder (only used for sfm_init.py I think, i.e., the script executed automatically when launching SFM),
     
  2. mainmenu, to create scripts launched from the Scripts dropdown menu in SFM (you should have there the Examples scripts. A good place to create scripts un related with a specific shot, just with your current session). You need to restart SFM for the new scripts to appear.
     
  3. dag, to create scripts accessed through RMB>Dag Utilities menu. There are two subfolders, exact\countX (when you have exactly X Dags selected (e.g. only X=1 dag selected)), and multi (when you have multiple Dags selected e.g. to create point/aim/orient/parent constraints quickly). New scripts appear automatically (no restart needed).
     
  4. animset, to create scripts accessed through RMB>Rig. That's the one I'm using for my scripts, as you can then get very easily some basic information like your current shot or the "thing" (Animation Set) you have selected. New scripts appear automatically (no restart needed).
     
  The scripts are written in Python (duh), if you don't know this language, here is a summary of some syntax rules (Python files end with .py, I recommend using Notepad++[notepad-plus-plus.org] to edit them) :
  
  • As most languages, it is case-sensitive (pro-tip : variableName, FunctionName)
    
  • No special char to end an instruction, the end of line suffice.
    
  • PYTHON IS ALL ABOUT TABULATION : hierachy is told through tabulations, there are no braces ('{'), only an opening with a colon (':'), and then the tabbed content.
    Example : this C/java code
    
    if (condition) { actionsIfTrue(); } else { actionsIfFalse(); } actionsAnyway();
    becomes in Python
    
    if condition: actionsIfTrue() else: actionsIfFalse() actionsAnyway()
    /!\If you copy-paste my code, you need to replace every 3 spaces by a tabulation /!\
    I did the opposite in order for my code to be good looking, because Steam-code tabulation is too large.
    In Notepad++, \t == 'a tabulation'
    
  • The for-loop in Python works with arrays. Here is an example from the Example scripts of SFM
    
    shots = sfmApp.GetShots() nCounter = 0 for shot in shots: flDuration = shot.timeFrame.duration.GetValue().GetSeconds() if flDuration == 0.0: nCounter += 1 track = sfmApp.GetParentTrack( shot ) track.RemoveClip( shot ) message = "Removed %d shots because of duration 0" % nCounter
  • #This is a comment.
    
  • To create a new function, use the def key word. Example :
    
    def HelloWorld(dumbName): print "Hello "+dumbName+" !" #Your function may or may not return something : return 0
  
  
• Some basic functions
  Of course, you must have a decent understanding of how the Element Viewer works, because Python scripts work messing with it (Dragons, blabla).
  
  If you are testing some code, of course,
  /!\ DO IT ON A BLANK SESSION WITH NOTHING TO LOSE /!\
  
  • Getting the shot you are working on (most important one, as almost all you add/edit will be related to its specific shot) :
    
    shot = sfm.GetCurrentShot() #This way you can get the Shot ID : shot.GetFileId()
  • Getting the model/camera/light/etc. (Animation Set actually) you are working on (the one you have right-clicked on) (very useful one, to edit one element easily) :
    
    animSet = sfm.GetCurrentAnimationSet()
  • Adding a New Model (the equivalent of Create Animation Set for New Model) :
    
    baseName = "secret_spy" #whatever we want. If this name is already taken, it will add "2" at the end of the name. modelPath = "models\\player\\hwm\\spy.mdl" #The short path to the model. If you use an antislash '\', be sure to double it, as it is a special char. newAnimSet = sfmUtils.CreateModelAnimationSet( baseName, modelPath )
    The function sfmUtils.CreateModelAnimationSet is very interesting per se, as it shows a more general method for adding stuff :
    
    def CreateModelAnimationSet( baseName, modelPath ): #We need the Shot ID when adding elements (vs.CreateElement) shot = sfm.GetCurrentShot() #We instance a pointer to the model, but it is not rendered yet and it has no Animation Set linked to it model = sfm.CreateModel( modelPath ) if ( model != None ): #If the model is created successfully ... #... we create an Animation Set for that model. animSet = sfm.CreateAnimationSet( baseName, target=model ) if ( animSet != None ): #The following code adds the model to the scene dag = vs.CreateElement( "DmeDag", baseName, shot.GetFileId() ) dag.AddChild( model ) shot.scene.AddChild( dag ) return animSet
  • Adding a New Light (the equivalent of Create Animation Set for New Light) :
    
    def CreateNewLight(lightName = "light"): shot = sfm.GetCurrentShot() #Create an abstract Light newLight = vs.CreateElement("DmeProjectedLight", lightName, shot.GetFileId()) #Add the light to the scene shot.scene.FindOrAddChild("Lights").children.AddToTail(newLight) #Create an Animation Set for the new Light newLightAnimSet = sfm.CreateAnimationSet(lightName, target=newLight) return newLightAnimSet CreateNewLight("fiatLuxEtLuxFuit")
  • Adding a New Camera (the equivalent of Create Animation Set for New Camera) :
    
    def CreateNewCamera(cameraName = "camera"): shot = sfm.GetCurrentShot() #Create an abstract Camera newCamera = vs.CreateElement("DmeCamera", cameraName, shot.GetFileId()) #Add the camera to the scene shot.scene.FindOrAddChild("Cameras").children.AddToTail(newCamera) #Create an Animation Set for the new Camera newCameraAnimSet = sfm.CreateAnimationSet(cameraName, target=newCamera) return newCameraAnimSet CreateNewCamera("bigBrotherIsWatchingYou")
  • Pasting As Reference (for you guys used to mess with the Element Viewer but new to scripting) : the function is AddToTail(dagElement)
    
  • Making the Selected Camera the RT Camera (e.g. for Portals or mirrors) :
    
    shot = sfm.GetCurrentShot() #We get the AnimationSet selected, and hope it is a Camera animSet = sfm.GetCurrentAnimationSet() if animSet.HasAttribute("camera"): #If the selected Animation Set was actually one of a camera #We get the Camera Dag (DmeCamera) : camera = animSet.camera shot.monitorCameras.AddToTail(camera) #Here is the "Paste as Reference equivalent of the Element Viewer ! shot.activeMonitor = shot.activeMonitor + 1 camera.name = "RTCamera" animSet.name = "RTCamera"
  • Getting all the Animation Sets (taken from revzin.net) :
    
    import vsUtils # Required for iterAttrs to work shot = sfm.GetCurrentShot() # Empty array allAnimSets = [] for shotAttribute in shot.iterAttrs(): if shotAttribute.GetName() == "animationSets": for animSet in shotAttribute: allAnimSets.append(animSet) break #Now all the Animation Sets are in the allAnimSets array : for animSet in allAnimSets : print animSet.GetName()

I did the opposite in order for my code to be good looking, because Steam-code tabulation is too large.
In Notepad++, \t == 'a tabulation'

IK rigging with scripts is nice, but I really think there is a whole world of automating general sfm tasks with Python. However, the official documentation lacks a lot of available functions, as it is only focused on rigging.

So, if there is anyone out there who knows how to do some basic stuff like creating a Camera (I have managed to create one but not to add it to the "Cameras" array, so it's not showing in the viewport dropdown list of cameras), doing the Element Viewer equivalent of "Pasting as Reference"), modifying sample values of the Motion Editor (for instance, someone in the group asked for a script that adds a focal plane bone to a camera, and I really think it would be doable with two scripts : the first one creates that bone/transform control, whose motion editor sample values would then tell the relative distance to the Camera, and then a second script which reads those values and edits the "Focal Plane" values accordingly. But that of course, requires motion editor values reading/writing, and I don't know how to do that),
please publish those basic examples, the Community would really appreciate it (I'd love to create a bunch of scripts for automating tasks, but I'm soslightly stuck right now).

EDIT : Figured out the stroken features while writing the tutorial :D